Review of Business and Economics Published quarterly Volume LIII
number 3
July-September 2008
CONTENTS Law and Economics in the Future of Legal Scholarship, Education, and Practice Th.S. Ulen
218
The Role of SMEs in Innovation in the EU: A Case for Policy Intervention? R. Veugelers
239
Corporate Governance and Earnings Management: Evidence from Europe A. Renders & S. Vandenbogaerde
263
The Business Failure Process: An Integrative Model of the Literature N. Crutzen & D. Van Caillie
288
Innovatief kosten/baten model voor optimalisatie van beslissingen aangaande veiligheidsinvesteringen in een chemische onderneming G.L.L. Reniers & A. Audenaert
317
Book Reviews
347
Editor-in-Chief: C. Crombez Naamsestraat 69 3000 Leuven ISSN 0772-7674 PB BC 30664 P808087 Afgiftekantoor: LEUVEN MAIL
Cover design: Danny Juchtmans © 2008 by Acco No part of this review may be produced in any form, by mimeograph, film or any other means without permission in writing from the publisher.
2008 / 3
Law and Economics in the Future of Legal Scholarship, Education, and Practice Th.S. Ulen
1
ABSTRACT Law and economics has become the dominant style of legal scholarship in the United States. This Article considers and illustrates the ways in which law and economics has affected and will affect legal scholarship, education, and practice. After first defining law and economics as a method by which to predict and measure law’s impact on behavior, the Article makes the claim that the central, lasting impact of law and economics on the field of law is through its importation of the scientific method into legal inquiry. The Article illustrates this claim through theoretical and empirical examples.
Review of Business and Economics
EL CODES
KEY WORDS
RESEARCH FUNDING
I.
A2, B4, K00 Law and Economics Swanlund Chair Fund, University of Illinois at Urbana-Champaign
Introduction
Let me begin with my warmest thanks to the Katholieke Universiteit Leuven for the honor they bestowed on me on October 5, 2007 in awarding me the degree of Doctor Honoris Causa. This means the world to me. I have a very deep – nearly religious – reverence for the power of reason to help us understand this world and make it better. And I have the greatest imaginable respect for those who have used their gifts of reason and creativity to advance human understanding and wellbeing, as has been going on in this great university since the 1420s. Great research universities are to me among the most precious institutions in our societies, the institutions of which we should be the most proud and the most reverential. So, when a great institution such as the Katholieke Universiteit Leuven chooses to honor me in this fashion, I feel that I, too, have contributed my small grain of sand
Law and Economics in the Future of Legal Scholarship, Education, and Practice
219
II.
2008 / 3
to the wonderful, enlightening, ongoing, and collaborative work of the scholarly community. In this essay I seek to use my nearly 30 years of experience in legal scholarship to speculate on the role that the field of law and economics will play in the future of legal scholarship, legal education, and the practice of law. Inevitably I shall focus on the contribution that law and economics has made to the academic study of law. But as my title indicates and as I deeply believe, what we do in the legal academy eventually has great effects on the practice of law and lawmaking. I hope that the connection between what I shall describe in the legal academy and the practice of law and of lawmaking will be evident.
The Impact of Law and Economics
Review of Business and Economics
Over the past 30 years I have had the great good luck to have participated in an intellectual revolution. That revolution is, of course, the one being effected in the legal academy and beyond by law and economics. That revolution is not complete. Indeed, in important senses it has just begun. And it is helpful, in assessing where we are and where we might go, to bear this relative youth in mind. When I was a graduate student in economics in the mid-1970s, no one in the Department of Economics at Stanford University seemed to know or care about the economic analysis of law. We had a classmate, Abbott B. “Tad” Lipsky, Jr., who was simultaneously doing a Ph.D. in economics and a J.D. That seemed to us to be faintly odd. It certainly necessitated his being absent from economics for long periods of time. On the rare occasions that we asked ourselves why someone would pursue joint degrees in law and economics, we contented ourselves with the thought that Tad must have a deep interest in antitrust issues. To my knowledge, Tad was the only joint J.D.-Ph.D. degree student in the several cohorts of economics graduate students at Stanford of whom I was aware.2 The first time I heard of the new intersection between law and economics (beyond the obvious ones in taxation, antitrust, and government regulation of natural monopolies) was at a reception for new professors at the University of Illinois. A new Assistant Professor of Law, Dan Farber, who has since gone on to a marvelous scholarly career and is currently the Sho Sato Professor of Law and Director of the Environmental Law Program at the University of California, Berkeley, School of Law, told me at that reception that I might be interested in a new book by Professor Richard A. Posner of the University of Chicago Law School entitled Economic Analysis of Law. I found the book in the University of Illinois Library and used it in some research that I was doing on the origins and effects of the Interstate Commerce Commission. I was so taken by the insights of the chapters on regulation that I read other parts of the book and was enchanted. Shortly thereafter, in Sum-
Review of Business and Economics
2008 / 3
220
Th.S. Ulen
mer 1980, I had the good fortune to be invited to a two-week seminar on law for economists put on by Professor Henry Manne, then of the Center for Law and Economics at the University of Miami Law School and later Dean of the George Mason University School of Law, and I was hooked. Upon returning from the law-and-economics boot camp, I put in a proposal to the Department of Economics at Illinois to teach an undergraduate course in this new field. Over the next several years I continued to teach the course (to an enthusiastic group of economics students), to develop a graduate course in law-andeconomics, to visit the University of California-Davis and teach both undergraduates and law students in the new field, and then to join the faculty of the University of Illinois College of Law in 1982 to teach law and economics to law students.3 Over the course of the ensuing nearly 30 years, law and economics has become one of the great success stories in modern legal scholarship.4 In this essay I seek to demonstrate the profound manner in which law and economics has altered legal scholarship, the more profound changes that are on the horizon, and the challenges that this particular revolution will present to the legal academy in the future. Even though this innovation has been, to this point, largely, although not exclusively, of interest within the legal academy and among legal scholars, I believe that it will eventually have important effects on the practice of law and on other areas of lawmaking. I further believe very deeply that this intellectual revolution will eventually have as profound an effect in other countries as it has had over the past 30 years in North America. Indeed, there are signs that that is already happening.5 Let me foreshadow my main conclusion. I do not think that the most important contribution of law and economics to law has to do with particular results (such as clarification of the distinction between the negligence and strict liability standards for determining liability in tort) or particular analyses of areas of law (such as the economic analysis of contract law) or doctrines within areas of law (such as the “true” meaning or real consequences of vicarious liability).6 Rather, I think that the most lasting effect of the revolution wrought or being wrought by law and economics on legal scholarship is methodological: its importation into law of the scientific method of inquiry. It is that alteration in the process by which we reach legal conclusions that is the most profound change going on in almost every law school today in the United States and, increasingly, around the world. What is law and economics all about? Law – to use economic jargon – creates incentives for people to behave in particular ways so as to realize their own and society’s goals. Economics is one of the fields of scholarly inquiry that investigates how people are likely to respond to incentives, and so, economics seems a natural place to look to examine whether the incentives created by law have their intended effects on targeted behavior. Suppose, for example, that the law instructs people that if they fail to take reasonable precaution and if there is an accident, then they will be held financially liable for the accident’s costs. It is not simply enough to announce the law in order for it to be obeyed. There may have been a day when an-
Law and Economics in the Future of Legal Scholarship, Education, and Practice
221
2008 / 3 Review of Business and Economics
nouncing a law was sufficient to induce compliance, but I very much doubt it. And yet traditional legal scholarship frequently seems to believe that the principal task of an academic lawyer is the coherent articulation of the law – once that is achieved, compliance seems to be presumed. Economics recognizes that coherent articulation is not enough. Human beings are extremely complex creatures, and a significant part of their complexity consists in confounding simple predictions about their behavior. So, simply because the law demands some action or inaction does not mean that real people will comply. Law certainly ought to be concerned with the question of whether people will comply with the law’s commands. This is where law and economics comes in. It does so in two steps, First, borrowing from economics, law and economics begins with a theory of how human beings make decisions – how, for instance, they learn about the law and decide whether to obey the law. How, that is, would they respond to the incentives that the law creates? Second, having made a prediction about how people, including those in groups and formal organizations, will respond to the law, law and economics, again borrowing from economics, provides methods for investigating empirically the extent to which the law has achieved its desired social objectives. This view of the role of law and economics brings into clear focus the connection between this modern field and the earlier scholarly innovation of “legal realism.” That phrase may have unfortunate meanings in European jurisprudence in that it is redolent of Nazism. But in American jurisprudence “legal realism” refers to an extremely influential development of the 1920s and 1930s. The American proponents of that new view were reacting to what they considered the stultifying aspects of legal formalism and particularly to the fact that formalism seemed to place doctrinal coherence far above actual social effects in legal analysis. The legal realists, for example, famously argued that the formalists’ concern with whether statutes limiting the hours that women could work in bakeries in Oregon were impermissible violations of the doctrine of freedom of contract inappropriately ignored the potentially adverse social consequences of having women (and children) working outside the home for a full workday.7 The point was not so much that the social effects of law were the only criterion by which to judge a law’s worth but that these effects deserved consideration every bit as much as (and sometimes more than) did doctrinal coherence. This concern with law’s actual effects on real behavior has become a standard part of North American legal analysis. In a real sense, all law professors in North America are (and have long been) legal realists. Nonetheless, from the beginnings of legal realism in the 1920s till the 1980s, there were no standard tools by which students of law judged the social effects of law. One of the most important contributions of law and economics to the study of law is the fulfillment of the legal realist goal of having a standardized set of tools to judge law’s actual effects. There is another important but unintended consequence of the fact that lawand-economics has provided a methodology for gauging law’s actual effects in the
Review of Business and Economics
2008 / 3
222
Th.S. Ulen
real world: that methodology is not tied to any one national or regional legal jurisdiction. Rather, that methodology is applicable to any legal system. This fact makes possible a cross-jurisdictional conversation about the law that was heretofore difficult, if not impossible. Doctrinal legal scholarship seems to be tied to particular state, regional, or national legal systems. German legal academics write almost exclusively about German legal issues; Vietnamese legal scholars write almost exclusively about Vietnamese legal matters. The result is that a doctrinal scholar in Moldova who is studying some legal subject in Moldova has very little of interest to say to a doctrinal scholar studying that same legal subject in New South Wales, Australia. Doctrinal legal scholars do not have a commonality of scholarly interest with other doctrinalists around the world. It is generally (although not exclusively) the particularities of time and place that interest doctrinal legal scholarship. It is not the generalities that span time and space that are of central interest to them.8 Contrast this state of affairs with that in almost any other academic discipline within the modern university.9 Scholars in the same subject areas – such as political science, medicine, anthropology, physics, and fine art – communicate frequently and to productive effect across national boundaries. Indeed, to academics in most of the modern university the notion of a jurisdictionally bound academic discipline would seem puzzling. There could, of course, be several good reasons for this difference between law and other academic disciplines. First, it might be the case that the study of law is so closely tied to the vocation of the practice of law that in every country the principal constituency for legal scholarship is the local practitioner community. Two pieces of casual empiricism support this contention. The first is the obvious fact that the vast majority of students of law in every country are going to be practitioners, not academics. Thus, those students demand and their professors supply very practical instruction that furthers the students’ desires to become gainfully employed as lawyers upon graduation. The second piece of casual empiricism in support of the proposition that law is different from other areas within the university is the practice common in many, if not most, countries of law professors’ simultaneously working as lawyers. If, for example, a law professor in Country A spends half of each workday at the law school (teaching, attending committee meetings, advising students, and the like) and the other half practicing law (meeting and counseling clients in his or her office, appearing in legal fora on behalf of clients, writing briefs, instructing associates, and the like), there is precious little time to left to devote to scholarship (although it is a remarkable fact that a few who do both of these jobs produce elegant and copious scholarship). It may well be the case that this dual-career aspect is motivated by a deep belief that legal education is and should be vocational education – the imparting of a set of skilled tools by master craftsmen to young applicants who want to join the mas-
Law and Economics in the Future of Legal Scholarship, Education, and Practice
223
2008 / 3 Review of Business and Economics
ters’ guild. One can imagine that a few other disciplines within the university – academic medical doctors, professors of education, social work professors, and academic veterinarians, among others – have a similar principled view about the inherently vocational aspect of their disciplines. There is, no doubt, in each of these areas, including law, a scholarly aspect of what professors do. Nonetheless, the vocational-education concerns seem to dominate the scholarly concerns. An additional explanation for the dual-career proposition – one that works with, not against, the prior point regarding the principled focus on the imparting of skills – might be that law professors’ salaries are low compared to what a practitioner can earn. One might argue that this difference in pay leads to either a separating equilibrium in which one of two things is true: (1) the less talented or productive teach while the more talented and productive practice or (2) those who prefer the life of a professor teach while those who prefer the life of a lawyer, practice. Alternatively, there might be a mixed equilibrium in which some teach exclusively, some practice exclusively, and some both teach and practice.10 One reason for a mixed equilibrium might be that even if it does not generate a substantial income, being a professor may generate great prestige, and people place a high value on prestige. People desire to be esteemed, and being a member of the professoriate frequently confers esteem.11 If, however, that is all that the title of “Professor” provides,12 then the title-holders must seek other, supplemental employment to generate income. Presumably, they find employment as practicing lawyers. And it is probably the case that the dual-careers are mutually enhancing: the professor’s prestige is enhanced by his or her being a prominent practitioner, and the lawyer’s attractiveness as an advocate is enhanced by his or her also having the prestigious title of “Professor.” One might speculate that there comes a time when the remuneration as a professor, while not exceeding that available in practice, becomes sufficiently high that more and more lawyers choose to abandon the dual-career track, practice less or not at all, and devote their energies entirely to being a professor. I do not know why it should happen that the relative income of the professoriate should rise in comparison to that of practitioners. Perhaps that happens because the nation recognizes that education is a tremendously important aspect of economic development and devotes more resources to higher education; perhaps it happens because professors are so poorly compensated that they are leaving the university, and the authorities of the universities must raise professorial incomes in order to attract talented faculty; or perhaps it happens because the domestic market for legal advice has become so competitive that the average earnings of practicing lawyers fall.13 Whatever the reason for rising relative incomes of professors, the result may well be the slow creation of a full-time legal professoriate. And once there is not the close relationship between the academy and practice that characterized the earlier mixed equilibrium phase, then what professors of law do and what practitioners do may begin to drift apart.14 It is well beyond the scope of the current article
Review of Business and Economics
2008 / 3
224
Th.S. Ulen
to explain the dynamics of the relationship between theory and practice in professional education, but I have little doubt that there is a fascinating study to be done on these matters. For the purposes of the present article, let us simply leave the matter in this unsatisfactory state: there are some countries in which there is a separating equilibrium between law professors and law practitioners (with the two occupations not overlapping very much) and others in which there is a mixed equilibrium (in which some practitioners are sometime professors and almost all professors are sometime practitioners). The clear implication is that in countries with a separating equilibrium between the academy and practice, law professors will be more free to consider subjects not obviously related to the practice of law. Whether they will in fact do that depends on a host of other factors, such as the presence of competition among law schools within the country and across national borders for productive and novel professional output. I have been suggesting some plausible reasons why the academic aspects of law should be so geographically specific and not transcend state, regional, and national boundaries. Ultimately, I do not believe that these reasons describe an immutable situation. That is, as I shall shortly explain in more depth, I do not believe that there is anything inherent in the field of law that prevents it from being a cross-national academic subject like any other academic discipline.15 Rather, I believe that the “localism” of much doctrinal legal scholarship may be viewed as an early stage in the life cycle of many academic disciplines. Let me illustrate this point by describing an international academic discipline and thereby contrasting it with the “localism” of doctrinal legal scholarship. Consider the field of economics. No one today believes that there is an economics of Belgium and a different economics of Russia and yet another economics of Argentina. Economic theory is the same all over this globe. Moreover, the economics of today are not only for explaining today’s economies. They can also help to explain historical episodes. In a nutshell, the tools of economics are global, not local, and timeless, not bound by a particular era or context. Students who study economics anywhere on the planet learn exactly the same subject, much as do students of physics. One result of this universality of the subject matter is that scholars who profess economics can communicate on any topic within the discipline across national boundaries. Most obviously, a specialist on macroeconomics is studying and writing about the same general topic whether he or she is in Canada, China, South Africa, or Scotland. Even where the economist is studying a particularly country – say, the efficiency of the health care system of Ukraine, her analysis is perfectly intelligible (and, indeed, of interest) to economists who focus on the efficiency of health care in Paraguay or Japan. One possible explanation for this difference between economics (and other academic disciplines) and law (and the other vocation-oriented academic disciplines) may be historical. Specifically, it may be the case that academic disciplines always
Law and Economics in the Future of Legal Scholarship, Education, and Practice
225
I think that in the modern legal academy, law and economics is playing the role of Pareto and traditional doctrinal scholars are playing the role of Schmoller. And I am confident that the role of Pareto will triumph in law, just as it has in economics.
III.
Theory in the Law
My central assertion is that the principal influence that law and economics is and will have on the academic study of law is the importation into the field of the scientific method of inquiry. There is an entire field of inquiry into the philosophy of science that wrestles with issues of just what the “scientific method” is.19 For my purposes here, I want to summarize the complexities of this matter by simply as-
Review of Business and Economics
Giving a lecture before a convention of scientists at Geneva, Pareto was interrupted from the audience by a patronizing call from Gustav von Schmoller, an economist of the then German Strassburg [sic]. “But are there laws in economics?” Schmoller had no personal acquaintance with Pareto at the time. After the lecture Pareto recognized the heckler on the street and sidled up to him in his shabby clothes and in guise of a beggar: “Please, sir, can you direct me to a restaurant where one can eat for nothing?” “Not where you can eat for nothing, my good man,” the German replied, “but there is a place where you can eat for very little!” “So, there are laws in economics!” laughed Pareto as he turned away.18
2008 / 3
or almost always begin as geographically localized studies and that only as they mature do they become more global and less context-dependent. If so, then it might be the case that the global and timeless aspect of economics is relatively new. In fact, about 100 years ago the field of economics was going through the same sort of debate about global versus local that seems to characterize the debate in law between law and economics and traditional doctrinalists. Here’s a true story that illustrates this. Around 1900 two economists – Gustav von Schmoller, a German scholar and the leading proponent of what was called the “German Historical School” of economics, and Vilfredo Pareto, a Swiss-Italian scholar and the leading proponent of the axiomatic-deductive “Austrian School” of economics – were famous in Europe for their contrasting views of the appropriate basis for the study of economics. Schmoller held that to understand a nation’s or region’s economy, one had to know its history, its institutions, its political arrangements, and many other details of its current context. He held, in a famous phrase, that “there are no universal truths in economics; only local truths.”16 In contrast, Pareto (who was trained as a lawyer and an economist) held that although the local details might help to explain the particularities of an economy, there were general properties that every economy shared, such as that factors of production that were in abundance and, therefore, inexpensive, would be more widely used.17 Here is how Pareto recalled one of their encounters:
Review of Business and Economics
2008 / 3
226
Th.S. Ulen
serting that the scientific method consists of two steps: first, the articulation by the investigator of a coherent hypothesis derived from an overarching theory (to take a legal example, increasing the money damages for breaching an enforceable contract will lead to fewer breaches and fewer consensual agreements), and second, gathering data (from actual behavior, questionnaires, experiments, and the like) designed to see whether the hypothesis is correct. I shall contend in this section that this method of inquiry has not been common in legal analysis until recently and that the principal reason that this method of inquiry into legal matters has become increasingly common is that those who profess law and economics have imported this method into their legal analyses. The first aspect of the scientific method that law and economics has brought to the study of law is theory – specifically, a theory of human decisionmaking. Let me illustrate how this commitment to theory and empirical work in law and economics has contributed to our understanding of the law. One of the most famous early theories in law and economics was the Coase Theorem.20 That theorem holds that where transaction costs – the costs of searching for someone with whom to bargain, negotiating with them once you have found them, and then monitoring and enforcing the deal that has been agreed upon – are zero, law is ineffectual and superfluous. An efficient result will occur, regardless of the law. The theory of this prediction was straightforward and not much contested. But there were relatively few examples of the theorem to be found in the legal context. In the early 1980s Professor Robert Ellickson, then at Stanford Law School, sought to test the Coase Theorem by investigating the practices of cattle ranchers and farmers in Shasta County, California, with respect to liability for damage done to farm property by unsupervised cattle.21 Shasta County presented ideal conditions for a natural experiment of the Coase Theorem because the law of liability for damage done by unsupervised cattle was different in the eastern and western parts of Shasta County. In the western half, cattle ranchers were liable for damage done to farm property by their cattle. In the eastern half, cattle ranchers were not liable. Ellickson theorized that if the costs of bargaining between ranchers and farmers were equally low in the two halves of the county, then the actual practices in the two halves of the county would not be different even though the law on the books was different. Moreover, the resulting practices – hypothesized to be the same in the two different parts of the county – would be efficient. When he went to investigate, Ellickson was amazed at what he found. Neither ranchers nor farmers throughout the county knew the law regarding legal obligations for stray cattle; more surprising still, attorneys in private practice in Shasta County did not know or were mistaken about the law. Apparently, ranchers and farmers in the county did not know the law because they were not seeking to conform their behavior to the law’s requirements. Rather, they strove to comply with the widely shared social norm of “neighborliness.” Good neighbors, that norm di-
Law and Economics in the Future of Legal Scholarship, Education, and Practice
227
2008 / 3 Review of Business and Economics
rected, did not sue one another; they helped each other. So, when a farmer found stray cattle on his property, the farmer typically telephoned the rancher, informed him that he had his cattle and would feed and shelter them until the rancher could come to pick them up. If the straying cattle had caused damage, the person who had suffered the damage typically incurred the costs of repair herself and never asked for indemnification from the cattle owner. Ranchers typically responded with thanks and a promise to come to pick up their cattle soon. If a rancher offered to compensate the farmer for damage or the costs of tending his cattle, the farmer frequently refused that offer, writing down those expenses to the costs of being neighborly. There were, nonetheless, law suits between aggrieved farmers and the ranchers whose cattle caused damage. In most instances, these law suits occurred for one of two reasons: first, one of the two parties was new to the area and did not know or follow the prevailing social norms, or second, one of the parties was a notoriously contentious person who was simply difficult to deal with. But these law suits were rare. People were not, as a famous phrase had it, “bargaining in the shadow of the law” – that is, settling their disputes through bargaining in full knowledge of what the law was likely to do if they were to litigate. Rather, farmers and ranchers were both striving to comply with what they took to be the prevailing social norms. They were ignorant of the law, and its prospective liability assignments had no impact on their behavior. The unavoidable implication was that law was not the central governance mechanism for farmers, ranchers, and others in Shasta County, California. Social norms were. The law was not the first resort when potential disputes arose. It was the last resort.22 The response among legal scholars (mostly law-and-economics specialists) to Ellickson’s article was to alter their contention that individuals assiduously sought to conform their behavior to the dictates of the law or to bargain around the law. Rather, it appeared to be the case that Ellickson’s work suggested that social norms, not law, were the central guiding principle of most people’s behavior. And that finding created an entire new body of scholarly inquiry. Now, in looking at many legal issues, we need to ask the extent to which it is social norms, not law, that are governing human behavior. I began this section by pointing out that the first step in the scientific method of inquiry is the framing of a hypothesis or theory about some legal matter. Ellickson’s investigation of the applicability of the Coase Theorem to the affairs of farmers and ranchers in Shasta County, California, began with the articulation of a hypothesis about what he might find in his investigation. And note well that in carrying through his investigation, Ellickson discovered something new – not at all what he expected to find or what he sought to investigate. The scientific method often has this felicitous unexpected result.23
2008 / 3
228
IV.
Th.S. Ulen
Empiricism in Law
The second aspect of the scientific method that law and economics has brought to the study of law (and ultimately, I hope, to the practice of law) is a deep concern for empirical information about the law. As I have said, the central reason for that concern is a concern to see whether the law is having the effects or consequences for which it was designed. Those effects cannot merely be presumed or hoped for on the basis of a coherent theory of what will happen. We vitally need to know what, in fact, they are. There is no better way in which to show the profound importance of empirical work than by considering two illustrations.
Review of Business and Economics
A. Litigation Versus Settlement
Many observers believe that the U.S. is a highly litigious society, that Americans are swift to resort to litigation for resolving the most trivial of disagreements. There are frequent stories in the press that seem to illustrate this over-litigiousness. The most famous of those is the instance of the elderly woman burned by spilling McDonald’s coffee on herself as she drove away from that restaurant’s drive-through window.24 Motivated by this picture, some Americans are frightened to provide the most common services, such as consultation on financial matters, without extensive waivers of liability from those to whom they are providing advice. City governments are said to be afraid to put play structures in public parks or public swimming pools because of fears of liability if children are injured while playing. This frightened and frightening view of litigation (without much systematic empirical support) has had important effects on public policy. There are annual bills in Congress and state legislatures to limit liability generally or for particular professions or products. For example, Congress has limited liability for nuclear-power generators in the Price-Anderson Nuclear Industries Indemnity Act.25 And almost annually there is a proposal introduced in Congress to give regulated industries a “regulatory compliance defense.”26 Excessive litigation and the costs that that imposes on American business have been recurrent themes in state and federal politics in the U.S. over the past 30 years. But is the allegation correct? Is the U.S. in some sense overly litigious? Has litigation exploded, creating economic hardship for some professions, such as physicians, and some industries, such as consumer product manufacturers? Law and economics has contributed a rich theoretical literature to this debate – and this literature should make one skeptical of the “excessive litigation” claim. That theoretical literature seeks to understand why people litigate rather than settle, given that litigation is expensive and unpleasant and that there is often a cooperative surplus between injurer and victim that can be divided relatively easily.27
Law and Economics in the Future of Legal Scholarship, Education, and Practice
229
2008 / 3 Review of Business and Economics
There are three fundamental conclusions of that theoretical literature. First, litigation should be relatively rare: the vast majority of disputes should settle, precisely because the costs of litigation may exceed the stakes of the controversy and because the controversy is unlikely to present novel or contested views of fact or law. Second, the theoretical literature predicts that litigation is unlikely to occur in cases in which it is obvious which party has the stronger claim. Where the law is clear and the facts are more obviously or uncontestedly favorable for one of the parties, there is not much to be gained from litigation. Settlement is cheaper and far more likely. An important implication of this theoretical point is that in the disputes that do result in litigation the law and the facts are contested. That is, litigation is more likely in instances in which each party has close to a 50-50 chance of prevailing at trial. This prediction is referred to as the “selective litigation hypothesis.”28 Third, litigation is likely to occur also when both parties are unrealistically optimistic about their chances of winning. Because most law-and-economics scholars view litigation as wasteful, by comparison to settlement, this theoretical explanation for litigation argues in favor of policies that discourage overoptimism. For instance, professional lawyers can play a significant role in deterring wasteful litigation by counseling their clients out of their unrealistic optimism. Additionally, expansive pre-trial discovery might make the true strength of the other side’s case more evident, making overoptimism more difficult to maintain. Note that this discussion has been entirely theoretical. The theory of when to litigate and when to settle is consistent and coherent. But as I argued in the last section, consistency and coherency are hallmarks of theory but not necessarily of the world in which we live. Or, better put, a different consistency and coherence from that in our theory may characterize the part of the world we are exploring. An absolute necessity is to perform careful empirical work to see the extent to which the world supports our theoretical explanation. That empirical work has been done with increasing vigor in the last 10 years. And the results are sometimes startling. A recent empirical study of litigation by Professor Marc Galanter of the University of Wisconsin Law School argues that litigation is rare and becoming even more rare.29 In that article Galanter studied all filings in all federal courts in the United States between 1962 and 2002 and reported two remarkable findings. First, less than five percent of all disputes in which a claim is filed actually go to trial. This is consistent with the predictions of the law-and-economics theory of litigation versus settlement. And second, the absolute number of trials in all federal and state courts in the U.S. has declined significantly since the early 1960s.30 For example, the average federal judge in the early 1960s presided over 39 trials per year. Today the average federal judge presides over 13 trials per year – two-thirds fewer than was the case one generation ago. Why has there been a decline in the number of trials? Galanter considered some obvious explanations – for example, that litigation has grown more expensive
Review of Business and Economics
2008 / 3
230
Th.S. Ulen
(making settlement far more attractive) and that cheaper methods of alternative dispute resolution (mediation and arbitration) have supplanted litigation. He found that neither of these fully explained the significant decline in the number of trials. I have an alternative theory: there is less litigation because the legal profession is doing a far better job of lawyering than it did before. Recognize the fact that the better job that a lawyer does in his or her counseling for a client (particularly in civil matters), the less likely that there will be a dispute that requires litigation to resolve.31 So, I hypothesize that there may be less litigation because lawyers are more adept at anticipating and heading off disputes than they used to be. Why should lawyering have become so much more effective? Here’s a sketch of a reason. More than 20 years ago, Professor Ron Gilson described commercial lawyers as “transaction cost engineers.”32 By that wonderful phrase I think that he meant that just as a civil engineer anticipates the structural problems that might cause a bridge to sway or collapse and designs the bridge to be immune to those adverse outcomes, so commercial lawyers recognize the costs of concluding agreements, monitoring those agreements, and bringing them to a mutually satisfactory end. As a result, they counsel and write agreements so as to make sure that unanticipated transaction costs cause no swaying or collapsing in those agreements. Note the process of scholarly inquiry that this example illustrates. We began with an observation about the adverse social effects of litigation and then briefly reviewed the law-and-economics theories of litigation. We confronted the theory with data. The data supported the theory but, as in the Ellickson example on social norms, with an unexpected twist. And so, we have to come up with a different theory that supports the empirical findings. This is a new way of studying law.
B. Variations in the Rate of Crime
The second example of the importance of empirical work comes from the study of crime. As I did in the previous illustration, let me begin with an observation about crime in the U.S., give a brief review of the law-and-economics theoretical literature designed to address the issue of crime, and then show how empirical work on the causes of variations in the incidence of crime have thrown fascinating light on the matter. Crime has gone in long cycles in the U.S. For example, in the 1920s and 1930s crime increased sharply; during the 1940s and 1950s crime declined; it rose again beginning in the late 1960s and through the 1970s and 1980s; and then in the 1990s crime began to decline again. What explains these cycles? Are they related to demographic changes, to political changes, to alterations in the vigor with which society deters crime, or to something else? One insight into why this may have happened was the Nobel Laureate Gary Becker’s new theory of the decision to commit a crime.33 Becker hypothesized that
Law and Economics in the Future of Legal Scholarship, Education, and Practice
231
2008 / 3 Review of Business and Economics
potential criminals make decisions in the same rational manner that most people make decisions about other actions. Specifically, a potential criminal calculates the benefits of the crime (both monetary and nonmonetary) and the expected costs of the crime (the probability of detection, arrest, and conviction times the sanctions – again, both monetary and nonmonetary – imposed if convicted). If the expected costs exceed the benefits, then the rational potential criminal does not commit the crime. If the benefits exceed the expected costs, then he or she commits the crime. Becker also related these individual optimizing decisions about whether to commit a crime to the social concerns about crime. Presumably, society’s concern is to minimize the social costs of crime. Becker hypothesized that crime could be deterred by manipulation of the expected costs of crime that individuals perceive and that this could be done in several different ways. First, the authorities could decrease the amount of crime by increasing the expected cost of crime through policies designed to increase the probability of detection or arrest or conviction. For example, greater use of closed-circuit TV might heighten the probability of detection; an increase in the number of police might increase the probability of arrest; and an increase in the number of prosecutors might increase the probability of conviction. Becker noted that any of these policies would cost real resources. Second, the authorities might deter crime by increasing the expected cost of crime through raising the sanction imposed on those guilty of a crime. For this policy to work without unintended consequences, the sanctions on all crimes ought to be increased so that there is no deflection away from one set or class of crimes to another. Becker pointed out that there would be no particular cost to deterring crime in this fashion.34 Third, the authorities could seek to deter crime by raising its opportunity cost. This would be the case if legal and gainful employment were to become more likely and more rewarding. Under those circumstances, the expected cost of crime would indirectly increase in that time spent in prison would be more costly. To the extent that potential criminals consider these opportunity costs, a full-employment policy and one designed to guarantee competitive compensation are effective anticrime policies. Finally, the authorities could use a combination of the preceding three policies. Note that there are testable implications from this model regarding variations in the rate of crime within a nation and differences in the amount of crime across national borders. So, for example, one such testable implication is to suggest that if crime has increased, then perhaps the expected cost of crime has fallen. That might happen if there are changes in the probabilities of detection, arrest, and conviction or if legal employment opportunities have become less available or less rewarding. We might apply these theoretical insights to the recent pattern of crime in the U.S. The rate of both non-violent and violent crime in the United States began to decline in the early 1990s. In the course of that decade, property crimes fell 30 per-
Review of Business and Economics
2008 / 3
232
Th.S. Ulen
cent, and violent crimes declined by 40 percent. Those declines have continued, though at a decreasing rate, through 2007, the latest year for which we have figures. Why did these changes occur? There have been plenty of theories – the economy has been robust; we have increased the number of prisoners by four-fold since 1980 (to over 2 million), and so on.35 In 2001 John Donohue, then of Stanford Law School, now at Yale Law School, and Steve Levitt of the Department of Economics at the University of Chicago published a startling alternative explanation for the decline in crime – the legalization of abortion in the early 1970s.36 What’s the connection between the legalization of abortion and crime levels? Donohue and Levitt find two connections. First, legalized abortion may reduce the number of 18-24 year old males, a group that accounts for almost 50 percent of crime in every society in the world. And that is, in fact, what Donohue and Levitt showed. Beginning in 1991, there was a much smaller cohort of 18-year-old males, and that decline, they calculated, accounted for about 25 percent of the decline in crime since the early 1990s. Second, they hypothesized that the quality of the children who were born after Roe v. Wade in 1973 was higher than was the case before 1973. The reason is that “women who have abortions are those most at risk to give birth to children who will engage in criminal activity – teenagers, unmarried women, and the economically disadvantaged.”37 The legalization of abortion allowed those women to choose a more propitious time at which to have their children. As a result, after January, 1972, more children grew up in better circumstances, making them less likely to commit crime. They calculated that this cohort quality effect accounted for another 25 percent of the decline in crime in the 1990s. I recognize that these are highly controversial results, and I am not at all recommending the legalization of abortion as a crime-control strategy. But can anyone deny that this is a startlingly original insight of great importance? I commend this Donohue and Levitt article as a spectacular example of the bright light that careful empirical work can throw on important legal issues.
V.
Conclusion
My crystal ball is small and clouded; so, I cannot be sure what the future of legal scholarship, education, and practice holds. But I have to say this: I think that the legal profession is one of the most important forces for good in civilized society. I do not and will never join the chorus that seeks to criticize lawyers generally. And I also believe that even though the relationship between the legal academy and the practicing bench and bar are today not congenial, at least in the U.S., that situation
Law and Economics in the Future of Legal Scholarship, Education, and Practice
233
1. Swanlund Chair, University of Illinois at Urbana-Champaign; Professor of Law, University of Illinois College of Law; and Director of the Illinois Program in Law and Economics. This is a revised version of the talk that I gave on Friday, October 5, 2007, at the University Hall on the occasion of being awarded a Doctorate honoris causa by the Katholieke Universiteit Leuven. I want to thank the Rector of KUL, the Faculties of Law and of Economics of KUL, and my dear friends Professor Dirk Heremans, Professor Herman Cousy, and Professor Stef Proost and their wives, Rea, Christiane, and Marjan, for their many kindnesses to me over the years. There are several other people whom I must thank for their help in my getting to Leuven on October 5, 2007. First and foremost, I want to thank my Mother and Father. They are both gone now, but in the hope that somehow they can know that this marvelous ceremony took place and that I didn’t turn out as badly as many would have confidently predicted, I want to acknowledge them. We carry not only the genetic material of our parents; we are, more importantly, the product of their love and affection. The love and nurturing that my Mother and Father lavished on my sisters and me was extraordinary. Another thank you that I want to extend is to faculty assistant at Illinois, Sally Cook. She is not here, but everyone in this business knows how important our support staff is. Sally has been not only the best assistant that I’ve ever had; she’s the best imaginable. The greatest thank you that I have to give is to my family – to my dear wife, Julia, and to our wonderful sons, Ted and Tim. Let me speak first of our sons. Both of our boys are training to be lawyers. My father was a lawyer; my great uncle, who was a great inspira-
Review of Business and Economics
NOTES
2008 / 3
will not last. I am confident that the new insights into law that come from law and economics are not only inherently interesting but are also of great practical value and will, therefore, find their ways into the practice of law, the art of judging, and the work of legislators and legal administrators. My central point has been that the greatest change that law and economics has brought to the study of law is the scientific method of inquiry. That method requires articulating a theory or hypothesis that is coherent and consistent and novel. And then the method requires confronting the theory with carefully designed empirical work. I want to end on an optimistic note. I have already mentioned the deep reverence that I have for reason and for the commitment to the scholarly life that characterizes the great research universities. But what about the future? What about the many difficult questions of human governance with which we must grapple? Can we deal with those? Are we equipped to understand the legal and other challenges before us? In the early 1930s, near the end of his life, the great German mathematician David Hilbert addressed an audience in his native Konigsberg about the optimism that every scholar must bring to even the hardest problems. He concluded with six famous words that are carved upon his tombstone and that I love and that I commend to you as words by which every scholar should live: Wir müssen wissen. Wir werden wissen. We must know. We shall know.
2008 / 3
234
Th.S. Ulen
tion to me, was also a lawyer. It’s clear that by not becoming a lawyer, I went dreadfully wrong. But that error will only last one generation. Julia has invariably supported and encouraged me over the past 34 years. She really is a co-recipient of this honor. 2. Tad went on to a distinguished career as an attorney for the Coca-Cola Company and then in private practice in Washington, DC. 3. I cannot resist recounting some aspects of the first several classes that I taught at the University of Illinois College of Law. There was some concern on the part of the law school administration that because I did not have a law degree, I might not teach the law students the material in an appropriate manner or with content that was sufficiently legal. (Since the founding of the University of Illinois College of Law in 1897 there had only been one non-lawyer who had taught law students – the distinguished sociologist Rita Simon – and then only in a research seminar with limited enrolment.) So, a young assistant professor of law was assigned to co-teach with me. He and I became great friends and taught very effectively together. I do not tell this story to illustrate anything other than the great distances that law-and-economics and legal education have traveled since the 1980s. Today nearly half of the full-time faculty at the University of Illinois has a graduate degree in some subject other than law in addition to a law degree. 4. For an account of the rise of law and economics and hypotheses about the reasons for its success in North America and its slow reception elsewhere, see Nuno Garoupa & Thomas S. Ulen, The Market for Legal Innovation: Law and Economics in Europe and the United States, ALA. L. REV. (forthcoming, 2008).
Review of Business and Economics
5. One sign of that spreading interest is the fact that a text on law and economics that Professor Robert D. Cooter and I have written – Law and Economics (5th ed. 2007) – has been translated into Polish, Russian, Chinese, Japanese, Korean, Spanish, Italian, and Slovenian. In addition, there are regional and national associations of law-and-economics scholars around the globe. 6. I should make clear that by the word “law” I mean all the governance structures, public and private, that society uses to put itself in order and to secure the good things in life for all its citizens. Law is, I truly believe, one of civilization’s greatest inventions – one that has long guided us toward a better life and holds out the promise of an even better life. 7. See Muller v. Oregon, 208 U.S. 412 (1908). That case is famous for, among other things, the “Brandeis brief.” That was an amicus curiae brief submitted by Louis Brandeis, then in private practice in Boston and later an Associate Justice of the United State Supreme Court, in which Brandeis focused on the “real” aspects of hours-limitations laws by, for example, enumerating the other states and nations that had enacted such limitations and citing literature on the social effects of long work hours. 8. Naturally I exempt jurisprudence from this characterization. The philosophy of law has long had a concern with generalities in the law and not with specific jurisdictional concerns. 9. I develop this notion in A Nobel Prize in Legal Science: Theory, Empirical Work, and the Scientific Method in the Study of Law, 2002 U. ILL. L. REV. 875. 10. The phrases “separating equilibrium” and “mixed equilibrium” come from the game theoretical analysis of situations of asymmetric information. “Asymmetric information” describes a situation in which two sides of a potential transaction have different information about some central aspect of the impending transaction but find it difficult to communicate that information truthfully. Imagine, for example, that some workers are very energetic and others are not but that it is difficult for workers to signal credibly to potential employers that they belong to the energetic group and difficult for employers to distinguish the two groups. (All potential employees would try to signal that they are energetic workers.) If a credible characteristic or signal for distinguishing energetic from nonenergetic workers could be found such that it could not be faked by nonenergetic
Law and Economics in the Future of Legal Scholarship, Education, and Practice
12.
13. 14.
15.
16.
18. 19.
20.
21.
22. 23.
24.
Review of Business and Economics
17.
workers, then the resulting employment situation will be a “separating equilibrium,” in which energetic workers get jobs and nonenergetic workers do not. However, if no clear distinguishing characteristic or signal between the two types of workers can be found, the resulting employment situation will be a “mixed equilibrium,” in which employers hire some energetic workers and some nonenergetic workers (and some energetic and some nonenergetic workers are unemployed). An old joke says that a low-paid job as a professor provides “all the prestige one can eat.” See Richard H. McAdams, The Origin, Development, and Regulation of Norms, 96 MICH, L. REV. 338 (1997), and GEOFFREY BRENNAN & PHILIP PETTIT, THE ECONOMY OF ESTEEM: AN ESSAY ON CIVIL AND POLITICAL SOCIETY (2004). Of course, if the average earnings for practicing lawyers fall, then the demand for legal education may also fall, reducing the demand for law professors’ services. Judge Harry T. Edwards of the U.S. Court of Appeals for the District of Columbia Circuit, a former law professor at the University of Michigan, wrote of this drift in The Growing Disjunction Between Legal Education and the Legal Profession, 91 MICH. L. REV. 34 (1992). A subsequent issue of the same volume of the Michigan Law Review contained numerous commentaries on Judge Edwards’ article. See particularly George L. Priest, The Growth of Interdisciplinary Research and the Industrial Structure of Legal Ideas: A Reply to Judge Edwards, 91 MICH. L. REV. 1929 (1993). The “disjunction” has been the cause of continuing commentary since the article appeared. See also ANTHONY KRONMAN, THE LOST LAWYER: FAILING IDEALS OF THE LEGAL PROFESSION (1993). Indeed, the fact that I, an American, had the great honor of recognition from the Katholieke Universiteit Leuven testifies to the fact that law and economics, in its study of the law, transcends the particularities of legal topic and of place that so occupy traditional legal scholarship. See generally YUICHI SHIONOYA, THE SOUL OF THE GERMAN HISTORICAL SCHOOL: METHODOLOGICAL ESSAYS ON SCHMOLLER, WEBER, AND SCHUMPETER (2005). See generally VILFREDO PARETO, THE MIND AND SOCIETY: A TREATISE ON GENERAL SOCIOLOGY (Arthur Livingston ed., 1935). Id. at xviii. See, for example, SAMIR OKASHA, THE PHILOSOPHY OF SCIENCE: A VERY SHORT INTRODUCTION (2002) and PETER GODFREY-SMITH, THEORY AND REALITY: AN INTRODUCTION TO THE PHILOSOPHY OF SCIENCE (2003). Ronald A. Coase, The Problem of Social Cost, 3 J.L. & ECON. 1 (1960). For a modern treatment of the theorem, see ROBERT D. COOTER & THOMAS S. ULEN, LAW AND ECONOMICS 85-99 (5th ed. 2007). See Ellickson, Of Coase and Cattle: Dispute Resolution Among Neighbors in Shasta County, 38 STAN. L. REV. 623 (1986). Ellickson elaborated on these findings and gave many other examples and an overarching theory of these matters in ORDER WITHOUT LAW: HOW NEIGHBORS SETTLE DISPUTES (1991). Note, too, that Ellickson’s original plan to investigate the predictions of the Coase Theorem in a natural experiment had given way to other concerns. The history of science is full of examples of unexpected discoveries that have resulted from the careful attempt to investigate a natural phenomenon. One of the most famous is the unexpected discovery that the Earth orbits the Sun and not the other way around. See THOMAS S. KUHN, THE COPERNICAN REVOLUTION: PLANETARY ASTRONOMY IN THE DEVELOPMENT OF WESTERN THOUGHT (1965). Liebeck v. McDonald’s Rest., P.T.S., Inc., No. CV-93-02419, 1995-WL 360309 (N.M. Dist. Ct. Aug. 18, 1994). Ms. Stella Liebeck purchased hot coffee, in a take-away cup from a McDonald’s restaurant in Albuquerque, New Mexico. Her grandson, who was driving,
2008 / 3
11.
235
2008 / 3
236
Th.S. Ulen
pulled over and stopped the car to allow Ms. Liebeck to put condiments in the coffee. During that attempt to fix her coffee. Ms. Liebeck spilled coffee onto herself, burning her thighs, groin area, and buttocks painfully. Ms. Liebeck went to a hospital suffering from what were diagnosed to be third-degree burns over six percent of her body. She was in the hospital for eight days, received skin grafts as treatment, and went through two years of additional care. After she was released from the hospital, her daughter, a nurse, took time off from work to minister to her mother at home. Ms. Liebeck approached McDonald’s for compensation for her medical expenses (approximately $10,000) and for her daughter’s lost income from her job. She sought a total of about $20,000 from McDonald’s. McDonald’s offered to pay $800. Ms. Liebeck brought an action against the restaurant and was successful. The jury, on a theory of comparative negligence, determined that McDonald’s was 80 percent responsible for Ms. Liebeck’s injuries and that Ms. Liebeck was 20 percent responsible for her injuries. The jury awarded her $2.86 million in compensatory and punitive damages ($200,000 in compensatory damages and almost $2.7 in punitive damages; the jury is said to have computed the punitive damages by determining one or two days’ worth of McDonald’s national coffee revenues). McDonald’s compensatory damages were 80 percent of $200,000 or $160,000. The trial judge reduced the punitive damages to $480,000, three times the compensatory damages. So, the total amount of damages was $640,000. Both parties appealed but then settled the dispute for an undisclosed amount. 25. The Act, first passed and renewed periodically since – most recently in the Energy Policy Act of 2005 – limits the liability for all non-military nuclear facilities constructed in the U.S. before 2026 to $10 billion.
Review of Business and Economics
26. Under that defense, a firm subject to federal regulation, such as a pharmaceutical manufacturer, could assert and demonstrate, as a complete defense, that their product, which had allegedly caused consumer injury, complied with all relevant regulations. For an economic analysis, see Alan Schwartz, Statutory Interpretation, Capture, and Tort Law: The Regulatory Compliance Defense, 2 AM. L. & ECON. REV. 1 (2000). 27. There is also an interesting theoretical literature on frivolous lawsuits. The puzzle to be explained is why actions that have “negative expected value” are pursued and sometimes succeed. See also Robert Bone, Modeling Frivolous Suits, 145 U. PA. L. REV. 519 (1997). 28. For the original statement of the hypothesis, see George Priest & Benjamin Klein, The Selection of Disputes for Litigation, 13 J. LEGAL STUD. 1 (1984). See also Cooter & Ulen, supra n. 5, at 474 – 84, for a summary discussion of the empirical literature on the predictions of the selective litigation hypothesis. 29. The Vanishing Trial: An Examination of Trials and Related Matters in Federal and State Courts, 1 J. EMP. LEGAL STUD. 459 (2004). 30. The decline was not slow and steady over the entire forty-year period. Rather, the absolute number of trials rose slowly from the early 1960s through the mid-1980s and then declined precipitously since the mid-1980s. 31. One of the few articles to stress this point is Robert E. Scott & George G. Triantis, Anticipating Litigation in Contract Design, 115 YALE L.J. 814 (2006). 32. Ronald J. Gilson, Value Creation by Business Lawyers: Legal Skills and Asset Pricing, 94 YALE L.J. 239, 255 (1984). 33. Gary S. Becker, Crime and Punishment: An Economic Analysis, 76 J. POL. ECON. 169 (1969). 34. Some point out that if the sanctions take the form of longer prison sentences, then more prisons might have to be built. That, obviously, would cost real resources. But if the greater sanctions are deterring, there will be less crime and, therefore, less demand for incarceration.
Law and Economics in the Future of Legal Scholarship, Education, and Practice
237
2008 / 3
35. See Steven D. Levitt, Understanding Why Crime Fell in the 1990s: Four Factors That Explain the Decline and Six That Do Not, 18 J. ECON. PERSP. 163 (2004), and Steven D. Levitt & Thomas J. Miles, “Empirical Study of Criminal Punishment,” in A. MITCHELL POLINSKY & STEVEN SHAVELL, EDS., HANDBOOK OF LAW AND ECONOMICS, v. 1 (2007). 36. John J. Donohue & Steven D. Levitt, The Impact of Legalized Abortion on Crime, 116 Q. J. ECON. 379 (2001). Although the article appeared in 2001, manuscript copies of the work were available in 1998, and there was a front-page New York Times discussion of the article in 1999: Erica Goode, “Linking Drop in Crime to Rise in Abortion,” New York Times, August 20, 1999. 37. Donohue & Levitt, supra n. 35, at 381.
REFERENCES
Review of Business and Economics
Becker, G.S., 1969, Crime and Punishment: An Economic Analysis, The Journal of Political Economy 76, 169-217. Bone, R., 1997, Modeling Frivolous Suits, University of Pennsylvania Law Review 145, 519605. Brennan, G. and Pettit, P., 2004, The Economy of Esteem: An Essay on Civil and Political Society, Oxford University Press, Oxford. Coase, R.A., 1960, The Problem of Social Cost, Journal of Law and Economics 3, 1-44. Cooter, R.D. and Ulen, T.S., 2007, Law and Economics 5th ed., Addison-Wesley Educational Publishers, Boston. Donohue, J.J. and Levitt, S.D., 2001, The Impact of Legalized Abortion on Crime, The Quarterly Journal of Economics 116, 379-420. Edwards, H.T., 1992, The Growing Disjunction Between Legal Education and the Legal Profession, Michigan Law Review 91, 34-78. Ellickson, R.C., 1986, Of Coase and Cattle: Dispute Resolution Among Neighbors in Shasta County, Stanford Law Review 38, 623-687. Ellickson, R.C., 1991, Order Without Law: How Neighbors Settle Disputes, Harvard University Press, Cambridge. Galanter, M., 2004, The Vanishing Trial: An Examination of Trials and Related Matters in Federal and State Courts, Journal of Empirical Legal Studies 1, 459-570. Garoupa, N. and Ulen, T.S., 2008, The Market for Legal Innovation: Law and Economics in Europe and the United States, Alabama Law Review, (forthcoming). Gilson, R.J., Value Creation by Business Lawyers: Legal Skills and Asset Pricing, The Yale Law Review 94, 239-313. Godfrey-Smith, P., 2003, Theory and Reality: An Introduction to the Philosophy of Science, University of Chicago Press, Chicago. Kronman, A., 1993, The Lost Lawyer: Failing Ideals of the Legal Profession, Harvard University Press, Cambridge. Kuhn, T.S., 1965, The Copernican Revolution: Planetary Astronomy in the Development of Western Thought, Harvard University Press, Cambridge. Levitt, S.D., 2004, Understanding Why Crime Fell in the 1990s: Four Factors That Explain the Decline and Six That Do Not, The Journal of Economic Perspectives 18, 163-190. Levitt, S.D. and Miles, T.J., 2007, Empirical Study of Criminal Punishment. In: Mitchell Polinsky, A. and Shavell, S. (eds.), Handbook of Law and Economics, vol. 1, North-Holland, Amsterdam, 343-390.
Review of Business and Economics
2008 / 3
238
Th.S. Ulen
McAdams, R.H., 1997, The Origin, Development, and Regulation of Norms, Michigan Law Review 96, 338-433. Okasha, S., 2002, The Philosophy of Science: A Very Short Introduction, Oxford University Press, Oxford. Pareto, V., 1935, The Mind and Society: A Treatise on General Sociology (Arthur Livingston ed.), Harcourt, Brace and Company, New York. Priest, G.L., The Growth of Interdisciplinary Research and the Industrial Structure of Legal Ideas: A Reply to Judge Edwards, 1993, Michigan Law Review 91, 1929-1944. Schwartz, A., 2000, Statutory Interpretation, Capture, and Tort Law: The Regulatory Compliance Defense, American Law and Economics Review 2, 1-57. Shionoya, Y., 2005, The Soul of the German Historical School: Methodological Essays on Schmoller, Weber, and Schumpeter, Springer, New York. Ulen, T.S., 2002, A Nobel Prize in Legal Science: Theory, Empirical Work, and the Scientific Method in the Study of Law, University of Illinois Law Review, 875-920. Priest, G. and Klein, B., 1984, The Selection of Disputes for Litigation, The Journal of Legal Studies 13, 1-55. Scott, R.E. and Triantis, G.G., 2006, Anticipating Litigation in Contract Design, The Yale Law Review 115, 814-879.
R. Veugelers
1
2008 / 3
The Role of SMEs in Innovation in the EU: A Case for Policy Intervention?
ABSTRACT
JEL CODES
L25, L26, 031, 038
KEYWORDS
SMEs, Innovation, Policy, YICs
RESEARCH FUNDING
I.
Financial support from FWO (ESF-STRIKE & G.0523.08), Belgian Federal Government PAI (P6/09), and the KUL-OF (OT/07/11) are gratefully acknowledged
Introduction
For innovation, the case of market failure is often taken as obvious and unnecessary to demonstrate. Estimates of social returns to innovation go up to twice the
Review of Business and Economics
Innovation policies, addressing the barriers that inhibit Small- and Medium-Sized Enterprises (SMEs) to innovate, have the potential to be extremely effective. Nevertheless, as this article will illustrate, market failure for innovating SMEs needs to be carefully identified, to better target government policies. And even if policy is well targeted, there could still be failures in implementation, as the evidence from actual policies being implemented in the EU and elsewhere suggest. In order to support policy making for stimulating innovation at SMEs, this article first reviews the empirical evidence on the contribution of SMEs to innovation, examining whether the barriers to innovation are related to market failures and whether SMEs are affected more by market failure induced barriers. We also take a look at trends in actual SME innovation policies in Member States and at EU level and how they have been evaluated on effectiveness. A concluding section channels the insights from this evidence into suggestions for improving policies towards SMEs and Innovation. The potential for specific SME Innovation policies, towards Young Innovative Companies (YICs) with radical innovations is identified.
2008 / 3
240
size of private returns (Griliches (1992)). Nevertheless, as this article will illustrate, market failure for innovating SMEs needs to be carefully identified, to better target government policies. And even if policy is well targeted, there could still be failures in implementation. In order to improve policy making for stimulating innovation at SMEs, this article first reviews the empirical evidence on the contribution of SMEs to innovation (section II), identifying the barriers for SMEs to innovate. Sections III & IV examine whether the barriers to innovation are related to market failures and whether SMEs are affected more by market failure induced barriers. The case for specific SME Innovation policies, especially towards Young Innovative Companies (YICs) with radical innovations is made. In an attempt to look at assessing the risk for government failure, we next take a look at trends in actual SME innovation policies in Member States and at EU level and how they have been evaluated on effectiveness (sections V & VI). A concluding section VII channels the insights from the previous sections into suggestions for improving policies towards SMEs and Innovation.
II.
Review of Business and Economics
R. Veugelers
Evidence on the contribution of SMEs to innovation
SMEs clearly have a role to play in the innovative system, if only because they are numerous, representing the bulk of economic activity in most economies. But they are even more promising actors in the Schumpeterian dynamics: in many fields, high-technology SMEs have a key role in developing and exploiting new technologies which will play a major role in the generation of new sectors and stir the future innovations of the economy (e.g. Baumol (2002)). Large firms are increasingly looking for external sourcing of new technologies and high-technology SMEs are one of the external sources which they aim to access. Countries without such innovative SMEs will be less attractive to large innovative, internationally mobile businesses. But are SMEs playing this role in practice in Europe? This section takes a closer look at the evidence on the contribution of SMEs to innovation in Europe.2
A. Are EU SMEs more or less innovative than large firms?
Looking at innovation inputs, about 75% of R&D expenditures in the EU are accounted for by large firms (> 500 employees). This still leaves a quarter of business R&D performed by SMEs (< 250 employees),3 a figure substantially higher than in the US (14%) and Japan (7%). Figure 1 shows the share of SMEs in total R&D employment, by EU-25 country.
The Role of SMEs in Innovation in the EU: A Case for Policy Intervention?
241
2008 / 3
Figure 1. Share of SMEs in Business Enterprise Researchers. Source:
Eurostat, R&D statistics, OECD-MSTI 2006-1.
Figure 2. Share of enterprises with innovation; By size class. Source:
Eurostat, Innovation in Europe, CIS3 results, 2004.
Review of Business and Economics
Since the share of SMEs is smaller in R&D than in sales or employment, this suggests that EU SMEs are less R&D intensive than large EU firms. Also, EU SMEs are less R&D intensive than US SMEs, but the SME gap is similar to the overall gap in R&D spending.4 Figure 2 shows that a lower percentage of EU SMEs successfully innovate (i.e. introduce new products or processes) as compared to large firms, both in services and in manufacturing.
2008 / 3
242
R. Veugelers
Although Figure 3 shows that this holds across all Member States,5 it also shows a high correlation between the intensity of innovation in SMEs and in large firms,6 suggesting that the importance of a country’s innovative potential is independent of the size of firms, and that there is a complementarity between innovative large and small firms.
Figure 3. Innovative firms by size class.
Review of Business and Economics
Source:
DGENTR, 2007, SMEs and Innovation: Note for the EPC (On the basis of CIS4).
Figure 4. Intensity of innovation per size class. Source:
DGENTR, 2007, SMEs and Innovation: Note for the EPC (On the basis of CIS4).
When looking at innovation intensity (i.e. share of turnover from innovative products), the differences between large and small firms are less strong, but still in fa-
The Role of SMEs in Innovation in the EU: A Case for Policy Intervention?
243
2008 / 3
vour of large firms in most countries (see Figure 4). With a lower SME innovative intensity, the share of SMEs in total innovative turnover is much smaller than their share in total economic activity for most EU countries. Only in a few (smaller) member states, like Portugal, Cyprus and Estonia are SMEs more important.
B. Are SMEs naturally disadvantaged in innovation?
Overall these results seem to indicate that SMEs and innovation are negatively correlated. This is consistent with the literature dating back to Schumpeter (1942), who argued that the large firm operating in a concentrated market is the main engine of technological progress (see also Kamien and Schwartz (1982), Cohen & Levin (1989), Symeonidis (2006)). Arguments in favour of large firms innovating are that R&D projects typically involve large fixed costs and these can only be covered if sales are sufficiently large; there are scale and scope economies in the production of innovations; large firms can undertake many projects at a time and hence spread the risks of R&D; large firms have better access to external finance and firms with greater market power are better able to finance R&D from own profits and more easily appropriate the returns from innovations. Counterarguments for large firms innovating are decreasing returns to scale in the production of innovation due to loss of managerial control and the bureaucratisation of innovation activity. And incumbency may lead to inertia and lack of incentives to innovate to avoid cannibalization of existing profits.
The issue therefore is not whether firm size is conducive to innovation, but rather which market and technology characteristics favour large and/or small firms. Characteristics like market concentration, technological opportunities, the stage of the technology life cycle, scope for diversification all matter here (e.g. Rothwell & Dodgson (1991), Utterback (1994), Cohen & Klepper (1996)). But perhaps the clearest case of disadvantage for small firms is identified for access to finance. With R&D typically invoking large and uncertain sunk costs, availability of internal and external finance is a critical issue for innovating firms (e.g. Czarnitzki (2006)).
Review of Business and Economics
However, much of the multivariate empirical analysis on the relationship between firm size and innovation, incorporating a wide set of firm and industry characteristics as control, has failed to find significant results for a positive (or negative) effect of firm size (see e.g. Acs & Audretsch (1987), Cohen & Levin (1989), Simeonidis (2006) for an overview of the empirical literature). They thereby seem to refute Schumpeter’s view, at least on average. This literature shows the importance of sectoral differences, with large firms important in specific sectors (like chemicals) and smaller firms in other sectors (like instruments), while in still other sectors there is a symbiosis between large and small firms (like biotech).
2008 / 3
244
R. Veugelers
And with imperfect capital markets, firm size (or market power) may matter. Capital market imperfections are due to asymmetric information problems, which are exacerbated the more risky the investments are. To mediate this problem, collateral and firm reputation helps. Small and young firms with less collateral and less reputation will face more financial barriers. Although larger firms may be financially constrained as well, this is still to a lesser degree than smaller firms.
Excessive perceived economic risks
Innovation costs too high
Lack of appropriate sources of finance
Organisational rigidities within the enterprise
Lack of qualified personnel
Lack of information on techonology
Lack of information markets
Insufficient flexibility of regulations or standards
Lack of customer responsiveness to new goods or services
Review of Business and Economics
Survey data for the EU confirm the importance of access to finance and too high economic risks as the major hampering factors for innovation, for all types of firms, but somewhat more for small than for large firms (see Figure 5).
Small
16
21
16
6
13
5
5
10
8
Mediumsized
13
19
13
5
14
4
4
8
6
Large
18
21
10
6
13
3
4
7
5
Figure 5. Proportion of enterprises that regard selected hampering factors as highly important (% of all enterprises); By size class. Source:
Eurostat, Innovation in Europe, CIS3 results, 2004.
A second tier of barriers is formed by access to skills, which again impedes both small as well as large firms. Also regulatory burden shows up in the second tier of barriers. Here, the data are suggestive of this burden being somewhat stronger for small firms. This could be a reflection of the regulation being designed such that they affect relatively more new and small firms, unintended or intended e.g. because they are strategically influenced through lobbying by large firms as raising rival’s costs and establishing barriers to entry. Or because the cost of compliance involve investments where smaller firms, because of the capital market imperfections, may face a higher cost of financing.
The Role of SMEs in Innovation in the EU: A Case for Policy Intervention?
245
(YICs) in radical innovations Up till now, we have considered the group of SMEs as a homogeneous block, identified by their small size, as being different from large firms in the costs and benefits from innovation they face. But clearly, SMEs differ according to other dimensions beyond size, that may matter more for shaping their innovative behaviour.
In the context of access to finance, perhaps more important than firm size, is the radicalness of the innovative project (of the Leading SMEs) and the effect of firm age, a fortiori if they are combined in Young Innovative Companies (YICs). Multivariate analysis on German CIS data which includes information on firm age, confirms that small firms and R&D intensive firms are more constrained by lack of internal and external funding, but that on top of this, YICs are even more financially constrained (Schneider & Veugelers (2008)). If YICs are more hampered in their innovative effort, because of financial constraints (or other), this is particularly problematic as these firms may play a very specific role in innovation, even if they are not large in numbers and size. Also Baumol (2002) notes the importance of small firms in large innovations,9 but at the same time recognizes the complementarity between small and large firms. According to Baumol, private sector innovations in the US come from two distinct sources,
Review of Business and Economics
It is important to recognize the heterogeneity within the SME population on the innovation dimension. The vast majority of SMEs undertake no or little R&D (Basic SMEs). But even within the group of SMEs innovating, there are differences in innovation profiles. Most of the SMEs have a low innovative profile, but regularly acquire, adapt, apply technology new to the firm (Adopting SMEs). Another part of SMEs however, is more innovative as technology user, developing or combining existing technology to develop innovations that are not only new to the firm, but also new to the market (Leading SMEs). Of these leading SMEs, only a very small number is involved in leading-edge high level research, leading to drastic innovations. These pioneering SMEs are often young start-up companies. Small, young firms, not having to be concerned with safeguarding incumbent profits, or restructure existing technology profiles of the company, can be more inclined to introduce radical innovations (Young Innovative Small Companies (YICs)). Table 1 provides an indication for Germany on the frequency of each of these types on the basis of CIS survey evidence. Although the data have a bias in favour of innovative firms, they nevertheless show the very small number of YICs for Germany.7 But these small number of YICs stand out in the innovative landscape, because of their critical contributions to radical innovations (see also Storey and Tether (1998)). In the German CIS sample, the share of sales from radically new innovations in total sales for YICs is on average 23%, 3 times higher than the total average.8
2008 / 3
C. Disaggregating SMEs: the role of small Young, Innovative Companies
2008 / 3
246
R. Veugelers
firstly from the activities of large firms and secondly from the efforts of independent inventors and their entrepreneurial partners. Baumol asserts that the active presence and interaction between both groups enhances the overall innovation process since their activities are complementary, with the independent inventors/entrepreneurs specialising in breakthrough innovations and with the R&D departments of the larger firms enhancing these breakthroughs and adding to their overall usefulness. He notes how fortunate the US has been to have this symbiosis of large and small firms.
Table 1. The heterogeneity within the SME population along the innovation dimension: some evidence from Germany. Basic SMEs
46% had not introduced product and process innovations in the last 3 years 45% had no (significant) own R&D (i.e. R&D to sales ratio of at least 1%)
Adapting SMEs
20% of SMEs that introduced new products and processes had no (significant) own R&D
Leading SMEs
45% of SMEs with (significant) own R&D introduced products new to the market
Young, Innovative Companies (YICs)
34 companies were < = 5 years old and highly innovative (R&D-to-sales ratio > 15%)
Review of Business and Economics
19 of these companies introduced products that were new to the market Share of sales from new-to-market innovations of YICs are three times higher than the sample average Note: (i) All SMEs (< 250) in the German CIS-IV 2005 survey response = 2868*; (ii) The CIS sample is not representative of the overall population of enterprises. The survey excludes very small firms and certain service sectors, like personal services, leaving a bias in favour of innovation. On top of the selection bias, a non-response bias in favour of innovation has to be factored in. Source:
Schneider & Veugelers (2008) KUL-WP on basis of Eurostat, CIS-IV Results from Germany.
There is, as yet, little analysis on how the EU is doing with respect to the complementarity between small and large firms. Collaborative agreements, joint networks, joint geographic location (in local clusters) are all mechanisms which foster the interaction between large and small firms. CIS survey data reveal the low propensity for innovative SMEs in the EU to be engaged in R&D cooperation with others (see figure 6). While on average about 14% of innovative SMEs are cooperating with others, this is much higher in Denmark, Sweden and Finland, with the latter two also characterized by a dominant share of large firms in R&D. These countries are also the three countries that are leading in overall innovative performance in Europe (EC-EIS Scoreboard (2008)), suggestive of the importance of a well connected innovation ‘system’.
The Role of SMEs in Innovation in the EU: A Case for Policy Intervention?
247
2008 / 3 Figure 6. Innovative SMEs co-operating with others (as % of all SMEs); By country. Source:
III.
Eurostat, Statistics in Focus, 116-2007, CIS 4 results.
A need for policy intervention: market failures?
In the case of private sector investment in innovation, the case of market failure is well argued. Several market failures exist, explaining why firms with new ideas and products may fail to enter or why incumbent firms will not invest enough in innovations (see e.g. Hall (2005)): (i) A market failure in providing financing. First, the degree of risk of an R&D project from an economy-wide point of view may be lower than that perceived by private firms; or, closely related, the risk premium demanded by private investors may be higher than “warranted” because of asymmetric information. Second, the degree of risk aversion by private investors may be higher than the social rate. As a result, the market may provide for too little risk taking in R&D. (ii) A lack of appropriation of the returns from innovation, because of spillovers. The incentive for firms to engage in risky and costly R&D investment is the expected returns from innovation. But with knowledge spilling over to other users, the innovator typically may not appropriate all the benefits from innovation. This divergence between the social and the private rate of return results in a too low private investment as compared to the socially desirable level.10
Review of Business and Economics
The empirical evidence presented supra suggests a large potential role of SMEs, particularly YICs, for innovation. However, significant barriers exist in Europe that inhibit SMEs and particularly YICs to innovate. The typical next step in the argumentation is a need for policies to address these barriers. But a first question to ask is whether this underperformance is caused by market failure. And secondly, even in case of market failure, there is still the risk of government failure to assess. Risk, uncertainty, imperfect, incomplete and asymmetric information are conditions increasing the likelihood of both market and government failure.
2008 / 3
248
R. Veugelers
These appropriation problems are more severe the more basic and generic the research is. In addition, depending on the nature of knowledge spillovers, firms may have a too low private incentive to engage in cooperative R&D, as compared to what would be socially optimal. The typical instruments deployed to tackle these market failures are: (i) direct support in the form of subsidies or tax credits; and the provision of public R&D resources to work with the private sector; (ii) indirect support through infrastructure design, including policy areas as competition policy (including towards cooperation in R&D); patent policy; procurement policy and trade policy.
IV.
A need for SME policy intervention: SME specific market failures
Review of Business and Economics
In addition to this general innovation – policy-agenda, is there is a need for a specific SME innovation policy? This requires a number of conditions: (i) there are specific barriers that are hindering smaller sized firms to innovate: i.e. a market failure specifically faced by SMEs or a larger market failure for SMEs. (ii) government can effectively target these SME market failures, either by devising SME specific instruments or by favouring SMEs in horizontal policies, that does not founder on government failure. The most important barrier to innovation, where market failure abounds, is access to finance. There might be substantial differences in this respect between small and large firms. First, problems of asymmetric information are usually more acute for younger/smaller firms, as discussed supra, and hence the risk premium that smaller firms are required to pay is often much higher. Second, R&D projects undertaken by small firms are, ceteris paribus, riskier than if done by larger firms, even if they are exactly the same in terms of technological goals. This is so because younger/smaller firms are disadvantaged relative to large firms in terms of a wide range of competencies and experience that are complementary to R&D, be it in marketing, pure management, access to complementary know-how, etc. Lastly, imperfections in capital markets usually affect small firms more than large firms. First, the availability of internal financing, which has been shown to be important in the context of R&D, is normally less constraining for older/larger firms than for smaller/younger ones. Second, access to global capital markets is easier/cheaper for larger firms. Thus, government support to R&D meant to bridge over those imperfections ought to be channelled more towards small firms than to larger ones (see also Hall (2005)). The second market failure, arising from the difficulty of appropriating the benefits from innovation, also may play differently for large than for small innovating
The Role of SMEs in Innovation in the EU: A Case for Policy Intervention?
249
Any other barriers which SMEs might face when innovating, such as access to skills, access to information, lack of absorptive capacity, regulatory burden, require investments to develop the critical capacities and to grow to the critical size needed to tackle these barriers. This trajectory can be impeded by financial constraints and failure to appropriate, where market failure exists. So far we have shown that the most severe barrier for companies to innovate, the financial barrier, is also the one most clearly associated with market failure. And even particularly for SMEs. But to access finance, reputation and collateral are important, which points to firm age as an even stronger constraint than firm size; and also riskiness of investment project matters. So companies likely to be most constrained are the young, radical small innovators. When policy makers want to target specifically SME sensitive barriers, where market failure exists, i.c. the financial constraints, they should recognize the heterogeneity within the SME population. The YICs being more affected by financial market imperfections deserve special
Review of Business and Economics
On whether small firms will be less likely to be engaged in cooperative R&D, the literature is less clear (Cassiman & Veugelers (2002), Belderbos et al. (2004)). But the literature does point out the need for an own absorptive capacity to be able to benefit from cooperation. SMEs may again lack the necessary scale to develop such an own absorptive capacity.
2008 / 3
firms. The patent system provides an improved appropriation of the returns from innovation. But even in the presence of a patent system, SMEs may find it more difficult to appropriate the returns from their innovations. First, the cost of patenting may be too high, especially since efficient appropriation often requires a portfolio of patents, which will be expensive, especially in Europe. In combination with the financial market failure, which will hamper SMEs to obtain funds for financing these patent costs, this will translate into a specific barrier to innovation for small innovative firms. Second, appropriation requires complementary strategies to patents, like trademarks, secrecy, lead time, complexity; building this portfolio of different appropriation strategies, might again require a critical scale which SMEs may lack (Cassiman & Veugelers (2002)). Thirdly, appropriation often requires beyond patent protection, the presence of complementary assets (like marketing, access to distribution channels etc...) for which again SMEs may lack the critical scale (Teece (1986)). Fourthly, SMEs may be unable to appropriate the surplus created by subsequent innovations that build on the knowledge introduced by the initial innovation. Particularly for initial innovations, these follow-up innovations may be quite large. Follow-up innovations may be done by large firms, which may not be fully compensating the pioneering SME, when the market for technology disfavours small, young and risky projects. Hence, as small firms are less able to capture the externalities that they generate, the larger would be the divergence between their social and private rate of return of their R&D.
2008 / 3
250
R. Veugelers
policy attention to support their introductions of more radical innovations. In addition, as these more radical innovations typically will induce follow-up innovations, the divergence between the social and private rate of return is higher for these firms, particularly if they are less able, because of their lack of size and experience, to appropriate the spillovers from their innovations.
V.
Which policies do we currently have? Trends in SME innovation policy
Combining financial and appropriation market failure stories, the case for government intervention is particularly strong for young radical innovators. But this is conditional on being able to avoid government failure. Even if policy is well targeted, there could still be failures in implementation. Government failure is all the more likely in settings characterized by risk, uncertainty, imperfect, incomplete and asymmetric information. This and the next section will try to assess the scope for government failure, by taking a look at actual policies being implemented in the EU and how they have been evaluated on their effectiveness.
Review of Business and Economics
A. Trends in innovation policy in general
Targets, such as the 3% Barcelona target for R&D spending, although few EU countries are on track to meet them, have nevertheless focused attention to and support for innovation at the highest levels of government. While some countries have made incremental changes in their main policy trends, others have made a significant break with past policy trends. Some of these reforms were prompted by evaluations, which have become a more important part of the governance and implementation of science and innovation policy in several countries (e.g. Austria, Finland, the Netherlands, and the United Kingdom.) Several countries with numerous small programmes have recently attempted to streamline support and focus programmes on systemic failures in innovation, especially in the area of networking and co-operation (OECD (2007)). Beyond the targeting of public R&D, governments, more than ever, wish to raise business investment in R&D. Direct support to business innovation in the form of competitive grants or subsidised or guaranteed loans remains important even if use of indirect schemes such as tax credits has tended to increase. Consequently, the main mechanisms for supporting business innovation remain competitive and merit-based grant programmes, but tax incentives and support for firm creation and start-ups and other programmes that focus on co-operation are gaining ground.
The Role of SMEs in Innovation in the EU: A Case for Policy Intervention?
251
2008 / 3
A lot of effort is going on at the policy level to boost access to venture capital. Many countries have embedded concepts of networking and consortium development into their innovation support programmes and initiatives in order to foster greater collaboration, especially around regions and clusters, among firms and between firms and public research bodies (OECD (2007)). As an illustration of these trends, we can have a look at some recent new measures taken by EU member states (Figure 7).
Source:
EC-Trendchart.
The new trend confirms the importance governments attach to policies and specific actions on building clusters and bridging business and science, especially as regards support to joint projects by SMEs and universities. As the discussion in section IV has demonstrated, the private incentives to collaborate may be too low, calling for a policy support to promote collaboration. The policy case for advantaging SMEs in collaborating with universities is less clear-cut. At least, differentiating among SMEs will be important here, as only a very small fraction of innovating SMEs are concerned.
Review of Business and Economics
Figure 7. New government actions in 2005 for Innovation in EU Member States.
2008 / 3
252
R. Veugelers
A number of new measures have been introduced in several Member States to support the creation and growth of innovative companies, especially by improving their access to funding as well as providing support for innovation commercialisation.11 Positive to note is the emphasis on access to risk capital, in public-private partnership with the financial sector, addressing the capital market failure. Also positive is the differentiating with respect to the profile of innovative SMEs (age (start-up) and (fast) growth), as the financial barriers play differently for these groups.
B. Preferential financial support for SMEs: how much of the public budget
for R&D support goes to SMEs?
Review of Business and Economics
The majority of support for business R&D is still in the form of grants or subsidies. These instruments are targeted at tackling the access to finance and the appropriation problem, both rooted in market failure. As section 4 has discussed, a preferential treatment of SMEs in government support for R&D can be justified by the larger barriers faced by SMEs to access finance and to appropriate the returns from innovation.
Figure 8. Share of SMEs in business R&D and in government-financed business R&D. Source:
OECD (2007), STI Indicators.
253
The Role of SMEs in Innovation in the EU: A Case for Policy Intervention?
2008 / 3
To get an idea of the preferential treatment of SMEs in actual government support for R&D, Figure 8 compares the importance of SMEs in government financed business R&D with the share of SMEs in expenditures on R&D.12 If the former far exceeds the latter, this is clear evidence of a favour/focus towards SMEs in government support for R&D policies. This would be consistent with governments trying to address the issue of financial constraints for SMEs caused by capital market failures and to bridge the gap between social and private returns for R&D for SMEs. In most countries, this seems to hold: e.g. for EU-15 countries like Portugal, Belgium, Ireland, Spain and for New Member States like Hungary. This focus on SMEs in government support for R&D also holds for countries where large firms are predominant in business R&D like Germany and Finland. While in the US and France, SMEs receive a more or less proportional share of government support, the UK is least focused on SMEs. For Germany, the share of SMEs in government support for business R&D has increased over time: while it was around 15% in 2001 (OECD, STI Scoreboard 2003), this has risen to over 20% in 2005. On the differentiation within the SME receiving grants population, particularly the share that goes to young, radical innovations, unfortunately no systematic information is available. The graph displays however some evidence on preferences for the very small enterprises (< 50), which prevails particularly in Finland, Germany, Hungary, Portugal, Belgium and Austria.
Proportion of enterprises with innovation activity having received public financial support for innovation activities, EU, 1988-2000 (%) Public funding (any source)
Local or regional authorities
Central government
The European Union
EU’s 4th or 5th framework programme
Total
29
15
15
7
4
Small enterprises
27
15
12
5
3
Medium-sized enterprises
31
14
18
8
5
Large enterprises
48
19
33
20
15
Figure 9. Public support for innovation: by size class; and by type of government. Source: Eurostat, NewCronos (theme9/innovat/inn_cis3). Source:
Eurostat, Innovation in Europe, CIS3 results, 2004.
Review of Business and Economics
Another piece of evidence comes from the innovative companies reporting in the EU-wide EUROSTAT Community Innovation Survey (CIS-III 2000-2002), whether they received any public funding (no information on size of the support) (Figure 9).
2008 / 3
254
R. Veugelers
This information shows that on average for the EU, large innovative firms are more likely to receive financial support than small innovative firms (see also Aerts & Schmidt (2008) for evidence on Belgium). The bias in favour of large firms diminishes with distance: local or regional funding is less big firm biased, as compared to national and European funding, underscoring the importance of especially the local policy level for SMEs.
Review of Business and Economics
C. Some recent EU policy initiatives for SMEs and innovation
Recognizing that most policy action for stimulating innovation at SMEs is at the level of Member States, the EU nevertheless has a few important policy levers. We list only the most important recent initiatives: As part of its Competition Policy competence, the Community State Aid rules set the conditions for Member States to grant aid to innovation, risk capital and to SMEs. The recently revised rules in the area of Innovation as well as Risk Capital, have introduced a balancing test, requiring to assess ex ante whether a market failure exists and in a second step, assessing that the aid is an appropriate instrument for correcting the market failure, before assessing the distortion of competition. The aim of the approach is to come to a better targeted aid. made it easier for Member States to grant innovation aid and promote access to risk capital for SMEs (by increasing the minimis thresholds, increasing the intensity of aid to SMEs, more flexible rules for SME costs for IPR, hiring staff, accessing innovation services, innovation clusters, collaborative R&D...). introduced the category of Young Innovative Companies, with a specific de minimis threshold for aid. The participation of SMEs in the Cooperation Programme of the 7th RTD Framework Participation is stimulated by various measures: a 15% participation target for SMEs, a higher aid intensity, administrative simplifications such as abandoning bank guarantee requirements. The EUROSTARs programme, launched in 2007, jointly with EUREKA, provides support for SMEs for international, market-driven, collaborative research projects. The Competitiveness and Innovation Framework Programme CIP, 2007-2013, earmarked part of its budget for improving access to finance for SMEs. A significant share of the Structural Funds (2007-2013) is earmarked for support to the creation and growth of SMEs, notably to improve their innovative capacity and investment in R&D and skills. Included is the JEREMIE initiative, launched jointly with the EIB and the EIF, to support Member States to develop financing instruments, such as venture capital, guarantees and micro-finance.
The Role of SMEs in Innovation in the EU: A Case for Policy Intervention?
255
These recent EU initiatives complement the general trend in Member States SME innovation policy towards greater focus on access to finance issues and stimulating cooperation.
VI.
2008 / 3
An integrated business and innovation support network to assist SMEs in crossborder activities.
Effectiveness of past policies
Have the policies that have been introduced in the past been successful in improving innovation? What do we know from evaluation studies of past policies? What can we learn from them for designing better targeted policies, avoiding government failure?
A. Perspective from the recipients on policies for innovation in the EU:
a call for more differentiation
Review of Business and Economics
Interesting insights to evaluate actual policies come from surveys of SMEs. These surveys need to be handled with the usual care, taking into account subjectivity and selection bias. In addition, they only relate to the private benefits of the individual recipients from policy intervention, ignoring the social benefits. The ECENTR’s Innobarometer 2004 reported results from a survey among SMEs across EU25 countries who had experiences of innovative activities (N = 4534), which includes questions on public support. A number of interesting findings emerged: Only 9% received government support to introduce innovations. And there is no higher probability of getting support for the very small firms (< 50 (8%)) or the young firms (< 5 years (8%)). Successful innovators (17%) reported a higher incidence of government support, which could be suggestive of the picking winners-bias or, more favourably, reflecting the beneficial outcome of public support. For those who received some form of public support, 71% responded negatively to the question whether this public support was crucial to any of its innovation projects, such that the innovation would not have been developed without the support. Again no significant difference for very small firms or young firms.13 A similar indication of “lack of effectiveness” is illustrated in Figure 10. For those firms which received some form of public support, 26% reported no notable effect on their innovation process. 20% reported support to have helped reduce costs, 18% to improve quality of innovations. This cost-reduction effect was even higher for Medium Sized Firms (250-500) (27%).
Review of Business and Economics
2008 / 3
256
R. Veugelers
For one in two recipient firms, the most valuable form of public support they received was for staff training. Following next, 44% of these firms consider that public support for implementing innovative processes was of most benefit. At the bottom end of the list we find “Advice services for innovation activities” and “Helping an innovation network in which you participated” which was rated as most valuable by only 21%. Only 13% participated in innovation networks. But this networking by SMEs is much more prevalent among successful innovating firms (36%) or in successful innovation countries, like Sweden (37%), Finland (28%) and Denmark (25%), suggestive of a link between networking and innovative performance. For those firms engaged in networks, 36% received public assistance or subsidies. For successful innovations this increased to 58%. For Medium sized (250-500), 42%.
Figure 10. Survey evidence on results from government support for SMEs in the EU. Source:
Innobarometer 2004
These results, although they have to be interpreted with caution given the small sample size of firms on sub-items, suggest that the overall effectiveness of policy cannot be taken for granted. The popularity of support for training and implementing innovation processes, reflects that for the majority of SMEs diffusion and absorption of innovation is of higher concern than creating cutting-edge new research. But the question nevertheless is why a market failure would exist which government intervention would and could address. The low value attached on average to support for innovation networks, and the low occurrence of participation
The Role of SMEs in Innovation in the EU: A Case for Policy Intervention?
257
Overall, the results can be interpreted against a “one size fits all” approach towards SMEs, in favour of a more targeted SME policy, applying specific instrument to target the market failure induced barriers for specific SMEs sub-populations.
2008 / 3
in innovation networks by SMEs, again reflects that for the majority of SMEs, innovation networks are not of major importance. This contrast with the increasing emphasis of innovation policy makers on these instruments (cf supra). The same holds for “advice services”.
B. Insights from academic studies: a call for more evaluation and
experimentation
But they continue to conclude that “the existing literature as a whole is subject to the criticism that the nature of the “experiment(s)” that the investigators envisage is not adequately specified”, leaving a lot of methodological issues (like sample selection bias). More recent studies have come up with better data and methodologies (e.g. Wallsten (2000) for the US, Czarnitzki & Aerts for Flanders & Germany (2007), Toivanen et al. for Finland (2007)), but the conclusions are still ambivalent, with some studies finding positive, others negative effects. Impeded by a lack of sufficiently detailed data, this line of research is also not able to identify whether and which characteristics of the policy intervention would be more favourable to positive effects on private R&D (such as details of the subsidy program, types of firms targeted (SME, YICs), the presence of complementary framework conditions) (Klette et al. (2000)). In a very recent study, Colombo et al. (2008), provide evidence on Italian data for different effects of funding depending on the stage of development of the recipient firm. They found that young new technology based firms benefit more (in terms of firm growth) than mature ones from financial support, especially if public funds are allocated through a selective evaluation process.
Review of Business and Economics
A growing body of economic papers are produced on evaluating the effects of public R&D spending, using econometric analysis on time series and cross section data at various levels of aggregation (firm, industry, country). Most of these studies examine the impact of subsidies on private R&D spending, correcting for other determining firm, industry and market characteristics affecting private R&D spending. The majority of the empirical literature thus focuses on the issue on whether public R&D spending is complementary and thus “additional” to private R&D spending, or whether it substitutes for and tends to “crowd out” private R&D? In a survey of this literature, David, Hall & Toole (2000), conclude that “the findings overall are ambivalent”, although on average there is more evidence in favour of positive effects, rather than negative.
2008 / 3
258
R. Veugelers
Increasingly, a number of funding agencies are building in ex post evaluation of specific programs, combining quantitative and qualitative analysis. Program evaluation is much more prevalent in the US than in Europe, for example the US Small Business Administration’s SBIR program. Lerner (1996), when evaluating the US-SBIR program, examined whether public funding of early-stage high-tech firms would stimulate significant growth for these firms. On the question how to deliver the necessary public funding in such a risky environment without losing the monitoring ability of private venture capital firms and without trying to ensure such monitoring with clumsy and costly contracts, he provides an excellent review of the literature suggesting the problems faced by government when it provides venture capital. If the expertise and resources required to monitor the early-stage firms are not available to government, then public funding at best just stimulates short-run gains. His analysis of the SBIR program confirms that positive findings were confined to firms located in areas with also substantial private venture capital activity. Wallsten (2000) trying to correct Lerner’s analysis on selection and endogeneity, however finds no significant evidence for positive effects.
Review of Business and Economics
Also in Europe the trend is towards more evaluation of specific programs.14 Nevertheless, an OECD (2007) survey of EU Member States performance of SME state aid evaluations, revealed a huge heterogeneity. While 11 countries did not respond, most EU Member States (N = 11) reported occasional evaluations of parts of their aid programs for SMEs. Only 2 countries reported a consistent evaluation of all aid policies (the Netherlands and Ireland). 3 Countries reported no evaluations.
VII.
Concluding with some elements for a good “policy approach” to SMEs and innovation
Given the lack of adequate data and the many problems with evaluation studies, the only conclusion that can be drawn from the empirical evidence at this stage is that we don’t know enough, yet, to tailor SME innovation policy and to evaluate the risk of government failure. Nevertheless, we can identify a number of characteristics which a good policy approach should have to stimulate to increase the contribution of SMEs to innovation.
i. SME innovation policy needs a systemic approach, creating the framework conditions for a favourable environment for innovation The high probability of government failure to affect growth and innovation, by a too detailed top-down interventionist approach, puts framework condition central when designing policy interventions.
The Role of SMEs in Innovation in the EU: A Case for Policy Intervention?
259
2008 / 3
Since SMEs need to find a complementary overall innovative environment to interact with in co-optition, SME policy needs to fit into an overall innovation policy. But the policy mix required goes well beyond innovation policy proper. Framework conditions need to be in place for firms having the incentives to be engaged in innovation activities. Innovation is fostered by policies that promote competition, and safeguards innovative firms access to markets, finance and skills. An EU policy should include a strong emphasis on a coordinated competition policy (including State Aid) that vigorously guards against incumbent behaviour that bars the entry and growth of more efficient, innovating firms and would provide an umbrella for inefficient firms from exiting. At the same time, a vigorous and continued pursuit of further market liberalisation and finishing the single market program is needed to provide dynamic firms with access to a large integrated business environment to source from and supply to. Particularly services remain an unavoidable policy priority in Europe. While competition policy, open trade policy and single market reforms are important, they are not sufficient to improve the creative destruction process. The effects of these will not be reached without some deep structural reforms of Europe’s labour markets and higher education system, which are needed to provide innovative firms a smooth access to skills and know-how.
The empirical evidence easily shows the barriers faced by SMEs when innovating (cf. figure 5). But this in itself does not yet make a case for government intervention in support of SMEs innovating. A first question to ask is whether these SME barriers are caused by market failure. Not all barriers are mapped into market failure. And secondly, if there is a case for market failure, can governments redress this failure, without introducing new barriers, i.e. assessing the potential for government failure. The recently revised EU state aid rules, with a “refined economic approach” and a “balancing test” carry in part a similar analysis. The first applications of the new procedure indicate that such test can be implemented in practice.
iii. SME innovation policy needs differentiation Firms face different barriers involving market failure during growth than during start-up. As a consequence, to support the creation of firms requires a different policy agenda than enabling subsequent development and growth. While policy ma-
Review of Business and Economics
ii. SME innovation policy needs a failure test, assessing first the case of market failure and secondly assessing whether the policy instrument is the most appropriate instrument to alleviate the market failure
2008 / 3
260
R. Veugelers
kers are usually keen to stress the importance of the creation of new firms, especially high-tech, there were still relatively little signs of a coherent strategy towards their subsequent development (and the development and exploitation of the technological and commercial capabilities which they create) and how they might be helped to grow to medium and large firm status. Especially the young, rapidly growing high-technology firm may experience difficulties raising second and subsequent tranches of finance, particularly when it needs to make a significant quantum increase in the size of its business. Supporting the development of private capital markets, especially the high-risk, early stage segments, and/or public funding can be warranted to solve the market failures faced by YICs. Effective IPR protection is often essential to enable a firm to raise finance, to access new markets and to appropriate the returns from new acquired market positions. These firms usually cannot afford to pay market salaries to experienced executives and can only attract them by giving them a share in the equity or profits of the growing business. Appropriate tax treatment of share options or profit related bonuses helps for such firms to attract the expertise they need. Favourable tax treatment of R&D can help finance growth particularly where payment of tax credits are made even when the firm is not making sufficient profits to offset the value of the credits against its potential tax bill.
Review of Business and Economics
iv. SME innovation policy needs experimentation and evaluation Efficiency and effectiveness of policy are of increasing importance in general for (innovation) policy making due to tight government budgets. We spend significant resources on “cures”, but don’t really know what “cures” work and if they work, under what circumstances. This also holds for SME innovation policy. It calls for an explicit build-in of evaluation, already at the design phase of the policy process. What is needed is an infrastructure (1) to collect and organize in a comprehensive and coherent way the data needed for policy making; (2) to set procedures for the ongoing evaluation of the effectiveness of the policies; (3) to evaluate, research and discuss long term policies. Cross-national coordination can help to build critical scale in evaluation expertise, learning from best-practices and benchmarking.
NOTES 1. Faculty of Business and Economics, K.U.Leuven, Naamsestraat 69, 3000 Leuven, Belgium,
[email protected]. KULeuven – MSI & SOOS, EC-BEPA and CEPR. The author wishes to thank Isabel Grilo, Josefina Monteguada, Vitor Gaspar, Gert-Jan Koopman, and Cedric Schneider, who have all contributed on various notes and working papers, which are used as inputs for the article, particularly, the contribution in the EC-BEPA Monthly January 2008 issue. This article does not necessarily reflect
The Role of SMEs in Innovation in the EU: A Case for Policy Intervention?
3.
4.
5. 6. 7.
8.
9.
11.
12.
13.
14.
Review of Business and Economics
10.
the view of the EC. All views, omissions and errors are on the account of the author only. This section draws heavily on the EC-BEPA Monthly January 2008 issue on SMEs and Innovation. Firms with < 250 employees represent 17% of total R&D expenditures, firms with less than 50 employees account for 5%. (Source: Eurostat, R&D statistics, Data for EU-27 (excl NL), 2004). The average R&D intensity of SMEs in Europe is 0.34% versus 0.53% in the US. This compares to an overall average R&D intensity in Europe of 1.17% versus 1.57% in the US. (DGRTD Key Figures 2007). Since the figure is on the basis of CIS (Community Innovation Survey) data, they are only available for EU countries. There are no comparable US data. The statistical correlation is very high (0.7) (DGENTR, SMEs and Innovation, Note for the EPC, 2007);. YICs are defined as firms < 6 years old, with < 250 employees, not part of a group, having an R&D intensity of > 15% and having introduced new products and/or processes. This is the definition of YICs used in EC-State Aid regulation. For sales of new products (incremental as well as radical), the superior performance of YIC is less outspoken: while YICs have an average of 51.7% this is “only” twice the average share. This superior performance of YICs is confirmed in multivariate analysis, correcting for other firm and industry characteristics (Schneider & Veugelers (2008), KUL-MSI Working Paper). In 1994 the Small Business Administration Office of Advocacy prepared a list of breakthrough innovations made by small firms during the 20th century (reported in Baumol (2002)). It is impressive going literally from A (airplane) to Z (zipper) with many innovations that have been crucial to the economy. Other studies on the distribution of innovations certified as “significant” by industry experts, confirmed that small (as well as large) firms outperform medium-sized firms for the US (see Simeonydis, 2006 for an overview). Econometric studies typically show that social returns are substantially larger than private rates of return from R&D (Griliches (1992) report estimates of social rates of return as high as 100%). Examples are the Enterprise Capital funds in the UK, the SME Credit Guarantee Scheme in the Netherlands, the TechnoPartner initiation in the Netherlands, the ERP Innovation Program in Germany. The share of SMEs in R&D expenditures differs markedly across countries (left panel), as was also demonstrated in section 2. In countries like Finland, Sweden, UK, France, France around or less than 20%, Germany around 10%. US also has a relatively small share of SMEs in total R&D, but within the SME population, almost half of their R&D is accounted for by very small firms < 50. There are some interesting differences across countries, but sample sizes are too small to be taken seriously. Countries with more “effective” support are Denmark (49%), Sweden (55%) UK (59%) and Germany (67%), countries with “less effective” support are Ireland (83%), Portugal (80%), Finland (78%), Netherlands (78%), Italy (77%). See OECD (2007), Framework for the Evaluation of SME and Enterpreneurship Policies and Programmes, for some examples of evaluations of SME-Innovation Programs. (E.g. the example of Science Parks in Sweden is a well designed evaluation study, finding no statistically significant effects for firms participating in these Parks).
2008 / 3
2.
261
2008 / 3
262
R. Veugelers
REFERENCES Acs, Z. and D. Audretsch, 1987, Innovation, market structure and firm size, The Review of Economics and Statistics, 71, 567-574. Aerts, K. and T. Schmidt, 2008, Two for the price of one? On additionality effects of R&D subsidies: A comparison between Flanders and Germany, Research Policy, 37, 5, 806-822. Baumol, W. 2002, The Free-Market Innovation Machine, Princeton University Press. Belderbos, R., Carree, M., Diederen, B., Lokshin, B. and R. Veugelers, 2004, Heterogeneity in R&D Cooperation Strategies, International Journal of Industrial Organisation, 22, 8-9, 1237-1263. Cassiman B. and R. Veugelers, 2002, R&D Cooperation and Spillovers: Some Empirical Evidence from Belgium, American Economic Review, 92, 4, 1169-1184. Cohen, W. and S. Klepper, 1996, Firm size and nature of innovation within industries: the case of process and product R&D, Review of Economics and Statistics, 232-243. Cohen, W. and R. Levin, 1989, Empirical Studies of Innovation and Market Structure, in Schmalensee R; and R. Willig (Eds.) Handbook of Industrial Organisation, North-Holland, chapter 18, p. 1060-1107. Czarnitzki, D. 2006, Research and Development in small and medium-sized enterprises, the role of financial constraints and public funding, Scottish Journal of Political Economy, 56, 3, 335-357. David, P., Hall, B. Toole, A., 2000, Is public R&D a complement or substitute for private R&D? A review of the econometric evidence, Research Policy, 29, 4-5, 497-529. Griliches, Z., 1992, The search for R&D spillovers, Scandinavian Journal of Economics, 94, 29-48.
Review of Business and Economics
Hall, B., 2005, The financing of innovation, in S. Shane (Ed.) Blackwell Handbook of Technology and Innovation management, Blackwell Publishers, Oxford. Kamien & Schwartz (1982), Market Structure and Innovation, Cambridge University Press. Klette, T., Moen, J. Griliches Z., 2000, Do subsidies to commercial R&D reduce market failures? Microeconometric evaluation studies, Research Policy, 29, 4, 471-495. Lerner, J., 1996, The government as venture capitalist: the long-run effects of the SBIR program, NBER WP 5753. Symeonidis, G., 2006, Innovation, firm size & market structure: Schumpeterian hyptheseses and some new themes, OECD Economics Department WP 161. Rothwell R. and M. Dodgson, 1991, Technology strategies in small firms, Journal of General Management, 17, 45-56. Storey, D. and B. Tether, 1998, New technology based firms in the European Union, Introduction to the special issue, Research Policy, 26, 9, 933-946. Takalo, T., Tanayama, T. Toivanen, O., 2007, Selection or self-rejection? Applications into a voluntary treatment program: the case of R&D subsidies, HECER DP, Helsinki. Utterback, J.M., 1994, Mastering the Dynamics of Innovation, Boston, Mass.: Harvard Business School Press. Schneider, C. and R. Veugelers, 2008, Why Young Innovative Companies matter? Microeconometric evidence from Germany, KUL-MSI Working Paper in progress. Teece, D. 1986, Profiting from technological innovation: implications for integration, collaboration, licensing and public policy, Research Policy, 15, 285-305. Wallsten, S. 2000, The effects of government-industry R&D programs on private R&D: the case of the Small Business Innovation Research Program, RAND Journal of Economics, 31, 1, 82-100.
1
A. Renders & S. Vandenbogaerde
2
2008 / 3
Corporate Governance and Earnings Management: Evidence from Europe
ABSTRACT3
JEL CLASSIFICATION
KEYWORDS
I.
M41, G34 Corporate governance, Earnings management
Introduction
Since the outburst of corporate scandals in the U.S. and Europe, regulators, investors, and companies have developed an increased interest in the good governance of companies. At the regulatory level, this has led to the development of corporate governance laws and recommendations, such as the Sarbanes-Oxley Act (2002) in the U.S. and the Code Lippens (2004) in Belgium. Since the outset, it has been clear that the approach of American and European regulators towards improved corporate governance is very different. While the U.S. Congress voted the
Review of Business and Economics
In response to the many corporate scandals, regulators developed laws and recommendations in order to improve the governance of firms. This paper studies the relation between different corporate governance mechanisms and earnings management in a European context. Prior research focused mainly on U.S. firms. However, Europe is quite distinct from the U.S. in terms of country- and company-level characteristics. Our results show that better corporate governance is associated with less (income-increasing) earnings management by European firms. More specifically, we find that better shareholder rights, fewer anti-takeover mechanisms, and a good functioning board of directors and director committees lead to higher accounting quality. We also find that the extent of corporate governance recommendations issued at the country level influences the effectiveness of corporate governance practices in constraining earnings management. These relations are especially evident in code law countries, where the opportunities for earnings management are larger than in common law countries.
Review of Business and Economics
2008 / 3
264
A. Renders & S. Vandenbogaerde
Sarbanes-Oxley Act into law, European regulators and stock exchange oversight boards developed best practices and recommendations. These best practices and recommendations usually incorporate the “comply-or-explain” principle, stimulating firms to follow the recommendations or explain why they deviate from it. As a consequence, compared to U.S. companies, European companies are to a large extent free to determine to which degree they will incorporate the recommendations.4 However, the goal of the laws and recommendations remains the same, namely to improve the quality of the accounting information provided to users of the financial statements. To this end, the codes and laws prescribe for instance that companies have an independent board of directors and install an audit committee in order to monitor management more closely. Prior research has already examined the relation between corporate governance practices and financial statement quality extensively. On the whole, this research finds that improved corporate governance leads to lower earnings management,5 fewer accounting restatements, and a lower likelihood of financial statement fraud. However, most prior studies have focused on U.S. companies and there is relatively little evidence on the relation between corporate governance mechanisms and accounting quality outside the U.S. Furthermore, earlier studies have concentrated on the independence of the board of directors and the audit committee, while other mechanisms such as shareholder rights and takeover defences have largely been ignored. In this paper, we add to prior research by focusing on European firms. Due to different institutional characteristics, not in the least in the way corporate governance is approached, the relation between corporate governance and earnings management may be different and the results from earlier studies may not be extended to Europe. We also contribute to the existing literature by looking at a broader range of corporate governance mechanisms. Besides the board of directors and the audit committee, we examine takeover defences and shareholder rights. We also combine these elements into one aggregated corporate governance score and test its relation with the level of earnings management. We study furthermore whether the extent of corporate governance recommendations formulated at the country level affects the effectiveness of corporate governance practices in constraining earnings management. Finally, we examine the relation between corporate governance and earnings management across countries from a different legal origin. A first reason why the relation between corporate governance and earnings management might be different in European companies is that managers of European firms may have fewer stock market incentives to manage earnings compared to their U.S. counterparts. European countries are characterised by small capital markets and firms rely heavily on bank financing. As a consequence, European companies face less market pressure to show strong financial results. Richardson et al. (2002) show, for instance, that capital market pressures are a primary motivation to adopt aggressive accounting policies. European managers are also rewarded to a
Corporate Governance and Earnings Management: Evidence from Europe
265
2008 / 3 Review of Business and Economics
lesser extent with shares and share options, leading to a weaker focus on stock prices. Secondly, European companies may have other incentives than market pressure to manage earnings. For example, the main conflict of interest in European firms lies between controlling and minority shareholders (Becht and Roell (1999)), while in the U.S. the main conflict lies between owners and managers of the firm (Jensen and Meckling (1976)). The controlling shareholders of European firms have been shown to expropriate value from their firms (Nenova (2003), Dyck and Zingales (2004)), resulting in earnings management to conceal this expropriation. So even in the absence of capital market pressures, European companies may still manage earnings. Research in Belgium has shown for instance that Belgian companies do manage earnings (Vander Bauwhede and Willekens (2003)). Last but not least, European companies may have more opportunities to manage earnings than U.S. firms. As mentioned before, corporate governance in Europe leaves companies a considerable amount of discretion. Second, the level of legal investor protection is generally lower than in the U.S. (La Porta et al. (1998)), resulting in lower financial statement quality (Leuz et al. (2003)). Finally, the quality of the European accounting standards is assumed to be relatively low, at least in the period we examine (i.e., before the introduction of IFRS) (Leuz and Verrecchia (2000), Barth et al. (2006)). In sum, Europe provides us with an interesting setting to test the relation between corporate governance and earnings management. Our sample countries are characterised by weak accounting standards and few governance regulations. They exhibit furthermore a wide variation in the level of investor protection and in the amount of corporate governance recommendations. Given the voluntary corporate governance, the low levels of investor protection, and the ample opportunities to manage earnings, it is a priori an open question whether better corporate governance leads to lower earnings management in this setting. Furthermore, the impact of corporate governance on earnings management might depend on the institutional characteristics of the countries. In this study, we address five research questions. First, are good corporate governance practices associated with less earnings management for European companies? Second, if so, which elements of corporate governance reduce the amount of earnings management? Third, does the impact of corporate governance practices on earnings management depend on the extent of corporate governance recommendations issued at the country level? Fourth, do European companies engage more in income-decreasing or income-increasing earnings management and does corporate governance constrain both types? And finally, does the relation between corporate governance practices and earnings management differ according to the legal origin of countries? Using a sample of the largest listed companies in Europe over the period 19992003, we find that good corporate governance leads to lower earnings manage-
Review of Business and Economics
2008 / 3
266
A. Renders & S. Vandenbogaerde
ment. The results also indicate that a good functioning board of directors, strong shareholder rights, and few takeover defences lead to less earnings management. Furthermore, in countries with extended corporate governance recommendations, corporate governance practices are found to have a stronger negative impact on the amount of earnings management than in countries with few corporate governance recommendations. Next, the results indicate that the majority of the companies engage in income-decreasing earnings management. However, good corporate governance only seems to constrain income-increasing earnings management. Finally, we find that the results are largely driven by companies in code law countries. Our study is closely related to the study by Peasnell et al. (2005). They focus on U.K. companies using earnings management to meet earnings targets in the preand post-Cadbury period. They find that the proportion of non-executive directors does not constrain earnings management in the pre-Cadbury period, though it does constrain earnings management in the post-Cadbury period. So after the introduction of the Cadbury Code, the board of directors has become a more effective tool in constraining earnings management. Our study differs however in three ways. First, we focus on a more general measure of earnings management. Second, our sample consists of companies from different European countries. And finally, we take the extent of corporate governance recommendations issued in each country into account by using an index to score countries on their recommendations and we test whether this influences the effectiveness of corporate governance practices. The remainder of this paper is organized as follows: in section II we review the existing literature. In section III, we describe our measure of earnings management and outline the model. Section IV presents the sample selection and section V discusses the multivariate results. Section VI summarizes the results and concludes.
II.
Literature review
In this paper we focus on internal corporate governance mechanisms, i.e., mechanisms that are installed by the company itself. We distinguish between different corporate governance mechanisms, such as board structure and functioning, including any committees installed by the company, takeover defences and shareholder rights. First, with regard to board structure and functioning, prior studies have shown that a board of directors comprising a majority of independent directors is associated with higher financial statement quality. For instance, Beasley (1996) and Dechow et al. (1996) find that firms with insider-dominated boards are more likely to have financial statement fraud or GAAP violations, respectively. Klein (2002) and Peasnell et al. (2005) show that companies with a more independent board of directors have lower levels of earnings management.
Corporate Governance and Earnings Management: Evidence from Europe
267
2008 / 3 Review of Business and Economics
Klein (2002) also finds that earnings management is less pronounced in firms that have an audit committee comprising a majority of independent directors. Abbott et al. (2004) shows that an independent and active audit committee results in a lower likelihood of financial restatements. Xie et al. (2003) concludes that audit committee activity and financial sophistication are important factors in constraining earnings management. In contrast, Agrawal and Chadha (2005) finds no evidence of a relation between an independent board or audit committee and the likelihood of a restatement. Second, the little evidence on the relation between corporate takeover defences and earnings management is very conflicting. Takeover defences may shield managers from the discipline of the market for corporate control, leading to managers placing their own private interests before those of the shareholders and hence engaging in higher levels of earnings management. This is evidenced by Yu (2006) finding that takeover pressures discipline managers and are thus associated with less earnings management. Also Barber et al. (2007) finds that companies with few anti-takeover mechanisms are less likely to have accounting restatements. However, the opposite relation may also hold. Managers in firms with few anti-takeover mechanisms are vulnerable to takeovers and may thus have incentives to boost earnings. As a consequence, mechanisms that reduce or eliminate takeover pressures may be associated with less earnings management. Jiraporn (2005) finds for example that poison pills reduce the amount of earnings management. Finally, at the country level, Leuz et al. (2003) shows that higher shareholder rights lead to less earnings management. At the company level, however, Chen (2007) finds that companies that do not adhere to the one share/one vote rule are less inclined to manage earnings. A possible explanation is that companies with weaker shareholder rights are protected from shareholder intervention, resulting in lower incentives to boost earnings. So in sum, the findings of prior studies indicate that corporate governance mechanisms are associated with earnings management. These prior studies have however focused mainly on U.S. companies which have not much scope for earnings management. In Europe, companies have more opportunities to manage earnings, but as corporate governance is largely voluntary, corporate governance practices may be merely window-dressing or box-ticking, resulting in no or only a small effect on the amount of earnings management. The effectiveness of corporate governance practices may depend in this case on the amount of corporate governance recommendations issued at the country level and the level of legal investor protection. The more recommendations are issued, the better companies know what is expected, and the more effective corporate governance practices will be in constraining earnings management. Peasnell et al. (2005) finds for instance that after the introduction of the Cadbury Code in the U.K., the board of directors becomes a more effective tool in reducing earnings management than before the Cadbury Code. Durnev and Kim (2005) and Klapper and
2008 / 3
268
A. Renders & S. Vandenbogaerde
Love (2004) show that corporate governance practices have a larger impact on firm value in countries with a strong level of investor protection.
III.
Research design
Review of Business and Economics
A. Measure of earnings management
As a measure of financial statement quality, we use in this study the amount of earnings management. Accounting earnings have two major components: cash flows from operations and accounting adjustments called accruals. These accruals are simply the difference between earnings and cash flows. They include for instance depreciation and amortization, changes in non-cash working capital, provisions, write-offs and write-downs. Accruals are specifically vulnerable to manipulation because their determination depends on managers’ or accountants’ judgement and estimation. So if we can measure the accruals, we are indirectly measuring the amount of manipulation. However, not all accruals are a result of earnings manipulation. Some are related to the normal activity and growth of a firm. So to detect earnings management, total accruals are usually decomposed into two parts: non-discretionary accruals and discretionary (or abnormal) accruals. Discretionary accruals are those accruals that cannot be explained by the firm’s normal operating activity and are used as a proxy for the amount of earnings management. In the accounting literature, different models have been developed to estimate the amount of discretionary accruals (see, for example, Jones (1991), DeAngelo (1986) and (1994)). The most frequently used models are the Jones (1991) model and the modified-Jones model (Dechow et al. (1995)). In this study, we use the modified-Jones model as a means to estimate the amount of discretionary accruals. This model filters out the non-discretionary accruals using cross-sectional regressions of total accruals on the change in sales, adjusted for the change in accounts receivables, and on property, plant, and equipment and this within industries. We follow recent literature and adjust our discretionary accruals for performance effects by incorporating the return-on-assets ratio in the model (Kothari et al. (2005)). The error terms of the regressions are then seen as the discretionary accruals (DAs) and hence as our measure of earnings management. Positive DAs suggest income-increasing manipulations, while negative DAs suggest income-decreasing manipulations. As we are interested in manipulations in both directions, we use the absolute value of the discretionary accruals. However, we also split the sample by
Corporate Governance and Earnings Management: Evidence from Europe
269
TAijt = ijt + 1ijt(1/Assetsijt-1) + + 3ijtPPEijt + 4ijt ROAijt + ijt
2ijt(
REVijt – ARijt) (1)
B. The effect of corporate governance
The model we use to test the effect of corporate governance practices on earnings management is as follows: CGratingit + 2it CGI + 3it CGratingit*CGIt + + 5it LNMVit + 6it LEVit + 7it USSEit + 8it LISTINGit + 9it MTBVit + 10it CFOit + 11it AUDITOR + 12t time fixed effects + it 0it
+
1it
where t indexes years and i indexes firms.
4it
IP (2)
Review of Business and Economics
where TAijt is total accruals,6 REVijt is the one year change in net revenues, ARijt is the one year change in accounts receivable, and PPEijt is gross property, plant, and equipment. All variables are scaled by lagged total assets. In order to estimate the DAs, we first collect for all our sample countries all listed firms in the Worldscope database between 1999 and 2003. We delete firms with missing data. Our final sample consists of 33,677 firm-year observations divided over 14 countries, five years and 73 industries. Then we calculate industryyear-specific expected non-discretionary accruals using all firms with the same two-digit SIC code, using at least 25 observations in each industry-year combination. Finally, we calculate the discretionary accruals as the prediction errors from equation (1) (|DA1|). Besides estimating discretionary accruals using total accruals, we also estimate discretionary accruals by regressing working capital accruals on the change in sales and accounts receivable (|DA2|).7
|DA|it =
2008 / 3
the sign of the discretionary accruals to see whether we find different results for different directions of earnings management. A drawback of the modified-Jones model is that it requires sufficient industryyear-specific observations. Because countries outside the U.S. have smaller capital markets, we have to work with rather small samples and this may lead to inaccuracy in estimating the discretionary accruals. Therefore, as a robustness check, we perform the analyses using abnormal working capital accruals following DeFond and Park (2001). We estimate the expected accruals for each firm i in industry j at time t, using ordinary least squares. The model is:
Review of Business and Economics
2008 / 3
270
A. Renders & S. Vandenbogaerde
|DA| is the absolute value of the discretionary accruals estimated from the modified-Jones model, using either total accruals (|DA1|) or working capital accruals (|DA2|). The three main variables of interest are: CGrating, CGI and an interaction between these two variables (CGrating*CGI). CGrating is the aggregate corporate governance rating of a company or one of the sub-ratings (shareholders’ rights, takeover defences, or board structure and functioning). These ratings are developed by Deminor Rating. Since 1998, Deminor scores each year the companies belonging to the FTSEurofirst 300.8 The ratings are based on a grid consisting of more than 300 criteria. The maximum score on the ratings is 30. Each of the sub-ratings is rated on 10. For all sub-ratings, except takeover defences, better corporate governance leads to a higher score. With regard to takeover defences, it is assumed that fewer defences are beneficial for investors, and hence companies with less takeover defences will have a higher score. The variable CGI is a country score on the extent of corporate governance recommendations issued within a country using an extensive corporate governance index (see Appendix 1 for an overview of the index). This index is constructed from 50 principles gathered from the OECD Principles of Corporate Governance (1999). At the beginning of each sample year, we collect the corporate governance recommendations issued within our sample countries. We give a country a score of 1 when a principle of the index is included in the corporate governance recommendations. We sum up all the scores to arrive at a total score per country and per year. The maximum score obtainable on the index is 50. Consistent with prior studies, the following variables are included in the model to control for earnings management incentives. First, we control for the amount of legal investor protection (IP) offered by the country (La Porta et al. (1998)). Leuz et al. (2003) and Daske et al. (2003) provide evidence that earnings management practices are more prevalent in companies from code law countries compared to common law countries. Second, we control for the size of a company proxied by the natural logarithm of its market value (SIZE). Larger firms prefer upward earnings management (Young (1999), Gore et al. (2001)). Third, leverage is included in the model (LEV). Highly leveraged firms may have an incentive to increase earnings to avoid violating debt covenants (DeFond and Jiambalvo (1994)). Fourth, we include a dummy variable indicating whether a company is listed on a U.S. stock exchange or not (USSE). A foreign listing might result in additional regulatory requirements reducing the opportunities for earnings management. On the other hand, international capital market pressures might also give additional incentives to increase the level of earnings management. Besides this variable, we also include the number of stock exchanges a company is listed on (LISTING). Next, firms with high growth opportunities or rapidly growing firms have large discretionary accruals, so we control for this using the market-to-book-value (MTBV) (Kothari et al. (2005)). Consistent with Dechow et al. (1995) and Young (1999), we
Corporate Governance and Earnings Management: Evidence from Europe
271
IV.
2008 / 3
control for cash flow from operating activities deflated by total assets (CFO) because there is a well-documented inverse relation between operating cash flows and accruals. Finally, we include an indicator variable indicating whether the company is audited by a big-five auditor or not (AUDITOR).9 Evidence has been provided in the U.S. and the U.K. that big-five audit firms constitute a constraint on earnings management (e.g., Becker et al. (1998), Gore et al. (2001)). In contrast, Maijoor and Vanstraelen (2006) do not find a big-four auditor effect for a number of European countries. We also include time fixed effects and cluster the standard errors at the firm level to adjust for heteroskedasticity (Rogers (1993), Petersen (2005)).
Sample selection
Review of Business and Economics
We have corporate governance ratings for the FTSEurofirst 300 companies over five years, from 1999 till 2003. This time period covers the period before and after the big accounting scandals in the U.S. and Europe, such as Enron, Ahold, and Lernout and Hauspie. This time period also covers the period before the mandatory adoption of IFRS by European listed companies in 2005. IFRS are assumed to lead to higher accounting quality, as has been shown by various papers (e.g., Auer (1996), Niskanen et al. (2000), Ashbaugh and Pincus (2001)). Hence, the scope for earnings management has been reduced since 2005 (Barth et al. (2006), Christensen et al. (2007)). Consistent with prior research, we exclude financial institutions. After deleting companies with missing data, our sample consists of 655 firm-year observations divided over 14 countries. In most cases, information was missing on cash flows from operations. In order to reduce the impact of outliers, we winsorize all variables on the 1 and 99%. Table 1 presents summary statistics for our sample companies for the variables included in model (2). We see that there is a wide variation in the amount of earnings management. The average corporate governance rating is rather low, about 15 compared to the maximum score of 30. So there is still room for improvement for European firms. Both the extent of corporate governance recommendations and the level of legal investor protection are also quite low and the sample countries exhibit a wide variation. On average the sample companies are listed on two stock exchanges, and only 8% of the European companies are listed on a U.S. stock exchange. Next, our sample consists of large companies with a good financial structure and strong performance. Finally, almost 89% of the companies are audited by a big-five auditor.
0.120
0.176
0.055
0.0004
0.893
Mean
Stdev
Median
Min
Max
|DA1|
0.924
0.0003
0.052
0.168
0.113
|DA2|
Table 1. Descriptive statistics.
26.980
4.970
11.390
5.309
15.473
CGrating
44.000
0.000
29.250
8.999
29.499
CGI
5.000
0.000
3.000
1.523
3.077
IP
Review of Business and Economics
20.791
8.635
16.090
1.205
16.292
LNMV
5.074
0.000
0.774
1.037
1.092
LEV
1.000
0.000
0.000
0.271
0.080
USSE
10.000
1.000
1.000
1.981
2.173
LISTING
26.325
0.538
2.504
4.527
4.068
MTBV
0.674
-0.042
0.100
0.082
0.114
OCF
1.000
0.000
1.000
0.316
0.888
AUDITOR
2008 / 3 272 A. Renders & S. Vandenbogaerde
Corporate Governance and Earnings Management: Evidence from Europe
273
2008 / 3
Table 2 presents descriptive statistics per year. The table shows that the amount of earnings management decreases steadily over the years for both measures of discretionary accruals. The decrease in the level of earnings management over time may be explained by the improvement in corporate governance practices over the years. Correspondingly, we see that more and better corporate governance recommendations are developed over time. Finally, this table shows that the market-tobook value of the companies decreases drastically from 1999 (6.058) to 2002 (2.771) and companies switched more and more to a big-five auditor over the years.
Table 2. Descriptive statistics per year.
|DA1| |DA2| CGrating CGI IP LNMV
USSE LISTING MTBV OCF AUDITOR N
2000
0.195 (0.247) 0.202 (0.261) 14.298 (5.390) 23.880 (11.399) 3.102 (1.506) 16.489 (1.160) 1.114 (0.990) 0.088 (0.284) 2.226 (2.022) 6.058 (6.397) 0.126 (0.109) 0.876 (0.331) 119
0.134 (0.129) 0.156 (0.143) 15.139 (5.263) 22.891 (4.364) 3.132 (1.547) 16.540 (1.155) 1.052 (1.035) 0.083 (0.277) 2.118 (2.002) 5.164 (5.248) 0.121 (0.090) 0.861 (0.347) 125
2001 0.117 (0.164) 0.129 (0.144) 15.503 (5.405) 29.034 (6.177) 3.039 (1.551) 16.297 (1.219) 1.081 (0.989) 0.085 (0.280) 2.157 (1.950) 3.642 (3.786) 0.107 (0.076)) 0.898 (0.315) 135
2002
2003
0.080 (0.112) 0.081 (0.125) 15.785 (5.389) 33.470 (6.409) 3.090 (1.514) 15.986 (1.218) 1.074 (1.038) 0.079 (0.255) 2.160 (1.974) 2.771 (2.628) 0.105 (0.066) 0.903 (0.297) 134
0.097 (0.184) 0.069 (0.106) 16.741 (4.887) 37.581 (5.234) 3.026 (1.514) 16.173 (1.199) 1.140 (1.131) 0.073 (0.261) 2.205 (1.988) 2.933 (2.409) 0.115 (0.065) 0.907 (0.291) 142
Means (stdev) are reported.
Table 3 shows the average level of earnings management and company corporate governance ratings per country and per legal origin. It shows that, on average, companies from code law countries have higher levels of earnings management and lower corporate governance ratings compared to companies from common law
Review of Business and Economics
LEV
1999
2008 / 3
274
A. Renders & S. Vandenbogaerde
countries, confirming prior findings. These differences are all statistically significant at less than the 10% level.
Table 3. Average level of earnings management and company corporate governance ratings per country and legal origin.
Belgium France Greece
|DA1|
|DA2|
CGrating
16
0.045
0.095
12.321
127
0.112
0.092
15.299
7
0.136
0.213
11.583
Italy
31
0.127
0.256
12.471
The Netherlands
57
0.134
0.119
10.808
8
0.227
0.180
7.448
Portugal Spain
29
0.113
0.102
11.689
275
0.140
0.151
10.599
3
0.355
0.127
11.780
88
0.128
0.121
13.647
German Code law
91
0.125
0.122
13.585
Denmark
19
0.130
0.171
10.521
Finland
18
0.050
0.065
16.938
Sweden
54
0.136
0.078
14.139
French Code law Austria Germany
Review of Business and Economics
N
Scandinavian Code law
91
0.142
0.105
13.866
457
0.137
0.118
13.451
11
0.056
0.070
21.187
UK
187
0.078
0.049
19.904
Common law
198
Code law Ireland
t-test on the differences between code and common law countries
0.122
0.103
19.976
p = 0.065
p = 0.029
p < 0.001
Table 4 shows the correlations between the regression variables. We find numerous significant correlations, however, for the regressions, we calculated the variance of inflation factors (VIF) and these are all below two. When we look at some specific correlations, we see first of all that the earnings management measures are highly correlated with each other. Next, the table shows that corporate governance practices are negatively related with the earnings management measures and the aggregate and sub-ratings of corporate governance are positively related with each other. With regard to the extent of corporate governance recommendations, we find that it is negatively related with earnings management and positively with the corporate
-0.036
-0.099***
-0.110***
-0.066*
-0.010
0.213***
0.091**
0.016
-0.058
0.199***
0.062
0.006
SHAR
TO
BOARD
CGI
IP
LNMV
LEV
USSE
LISTING
MTBV
CFO
AUDITOR
,
,
0.010
0.126
0.190***
-0.038
-0.020
0.102***
0.169***
-0.096**
-0.159***
-0.163***
-0.141***
-0.039
-0.163***
1.000
|DA2|
0.013
-0.030
0.026
-0.030
0.114***
-0.037
-0.227***
0.546***
0.378***
0.738***
0.883***
0.550***
1.000
CGrating
0.004
0.087**
0.009
-0.071*
0.058
-0.101***
-0.112***
0.098***
0.295***
0.165***
0.434***
1.000
SHAR
0.004
-0.052
0.065*
0.073**
0.145***
-0.048
-0.122***
0.430***
0.154***
0.438***
1.000
TO
0.009
-0.029
-0.016
-0.211***
0.006
0.017
-0.298***
0.649***
0.444***
1.000
BOARD
Review of Business and Economics
, indicates significance at the 10%, 5% and 1%, respectively.
-0.130***
CGrating
* ** ***
0.656***
|DA1|
|DA2|
1.000 -0.270*** 0.025 0.068* -0.318*** 0.121*** -0.031 0.044
0.271***
0.021 0.022 -0.088** -0.140*** -0.072* -0.008
IP
-0.259***
1.000
CGI
0.089**
0.168***
0.196***
0.333***
0.112***
-0.099***
1.000
LNMV
-0.063*
-0.196***
0.074**
-0.001
-0.018
1.000
LEV
0.105***
0.018
-0.005
0.384***
1.000
USSE
0.110***
-0.113***
-0.124**
1.000
LISTING
-0.29
0.412
1.000
MTBV
-0.017
1.000
CFO
1.000
AUDITOR
2008 / 3
Table 4. Pearson correlations.
Corporate Governance and Earnings Management: Evidence from Europe
275
2008 / 3
276
A. Renders & S. Vandenbogaerde
governance ratings. The same holds for the level of legal investor protection. The table also shows that leverage is positively correlated with the level of earnings management. This indicates that as the amount of debt financing increases, companies will manage earnings more in order to avoid violating debt covenants. This indicates that companies have other incentives, besides stock market pressure, to manage earnings. Another interesting fact we can see from this table is that when a company is listed on a U.S. stock exchange, this company will have better corporate governance.
V.
Multivariate analysis
Review of Business and Economics
1. Aggregate corporate governance rating Table 5 presents the regression results of earnings management on the aggregated corporate governance ratings. We first report the results of a basic model in which we only include the control variables. The R² of this basic model is about 12% for |DA1| and almost 16% for |DA2|. In both models, we find that size has a significant positive impact on earnings management. So larger companies appear to manage earnings more. Next, companies with a higher debt-to-equity ratio are more likely to manipulate earnings. In contrast, companies that are listed on multiple stock exchanges seem to have less incentive to manage earnings, although we do not find an effect for companies listed in the U.S. So the additional regulatory requirements associated with listing on multiple stock exchanges reduce the opportunities for earnings management. Finally, most of the time effects are significant. When we include our test variables, the R²s increase to 13% and 17%, respectively. We find first of all that corporate governance mechanisms installed in companies have a significant negative impact on earnings management. This indicates that even in countries with voluntary corporate governance, better corporate governance practices are associated with higher financial statement quality. Second, the results indicate that in countries with extensive corporate governance recommendations, corporate governance practices are more effective in constraining earnings management. We do not find, however, any direct impact of institutional characteristics, i.e. the extent of corporate governance recommendations or the degree of legal investor protection, on the level of earnings management. The results of the control variables are largely consistent with the basic model, with the exception that the size-effect has disappeared in the extended model.
Corporate Governance and Earnings Management: Evidence from Europe
277
Expected coefficient –
CGI
?
CGrating*CGI
–
IP
–
LNMV
+
LEV
+
USSE
?
LISTING
?
MTBV
+
CFO
–
AUDITOR
–
DUM99
?
DUM00
?
DUM01
?
DUM02
?
R²
|DA1|
-0.007 (-0.74) 0.024* (1.83) 0.021* (1.84) 0.047 (1.03) -0.015** (-2.19) 0.004 (1.47) -0.111 (-0.77) 0.027 (1.08) 0.069*** (3.42) 0.031* (1.77) 0.027** (2.10) -0.026*** (-2.35) 0.115
Full model
|DA2|
|DA1|
|DA2|
-0.016 (-1.43) 0.013* (1.71) 0.025*** (2.35) 0.022 (0.61) -0.010* (-1.75) 0.002 (0.57) -0.170 (-1.34) 0.034 (1.36) 0.132*** (3.71) 0.029** (2.14) 0.042*** (3.42) 0.013 (1.38) 0.159
-0.011*** (-2.44) -0.004 (-1.36) -0.002* (-1.77) -0.004 (-0.04) 0.021 (1.58) 0.023** (2.25) 0.046 (1.08) -0.012* (-1.95) 0.004 (1.47) -0.083 (-0.56) 0.027 (1.09) 0.063** (2.29) 0.041* (1.90) 0.022** (2.01) -0.021* (-1.74) 0.132
-0.009* (-1.88) -0.004 (-1.18) -0.002* (-1.70) -0.013 (-1.16) 0.012 (1.03) 0.024** (2.14) 0.022 (0.65) -0.009* (-1.81) 0.001 (0.53) -0.188 (-1.44) 0.033 (1.33) 0.134*** (3.91) 0.049** (2.92) 0.050*** (3.44) 0.020* (1.83) 0.169
*
, **, ***, indicates significance at the 10%, 5%, and 1%, respectively. Coefficients (t-stat) are reported. All standard errors are heteroskedasticity robust, clustered at the firm level.
2. Sub-ratings Table 6 presents the regression results for each of the sub-ratings separately. We only report the results for |DA1|; the results for |DA2| are similar. For brevity reasons, we also do not report the time effects. We find first that as the amount of shareholder rights increases, the level of earnings management decreases (t = 1.71). Second, we find that takeover defences are also negatively associated with the level of earnings management (t = -2.85). This indicates that when companies
Review of Business and Economics
Cgrating
Basic model
2008 / 3
Table 5. Regression results for earnings management on aggregate corporate governance ratings.
2008 / 3
278
A. Renders & S. Vandenbogaerde
are protected from the market of corporate control through the use of takeover defences, insiders will place their own benefits before those of the outside shareholders and manage earnings more. This is consistent with the findings of Yu (2006) and Barber et al. (2007). Third, the results show that a good functioning board of directors and board committees result in lower earnings management (t = -2.00). Finally, the table shows that the extent of corporate governance recommendations affects the effectiveness of takeover defences and the board of directors and committees, but not of shareholder rights. Table 6. Regression results for earnings management on shareholder rights, takeover defences, and board structure and functioning. Coef. (t-stat) SHAR
Coef. (t-stat)
-0.027* (-1.71)
TO
-0.019*** (-2.85)
BOARD CGI Sub-rating*CGI
Review of Business and Economics
IP LNMV LEV USSE LISTING MTBV CFO AUDITOR R²
Coef. (t-stat)
-0.005 (-1.14) -0.001 (-1.32) -0.007 (-0.89) 0.024* (1.70) 0.024*** (2.40) 0.048 (1.06) -0.015** (-2.16) 0.004 (1.44) -0.100 (-0.69) 0.022 (0.90) 0.119
-0.001 (-0.83) -0.004** (-2.20) -0.000 (-0.12) 0.022 (1.63) 0.023** (2.26) 0.047 (1.11) -0.011* (-1.84) 0.004 (1.52) -0.091 (-0.61) 0.022 (0.83) 0.136
-0.027** (-2.00) -0.003 (-0.38) -0.003** (-1.96) -0.003 (-0.26) 0.022* (1.65) 0.025*** (2.51) 0.041 (0.93) -0.015** (-2.11) 0.004 (1.37) -0.090 (-0.63) 0.025 (1.01) 0.126
* ** ***
, , , indicates significance at the 10%, 5%, and 1%, respectively. All standard errors are heteroskedasticity robust, clustered at the firm level. Time effects are not reported.
With regard to the control variables, we find largely the same results as with the aggregated measure of corporate governance.
Corporate Governance and Earnings Management: Evidence from Europe
279
Table 7. Regression results for income-increasing and income-decreasing earnings management separately.
CGrating CGI CGrating*CGI IP LNMV LEV USSE LISTING MTBV CFO
R² N
DA > 0
Coef. (t-stat)
Coef. (t-stat)
-0.004 (-0.90) -0.002 (-0.41) -0.001 (-0.11) -0.010 (-0.95) 0.017 (1.18) 0.009 (0.69) 0.032 (0.65) -0.010 (-1.28) 0.002 (0.51) -0.326 (-1.56) -0.003 (-0.09) 0.093 356
-0.014*** (-2.37) -0.004 (-0.94) -0.003* (-1.89) -0.009 (-0.69) 0.030*** (2.41) 0.023** (2.20) 0.069 (1.03) -0.017** (-2.01) 0.004 (1.00) -0.080 (-0.46) -0.051* (-1.91) 0.269 299
* ** ***
, , , indicates significance at the 10%, 5%, and 1%, respectively. All standard errors are heteroskedasticity robust, clustered at the firm level. Time effects are not reported.
Table 7 shows the regression results when we split our sample into companies with income-increasing and companies with income-decreasing earnings management. Again we only report the results for |DA1|; the results for |DA2| are similar. The majority of our sample companies (54%) manage earnings downwards. We find that better corporate governance is associated with less income-increasing earnings management (t = -2.07), but not with income-decreasing earnings management (t = -0.90). The results also show that corporate governance recommendations only increase the effectiveness of corporate governance practices with regard to incomeincreasing earnings management (t = -1.89). Next, we find that larger companies with a higher leverage are more likely to manage earnings upwards (t = 2.20).
Review of Business and Economics
AUDITOR
DA < = 0
2008 / 3
3. Income-increasing and income-decreasing earnings management
2008 / 3
280
A. Renders & S. Vandenbogaerde
This is consistent with prior findings. The findings also show that companies that are listed on more stock exchanges will engage less in income-increasing earnings management (t = -2.01). Finally, the presence of a big-five auditor seems to constrain income-increasing (t = -1.91) but not income-decreasing earnings management (t = -0.09).
4. Code law and common law countries Finally, when we split our sample into code and common law countries (see Table 8) we find that our results are mainly driven by the code law countries. For instance, we find that better corporate governance is associated with less earnings management in code law countries (t = -2.04), but not in common law countries (t = -1.33). More corporate governance recommendations also lead to more effective corporate governance practices in code law countries (t = -1.83), but not in common law countries (t = -0.001).
Table 8. Regression results for code law and common law countries separately.
CGrating
Review of Business and Economics
CGI CGrating*CGI LNMV LEV USSE LISTING MTBV CFO AUDITOR R² N * ** ***
Code law
Common law
Coef. (t-stat)
Coef. (t-stat)
-0.014** (-2.04) -0.006 (-1.31) -0.004* (-1.83) 0.029 (1.29) 0.017 (2.55) 0.064 (1.04) -0.015* (-1.93) 0.002 (0.42) 0.026 (0.15) 0.024 (0.88)
-0.009 (-1.33) -0.002 (-0.31) -0.001 (-0.83) 0.008 (0.39) 0.011 (1.85) 0.024 (0.22) -0.012* (-1.74) 0.007 (0.80) -0.366 (-1.07) 0.050 (0.39)
0.148
0.143
457
198
, , , indicates significance at the 10%, 5%, and 1%, respectively. All standard errors are heteroskedasticity robust, clustered at the firm level. Time effects are not reported.
Corporate Governance and Earnings Management: Evidence from Europe
281
2008 / 3
One possible explanation is that we have too little observations for our common law countries and our tests have too little power. Another explanation might be that in our common law countries the amount of earnings management is lower and the level of corporate governance higher (see Table 3), and as a consequence, an improvement in corporate governance has no or only a very small marginal effect on the amount of earnings management. Whereas in code law countries, the amount of earnings management is higher and the governance of firms weaker, leading to an improvement in corporate governance having a strong effect on earnings management.
5. Robustness tests We use as an alternative earnings management to measure abnormal working capital accruals following DeFond and Park (2001). When using this alternative to measure of earnings management, we find results that are largely consistent with the results found above (results not reported).
VI.
Conclusion
Review of Business and Economics
In this paper, we study the relation between corporate governance practices and earnings management for a sample of European companies. Prior evidence on this relation in Europe is rather limited as most existing studies focus on U.S. companies. However, as the institutional setting in Europe is quite different from that in the U.S., prior results may be difficult to generalise. For instance, where the U.S. is characterised by strong investor protection laws and extended corporate governance regulation, European countries typically have weak laws protecting investors and tend to leave companies considerable options with regard to corporate governance. Furthermore, European companies may have different incentives to manage earnings than U.S. companies and have considerable accounting discretion. As a consequence, it is not clear ex ante whether the negative relation between corporate governance and earnings management found by prior studies will also hold in a European setting. In this study we address five research questions. First, are good corporate governance practices associated with less earnings management for European companies? Second, if so, which elements of corporate governance reduce the amount of earnings management? Third, does the impact of corporate governance practices on earnings management depend on the extent of corporate governance recommendations issued at the country level? Fourth, do European companies engage more in income-decreasing or income-increasing earnings management and does corporate governance constrain both types? And finally, does the relation between corporate
Review of Business and Economics
2008 / 3
282
A. Renders & S. Vandenbogaerde
governance practices and earnings management differ according to the legal origin of countries? As a measure of the amount of earnings management, we use performance-adjusted discretionary accruals estimated by employing the modified-Jones model (Kothari et al. (2005)). As our measure of corporate governance practices, we use the aggregate corporate governance score developed by Deminor Rating, and also the sub-ratings with regard to shareholder rights, takeover defences, and board structure and functioning. We also score countries on the extent of corporate governance recommendations issued at the country level. Our results show that better corporate governance is associated with less earnings management. So even when corporate governance is not regulated by law, it has an impact on accounting quality. These results also hold for the three sub-ratings. We find furthermore that in countries with more extensive corporate governance recommendations, corporate governance practices are more effective in constraining earnings management. When we split our sample into companies that manage earnings upwards and companies that manage earnings downwards, the results show that corporate governance has an impact on income-increasing earnings management, but not on income-decreasing earnings management. This is surprising, given that the majority of the sample companies engage in income-decreasing earnings management. Finally, the findings indicate that the negative relation between corporate governance and earnings management is particularly evident in code law countries and not in common law countries. There are of course a number of limitations associated with this study. First, we only consider earnings management as a measure of accounting quality. Second, our sample period is limited to the period before the mandatory adoption of International Financial Reporting Standards (IFRS) by European listed companies. An opportunity for future research would be to test whether the results still hold in the period after IFRS-adoption, i.e., when companies have less opportunities to manage earnings. A final limitation is that we do not examine neither control for the possible endogenous relation between earnings management and corporate governance.
Appendix 1. Structure of the Corporate Governance Index (CGIj) General principles I. II.
Legal basis and compliance (1) 1. Comply-or-explain principle Scope (1) 2. Encouraged to all companies
Yes/No Yes/No
Corporate Governance and Earnings Management: Evidence from Europe
283
III.
IV.
V.
Mission of the board (5) 3. Shareholder value maximization 4. Long-term viability of the company 5. Good relationship with stakeholders 6. Effective monitoring of management 7. Compliance with laws
Yes/No Yes/No Yes/No Yes/No Yes/No
Key functions of the board (6) 8. Guide corporate strategy 9. Monitor and replace key executives 10. Review remuneration 11. Manage potential conflicts of interest 12. Guard integrity of financial reporting 13. Increase effectiveness of governance practices
Yes/No Yes/No Yes/No Yes/No Yes/No Yes/No
Independence of the board (3) 14. Separation of chairman and CEO 15. Mix of inside and outside directors Non-executive directors Independent directors 16. Stock options not allowed as compensation
2008 / 3
Board of directors
Yes/No None/minimum number/majority None/minimum number/majority Yes/No
Board committees VI.
Recommended committees (3) 17. Appointment committee 18. Remuneration committee 19. Audit committee
VIII. Independence of the committees (6) 27. Appointment committee: non-executive directors 28. Appointment committee: independent directors 29. Remuneration committee: non-executive directors 30. Remuneration committee: independent directors
Yes/No Yes/No Yes/No Yes/No Yes/No Yes/No Yes/No
None/minimum number/majority None/minimum number/majority None/minimum number/majority None/minimum number/majority
Review of Business and Economics
VII. Key functions of the committees (7) 20. Appointment committee: propose appointment of directors 21. Remuneration committee: recommend remuneration for directors 22. Audit committee: report to the board 23. Audit committee: hear the company auditors 24. Audit committee: ensure appropriateness and consistency of accounting policies 25. Audit committee: verify accuracy of internal procedures 26. Audit committee: appoint auditor and determine audit fee
Yes/No Yes/No Yes/No
2008 / 3
284
A. Renders & S. Vandenbogaerde
31. Audit committee: non-executive directors 32. Audit committee: independent directors
None/minimum number/majority None/minimum number/majority
Shareholders IX. Shareholders' protection (4) 33. Equal treatment of shareholders 34. One share/one vote 35. No anti-take-over devices 36. Proxy voting allowed X.
General meeting (4) 37. Select new directors 38. Participate in decisions concerning fundamental changes 39. Decide on distribution of profits 40. Ask questions
Yes/No Yes/No Yes/No Yes/No Yes/No Yes/No Yes/No Yes/No
Review of Business and Economics
Disclosure XI. Quality (2) 41. Use high quality accounting standards 42. Audited by an independent auditor
Yes/No Yes/No
XII. Timing (1) 43. Timely disclosure of relevant information
Yes/No
XIII.Contents (7) 44. Financial situation 45. Performance 46. Ownership 47. Governance 48. Relevant interests of directors 49. Composition of the board 50. Remuneration of key executives
Yes/No Yes/No Yes/No Yes/No Yes/No Yes/No Yes/No
Total
NOTES 1. K.U.Leuven, Department Accounting, Finance en Insurance (AFI), Assistant Professor, Naamsestraat 69, 3000 Leuven,
[email protected], tel.: 016/32 67 68, fax: 016/32 66 83, research interests: IFRS, corporate governance, accounting quality. 2 K.U.Leuven, Department Accounting, Finance en Insurance (AFI), PhD student, Naamsestraat 69, 3000 Leuven,
[email protected], tel.: 016/32 69 34, fax: 016/32 66 83. 3 We thank Deminor Rating for providing the corporate governance ratings. 4. Healy and Wahlen (1999, p. 368) define earnings management as: “Earnings management occurs when managers use judgement in financial reporting and in structuring transactions to alter financial reports to either mislead some stakeholders about the underlying economic performance of the company or to influence contractual outcomes that depend on reported accounting numbers”.
Corporate Governance and Earnings Management: Evidence from Europe
285
2008 / 3
5. Prior studies have shown that companies comply to some extent with the corporate governance recommendations issued in their country, but they also find that there is a wide variation between companies within one country (Canyon and Mallin (1997), Weir and Laing (2000), Pellens et al. (2001)). Some countries, such as Belgium, do have laws with regard to corporate governance practices (e.g., the law on Corporate Governance of August 2, 2002). However, the impact of these corporate governance laws is very small compared to the impact of the Sarbanes-Oxley Act in the U.S. 6. Total accruals are calculated as the change in current assets – change in current liabilities – change in cash + change in current maturities of long-term debt – depreciation and amortization expense (Dechow et al. (1995)). 7. Working capital accruals are calculated as the change in noncash current assets – the change in current liabilities. 8. These are the 300 largest listed companies in Europe in terms of market capitalization (FTSEurofirst (2006)). 9. Until 2001 Arthur Andersen was still active in the market, therefore we use the term bigfive auditor.
REFERENCES
Review of Business and Economics
Abbott, L., Parker, S., and Peters, G., 2004, Audit Committee Characteristics and Restatements, Auditing: A Journal of Practice and Theory 23, 69-87. Agrawal, A., and Chadha, S., 2005, Corporate Governance and Accounting Scandals, Journal of Law and Economics 48, 371-406. Ashbaugh, H., and Pincus, M., 2001, Domestic Accounting Standards, International Accounting Standards, and the Predictability of Earnings, Journal of Accounting Research 39, 417-434. Auer, K., 1996, Capital Market Reactions to Accounting Earnings Announcements – Empirical Evidence on the Difference in Information Content of IAS-based Earnings and EC-Directives-based Earnings, European Accounting Review 5, 587-623. Baber, W., Kang, S., and Liang, L., 2007, Shareholder Rights, Corporate Governance, and Accounting Restatements, Working Paper, Georgetown University. Barth, M., Landsman, W., Lang, M., and Williams, C., 2006, Accounting Quality: International Accounting Standards and US GAAP, Working Paper, Stanford University. Beasley, M., 1996, An Empirical Analysis between the Board of Director Composition and Financial Statement Fraud, The Accounting Review 71, 443-466. Becht, M., and Roell, A., 1999, Blockholdings in Europe: An International Comparison, European Economic Review 43, 1049-1056. Becker, C., DeFond, M., Jiambalvo, J. and Subramanyam, K., 1998, The Effect of Audit Quality on Earnings Management, Contemporary Accounting Research 15, 1-24. Canyon M., and Mallin, C., 1997, A Review of Compliance with Cadbury, Journal of General Management 2, 24-37. Chen, F., 2007, Capital Market Pressure and Earnings Management: Evidence from Dualclass Firms, Working Paper, Columbia University. Christensen, H., Lee, E., and Walker, M., 2007, Incentives or Standards: What Determines Accounting Quality Changes around IFRS Adoption, Working Paper, Manchester Business School. Cohen, D., Dey, A., and Lys, T., 2007, Real and Accrual-based Earnings Management in the Pre- and Post-Sarbanes Oxley Periods, Working Paper, Northwestern University.
Review of Business and Economics
2008 / 3
286
A. Renders & S. Vandenbogaerde
Daske, H., Gebhardt, G. and McLeay, S., 2003, The Distribution of Earnings Relative to Targets: A Comparative Study across and within the European Union Countries, Working Paper, Frankfurt University. DeAngelo, L.E., 1986, Accounting Numbers as Market Valuation Substitutes: A Study of Management Buyouts of Public Stockholders, The Accounting Review 61, 400-420. DeAngelo, L.E., 1994, Accounting Choice in Troubled Companies, Journal of Accounting and Economics 17, 113-143. Dechow, P., Sloan, R., and Sweeney, A., 1995, Detecting Earnings Management, The Accounting Review 70, 193-226. Dechow, P., Sloan, R., and Sweeney, A., 1996, Causes and Consequences of Earnings Manipulation: An Analysis of Firms Subject to Enforcement Actions by the SEC, Contemporary Accounting Research 13, 1-36. DeFond, M., and Jiambalvo, J., 1994, Debt Covenant Violation and Manipulation of Accruals, Journal of Accounting and Economics 17, 145-176. DeFond, M., and Park, C., 2001, The Reversal of Abnormal Accruals and the Market Valuation of Earnings Surpises, The Accounting Review 76, 375-404. Durnev, A., and Kim, E., 2005, To Steal or Not to Steal: Firm Attributes, Legal Environment, and Valuation, Journal of Finance 60, 1461-1493. Dyck, A., and Zingales, L., 2004, Private Benefits of Control: An International Comparison, Journal of Finance 59, 537-600. FTSEurofirst, 2006, Ground Rules for the Management for the FTSEurofirst Indices, www.ftseurofirst.com. Francis, J., Maydew, E., and Sparks, H., 1999, The role of Big 6 Auditors in the Credible Reporting of Accruals, Auditing: A Journal of Practice and Theory 18, 17-34. Gore, P., Pope, P., and Singh, A., 2001, Non-audit Services, Auditor Independence, and Earnings Management, Working Paper, Lancaster University. Healy, P., and Wahlen, J., 1999, A Review of the Earnings Management Literature and its Implications for Standard Setting, Accounting Horizons December, 365-383. Jensen, M., and Meckling, W., 1976, Theory of the Firm: Managerial Behavior, Agency Costs, and Ownership Structure, Journal of Financial Economics October, 305-360. Jiraporn, P., 2005, An Empirical Analysis of Corporate Takeover Defences and Earnings Management: Evidence from the US, Applied Financial Economics, 293-303. Jones, J., 1991, Earnings Management During Import Relief Investigations, Journal of Accounting Research 29, 193-228. Klapper, L., and Love, I., 2004, Corporate Governance, Investor Protection, and Performance in Emerging Markets, Journal of Corporate Finance 10, 703-728 Klein, A., 2002, Audit Committee, Board of Director Characteristics, and Earnings Management, Journal of Accounting and Economics 33, 375-400. Kothari, S.P., Leone, A., and Wasley, C., 2005, Performance Matched Discretionary Accrual Measures, Journal of Accounting and Economics 39, 163-197. La Porta, R., Lopez-de-Silanes, F., and Shleifer, A., 1998, Law and Finance, Journal of Political Economy 106, 1113-1155. Leuz, C., Nanda, D., and Wysocki, P., 2003, Investor Protection and Earnings Management: An International Comparison, Journal of Financial Economics 69, 505-527. Leuz, C., and Verrecchia, R., 2000, The Economic Consequences of Increased Disclosure, Journal of Accounting Research 38, 91-124. Maijoor, S. and Vanstraelen, A., 2006, Earnings Management within Europe: The Effects of Member State Audit Environment, Audit Firm Quality and International Capital Markets, Accounting and Business Research 36, 33-52.
Corporate Governance and Earnings Management: Evidence from Europe
287
2008 / 3 Review of Business and Economics
Nenova, T., 2003, The Value of Corporate Voting Rights and Control: A Cross-Country Analysis, Journal of Financial Economics 68, 325-351. Niskanen, J., Kinnunen, J., and Kasanen, E., 2000, The Value Relevance of IAS Reconciliation Components: Empirical Evidence from Finland, Journal of Accounting and Public Policy 19, 119-137. Organisation for Economic Co-operation and Development (OECD), 1999, OECD Principles of Corporate Governance, Doc. no. SG/CG(99)5. Peasnell, K., Pope, P., and Young, S., 2005, Board Monitoring and Earnings Management: Do Outside Directors Influence Abnormal Accruals?, Journal of Business Finance and Accounting 32, 1311-1346. Pellens, B., Hillebrandt, F., and Ulmer, B., 2001, Implementation of Corporate Governance Codes in German Practice – An Empirical Analysis of the DAX 100 Companies, BetriebsBerater 56, 1243-1250. Petersen, M., 2005, Estimating Standard Errors in Finance Panel Data Sets: Comparing Approaches, Working paper, Northwestern University. Renders, A., and Gaeremynck, A., 2007, The Impact of Legal and Voluntary Investor Protection on the Early Adoption of IFRS, De Economist 155, 49-72. Richardson, S., Tuna, I., and Wu, M., 2002, Predicting Earnings Management: The Case of Earnings Restatements, Working Paper, University of Pennsylvania. Rogers, W., 1993, Regression Standard Errors in Clustered Samples, Stata Technical Bulletin 13, 19-23. Vander Bauwhede, H., and Willekens, M., 2003, Earnings Management in Belgium: A Review of the Empirical Evidence, Tijdschrift voor Economie en Management 68, 199-218. Weir, C. and Laing, D., 2000, The Performance-Governance Relationship: The Effects of Cadbury Compliance on UK Quoted Companies, Journal of Management and Governance 4, 265-281. Xie, B., Davidson, W., and Dadalt, P., 2003, Earnings Management and Corporate Governance: The Role of the Board and the Audit Committee, Journal of Corporate Finance 9, 295-316. Young, S., 1999, Systematic Measurement Error in the Estimation of Discretionary Accruals: An Evaluation of Alternative Modelling Procedures, Journal of Business, Finance and Accounting 26, 833-862. Yu, F., 2006, Corporate Governance and Earnings Management, Working Paper, University of Minnesota. Zhou, J., and Lobo, G., 2001, Disclosure Quality and Earnings Management, Asia-Pacific Journal of Accounting and Economics 8, 1-20.
2008 / 3
The Business Failure Process: An Integrative Model of the Literature 1
N. Crutzen & D. Van Caillie
2
ABSTRACT
Review of Business and Economics
Considering the fragmented literature related to business failure prevention and considering the need for a unifying conceptual framework on how and why companies fail, the current paper proposes an integration of the multiple similar, complementary or supplementary findings of previous preventive contributions, i.e. the studies focused on the identification of failure factors and of failure paths. An in-depth analysis of the existing literature and of its limitations leads us to the elaboration of an original integrative model on how and why companies fail. It is founded on a resource-based view of the firm (Wernerfelt, 1984) and it represents, in a deterministic perspective, the different general phases which a failing firm goes through until it eventually exits the failure process. JEL CODES
KEY WORDS
RESEARCH FUNDING
I.
G33, M13, M41 Failure, causes and symptoms of failure, dynamics of failure, failure path(s) This research was financed by a doctoral grant “non-Fria” received from the University of Liège
Introduction
Business failure3 has been one of the most investigated topics in the business literature during the last seven decades (Balcaen and Ooghe, 2006a). The interest for the analysis and for the understanding of business failure can notably be explained by the fact that a large number of stakeholders are concerned with the firm’s activity and with its evolution over time. Its customers, its suppliers, its creditors, its workers are all interested in the potential future of the firm (Weitzel and Jonsson, 1989;
The Business Failure Process: An Integrative Model of the Literature
289
The present paper integrates, in an interpretative perspective (Thiétart, 2003), the findings of previous studies in order to propose a integrative model of the business failure process explaining why and how companies fail. It is based on an in-depth analysis of the existing literature and of its limitations and it represents the different general phases which a company goes through in its evolution towards a potential bankruptcy. More precisely, this model presents for each chronological phase a series of failure factors (causes, symptoms or warning signals) that can be presented by failing firms. Furthermore, the failure process can be very different from one firm to another, i.e. different sequences of failure factors are possible (Ar-
Review of Business and Economics
Since the early 1930s and the initial contribution of Fitz Patrick (1932), a specific research field focused on the study of business failure has emerged and the number of research papers concerning the phenomenon has considerably increased over time (Guilhot, 2000). A large part of these contributions have been elaborated into a predictive perspective (Van Caillie, 2004). The business failure prediction literature, also called forecasting literature (Shumway, 2001) or literature on default and credit risk modelling (Carling et al., 2007), develops or analyses statistical models that tend to determine each firm’s bankruptcy risk as accurately as possible (Beaver, 1966; Altman, 1968; Shumway, 2001; Carling et al., 2007). Nevertheless, predictive studies have a narrow objective, i.e. to propose a classification of firms in a default and credit risk management perspective without giving any information about the potential causes or process of their failure, and strong limitations (Keasey and Watson, 1991; Daubie and Meskens, 2001; Van Caillie, 2004; Balcaen and Ooghe, 2006a). Consequently, the present paper only concentrates on business failure research that has been elaborated into a second perspective: the business failure prevention literature. Fundamentally, the business failure prevention literature intends to explain why (causes) and how (dynamic process) companies fail. It focuses on the ability to avoid serious problems and not only on their simple diagnosis (Daubie and Meskens, 2001). Yet, a previous examination of research written in a preventive perspective shows that the literature on business failure prevention has currently important limitations. In particular the literature concerning business failure prevention is highly fragmented, due to the fact that most authors have constructed their results through inductive reasoning, with little (or even no) reference to previous theories and research. As a result, researchers propose similar, complementary or supplementary results which have still not been integrated. So, as already stated by Dimitras et al. in 1996, to date, a convergence is needed to build a true integrative framework explaining how firms may fail, especially if we want to make further progress in the study of business failure from a preventive perspective.
2008 / 3
Daubie and Meskens, 2001). Therefore, its possible disappearance may lead both to dramatic macroeconomic and disastrous microeconomic consequences for the multiple partners of the firm (Charest et al., 1990).
2008 / 3
290
N. Crutzen & D. Van Caillie
genti, 1976), and thus the general model can be used as a basis to identify specific failure paths in the future. The paper is organised in two parts. In the first part, an analysis of the literature on business failure prevention is proposed. In fact, the main characteristics and the limits of this literature are highlighted and the need for an integrative model is emphasised. In the second part, an original integrative model explaining how and why firms do fail is proposed on the basis of the conclusions made in the first part of the paper.
II.
The literature on business failure prevention: an analysis
Review of Business and Economics
A. Introduction
Two main chronological perspectives concerning the phenomenon of business failure can be distinguished: the predictive and the preventive perspectives. Nevertheless, considering the particular purpose of business failure prediction studies, i.e. to elaborate statistical models in order to classify firms in a credit-risk management perspective, and considering the limitations of this perspective to failure (Keasey and Watson, 1991; Daubie and Meskens, 2001; Crutzen and Van Caillie, 2007a), up to now, we have chosen to exclude these studies from our scope of analysis. Therefore, this paper concentrates exclusively on the literature dealing with business failure prevention. Since the mid 1970s, a specific research field focused on the prevention of business failure has emerged. More specifically, instead of trying to discriminate between failed and non-failed companies (prediction of failure), a series of researchers have tried to understand and to explain in-depth why (i.e. causes) and how (i.e. process(es)) companies fail (Dimitras et al., 1996) in order to avoid a succession of serious problems. Formally these preventive studies concentrate on two elements. On the one hand, they try to understand and go back to the origins of the failure because only corrective actions that would solve the fundamental causes of the difficulties would really lead to a lasting recovery of the firm (Argenti, 1976). On the other hand, in line with the literature, which suggests that the concepts of failure “process” and failure “path” are extremely important in the study of business failure (Luoma and Laitinen, 1991), they take the dynamics of the failure process into account, i.e. how constitutive factors (causes, consequences or symptoms of business failure) combine over time.
The Business Failure Process: An Integrative Model of the Literature
291
B. The literature on business failure prevention: main characteristics
2008 / 3
Business failure prevention is thus interested in the identification and in the analysis of the causes, processes and remedies allowing to prevent failure or bankruptcy (Daubie and Meskens, 2001).
In this section, the main characteristics of studies on business failure prevention are analysed with reference to five criteria: the dynamics of their results, the failure stages they take into consideration, the organizational or financial approach they favour, their research object and their methodology. 1. The dynamics of the results proposed: static versus dynamic contributions
Review of Business and Economics
As Van Caillie (2004) states, two kinds of failure prevention studies can be distinguished: the static and the dynamic ones. In the first category, researchers such as Larson and Clute (1979), Cochran (1981), Wichman (1983), Koenig (1985), Jaminon (1986), Brüderl and Schüssler (1990), Sullivan et al. (1999) and Turner (2005) make an inventory of failure factors, i.e. causes and symptoms reflecting business failure, or analyse the impact of some variables (age or size) on the failure of firms. Thus, they propose static results. For example, Koenig (1985) classifies failure factors into some classical categories (mostly organizational, structural, environmental and financial factors) and presents them as independent elements without any chronological consideration: the relations between factors are not clearly pointed out and one or several possible combinations of the diverse elements are not proposed (Argenti, 1976). In the second category, researchers like Argenti (1976), Hambrick and D’Aveni (1988) or Laitinen (1991), Levinthal (2001), Mellahi and Wilkinson (2004), Ooghe and Waeyaert (2004) and Mellahi (2005) go further in the analysis of business factors and propose one or several dynamic models that show how events combine over time. These models propose different chronological phases through which companies evolve when they are failing. Amongst them, two subcategories of dynamic studies can be distinguished. Firstly, researchers like Sharma and Mahajan (1980), Ooghe and Van Wymeersch (1986), Sutton (1987), Hambrick and D’Aveni (1988), Weitzel and Jonsson (1989), Marco (1989), Sheppard (1994) or Mellahi and Wilkinson (2004) represent the failure process in a very general way: they point out some possible business failure factors and some important steps through which all failing companies evolve but they do not present different specific combinations of these factors. Alternatively, a few researchers like Argenti (1976), Laitinen (1991), Van Wymeersch and Wolfs (1996) or Ooghe and De Prijcker (2006) propose a description of some different failure paths after having demonstrated the evidence of multiple and diverse failure paths.
2008 / 3
292
N. Crutzen & D. Van Caillie
2. The failure stage(s) taken into consideration: early stage(s) versus last stage(s)
The literature frequently proposes a distinction between the early stage(s) and the last stage(s) of the failure process (Hambrick and D’Aveni, 1988; Weitzel and Jonsson, 1989). The early stages refer to the period before any sign of distress, namely before the appearance of any symptom that would be observable externally, while the last stages refer to the period between the appearance of observable failure symptoms and the disappearance of the firm. In the literature, these different phases are considered either globally, i.e. study of the whole failure process (Argenti, 1976; Larson and Clute, 1979; Levinthal, 1991; Mellahi, 2005), or sequentially, i.e. focus on the early stages (Carroll and Delacroix, 1982; Miller and Friesen, 1983; Singh et al., 1986a/b; Fichman and Levinthal, 1991; Sheth, 2005) or on the last stages of the process (Ooghe and Van Wymeersch, 1986; Sutton, 1987; Laitinen, 1991; Laitinen, 1992; Van Wymeersch and Wolfs, 1996).
3. The privileged approach: an organizational versus a financial approach of
Review of Business and Economics
failure The literature on business failure prevention favours almost systematically either an organizational or a financial approach of the business failure process and rarely combines these two complementary views. The privileged approach is strongly linked to the failure stage(s) studied by the researcher. The early stages of the failure process are characterized by a deterioration of the organisation and structure of the failing firm, which is not externally observable: so researchers who have focused on these stages have generally adopted an organizational approach towards the failure process (Carroll and Delacroix, 1982; Nystrom and Starbuck, 1984; O’Neill and Drucker, 1986; Bates and Nucci, 1988; Henderson, 1999; Sheth, 2005). In contrast, when a firm enters into the last stages of a failing process, the increasing internal deterioration of its organisation and structure gives rise to clear failure symptoms, which are mainly visible in the financial indicators. So researchers working on these last stages have usually selected a financial approach to study business failure (Ooghe and Van Wymeersch, 1986; Sutton, 1987; Laitinen, 1991; Laitinen, 1992; Van Wymeersch and Wolfs, 1996).
4. The research object
Researches proposed in the literature about business failure prevention have different research objects, even if the firm may be considered as the dominant research object. Indeed, most researchers focus their research on specific kinds of firms according to their age, size or industry.
The Business Failure Process: An Integrative Model of the Literature
293
From a methodological point of view, two main categories of research on business failure prevention can be discerned. Firstly, some researchers as Whetten (1980), Guilhot (2000) and Turner (2005) describe and summarize previous literature. In fact they propose a descriptive state of the art but do not provide any new contributions. Secondly, some researchers propose studies that provide new lessons for understanding the business failure process. Their studies are based either on deductive or inductive reasoning (Morris, 1997). Studies based on deductive reasoning derive hypotheses from previous literature, integrate them into a conceptual model and the validity of the deducted lessons is then generally empirically tested (Sharma and Mahajan, 1980; Miller and Friesen, 1983; Singh et al., 1986a/b; Cameron et al., 1987a/b; Sheppard, 1994). Simultaneously, studies based on inductive reasoning gather empirical or survey evidence with little or no reference to previous research. They rationalise inductively which phenomena may have led to bankruptcy (Argenti, 1976; Larson and Clute, 1979; Peterson et al., 1983; Nystrom and Starbuck, 1984; Sutton, 1987; Amburgey et al., 1993; Sullivan et al., 1999; Turner, 2005). Furthermore, when considering the empirical evidence proposed in a study, a clear distinction appears between qualitative data analysis (Argenti, 1976; Sutton, 1987; Ooghe and De Prijcker, 2006) and quantitative ones (Wichman, 1983; Cameron et al., 1987a/b; Carter et al., 1997; Thornhill and Amit, 2003).
C. The literature on business failure prevention: current limitations
Even if conducting a research on business failure in a preventive perspective allows to by-pass the main limitations related to the prediction perspective (Crutzen and Van Caillie, 2007a), to date, few authors have worked on causes, process or remedies for business failure (Daubie and Meskens, 2001; Van Caillie, 2004). Furthermore this literature has some considerable limitations. After an analysis of the main characteristics of the representative studies (see table 1) and according to the different statements made in the literature, five categories of limitations can be discerned: limits related to the dynamics of the results proposed, limits related to the failure stage(s) taken into consideration, limits related to the conceptual approach chosen, limits related to the research object and limits related to the methodology used.
Review of Business and Economics
Table 1 Summarizes the main characteristics of representative preventive studies.
2008 / 3
5. The methodology
–
–
Static
Dynamic (Failure process)
Static
Static
Static + Dynamic (Failure process)
Static
Static
Static
Static
Static
Static
Dynamic
Larson and Clute (1979)
Sharma and Mahajan (1980)
Whetten (1980)
Cochran (1981)
Malecot (1981)
Carroll and Delacroix (1982)
Miller and Friesen (1983)
Peterson and al. (1983)
Wichman (1983)
Marco (1984/5)
Nystrom and Starbuck (1984)
Newton (1985)
No info
–
–
–
–
No info
–
–
No info
–
Dynamic From 2 to 20 (Failure process years, depending + 3 specific paths) on the failure paths
Time horizon (dynamic)
Argenti (1976)
Type of studies – Static – Dynamic
Main approach – Organizational – Financial
Global but stress on the early stages
Early stages
Early stages
Early stages
Early stages
Early stages
Early stages
2. Last stages – Degradation of symptoms
1. Early stages – Entry paths
–
–
Global but little data about the exit phases
Organizational
Organizational
Organizational
Organizational
Organizational
Organizational
Organizational
Organizational
Organizational
Organizational
Organizational
Global – stress on the Organizational early stages
Global (link between Organizational causes and financial symptoms) – stress on the early stages
Studied phase of the failure process
Table 1. Main characteristics of business failure prevention studies.
Review of Business and Economics
Inductive reasoning
State of the art
State of the art
Deductive reasoning (adaptation of Argenti’s model)
Inductive reasoning
Inductive reasoning
Main methodology
–
–
No info
Small US firms
Small firms
Corporations older than 20 years
State of the art
Inductive reasoning
Inductive reasoning
Deductive reasoning
Inductive reasoning
Deductive reasoning
Argentinean and Deductive reasoning Irish firms in Newspaper
French firms
Small US firms
–
–
Small US firms
Large firms
Research object
–
Authors’ experiences + Case studies
Statistical technique
Statistical technique
Interviews and statistical technique
Statistical technique
Statistical technique
French national surveys
–
–
No information
Statistical technique
Case studies
Analysis technique
2008 / 3 294 N. Crutzen & D. Van Caillie
–
Static
Static
Dynamic (Failure process)
Static
Static
Static
Static
Dynamic
Static
Dynamic (Failure process)
Dynamic (Several patterns of decline)
Jaminon (1986)
O’Neill and Drucker (1986)
O and VW (1986)
Singh et al. (1986a)
Singh et al. (1986b)
Cameron et al. (1987a)
Cameron et al. (1987b)
Sutton (1987)
Bates and Nucci (1988)
Hambrick and D’Aveni (1988)
D’Aveni (1989)
Voluntary social Deductive reasoning services organizations located in Toronto Voluntary social Deductive reasoning services organizations located in Toronto US higher educa- Deductive reasoning tion organizations US higher educa- Deductive reasoning tion organizations Unambiguous orga- Inductive reasoning nizational deaths in USA US small firms
Organizational
Organizational
Organizational Organizational Organizational
Organizational
Organizational
Large US firms
–
Financial
Consequences of the decline (how events may combine when a firm is involved in a failure process
Small US firms
Organizational
Large US private firms
Belgian firms
Organizational
Statistical technique
Case studies
Statistical technique
Statistical technique
Statistical technique
Statistical technique
Statistical technique
Statistical technique
Case studies
–
Analysis technique
Deductive reasoning
Statistical technique
Inductive reasoning Statistical technique But the main constructs are chosen according to the literature
Inductive reasoning
Inductive reasoning
Deductive reasoning
State of the art + deductive reasoning
State of the art
–
Organizational
Main methodology
Research object
Main approach – Organizational – Financial
Identification of 4 Organizational phases but study of the last two phases (5 years prior bankruptcy)
Early stages
Last stages
Early stages
Early stages
Early stages
Early stages
Last stages
Early stages
Early stages
Early stages
Studied phase of the failure process
Review of Business and Economics
5 years
10 years
–
No info
–
–
–
–
No info
–
–
Static
Koenig (1985)
Time horizon (dynamic)
2008 / 3
Type of studies – Static – Dynamic
The Business Failure Process: An Integrative Model of the Literature
295
Static
Static
Dynamic (3 financial paths)
Dynamic
Dynamic (Failure process)
Static (link between – changes and failure)
Dynamic (Failure process)
Dynamic (financial failure paths)
Bruderl and Schussler (1990)
Fichman and Levinthal (1991)
Laitinen (1991)
Levinthal (1991)
Laitinen (1992)
Amburgey at al. (1993)
Sheppard (1994)
Van Wymeersch and Wolfs (1996)
No info
Static + Dynamic (Failure process)
Static
Liefhooghe (1997)
Henderson (1999)
–
–
Carter et al. (1997) Static
5 years
5 years
No info
–
4 years
–
–
No info
Dynamic (Failure process)
Marco (1989)
No info
Dynamic (Failure process)
Time horizon (dynamic)
Weitzel and Jonsson 1989
Type of studies – Static – Dynamic
Main approach – Organizational – Financial
Organizational
Organizational
Organizational
Early stages
Global – stress on early stages
Early stages
Last stages
Early stages
Early stages
Last stages
Global
Organizational
Organizational
Organizational
Financial
Organizational
Organizational
Financial
Organizational
Degradation of finan- Financial cial ratios – Last stages
Early stages
Early stages
Global but few details
Global: “from early Organizational stages of decline to later stages”
Studied phase of the failure process
Review of Business and Economics Analysis technique
Deductive reasoning
Deductive reasoning
Inductive reasoning
Deductive reasoning
Inductive reasoning
Inductive reasoning
State of the art US firms in personal Deductive reasoning computer industry
–
New firms in retail Deductive reasoning in USA
Large commercial and industrial firms
Large firms older than 5 years
Young Finnish firms
Argentinian and Deductive reasoning Irish firms in Newspaper
Statistical technique
–
Statistical technique
Statistical technique
Statistical technique
Statistical technique
Statistical technique
Statistical technique
Mainly small Finn- Inductive reasoning but Statistical technique ish firms the financial variables are chosen according to the literature
Statistical technique
Statistical technique
No complete published empirical validation (Van Caillie, 1992)
State of the art and inte- – gration of previous contributions
Main methodology
Voluntary social Deductive reasoning services organizations located in Toronto
German firms
No info
–
Research object
2008 / 3 296 N. Crutzen & D. Van Caillie
Static
Static
Static
Dynamic
Static (+ dynamic: link between causes)
Dynamic
Static
Static
Dynamic (4 failure paths)
Guilhot (2000)
Thornhill and Amit (2003)
Mellahi and Wilkinson (2004)
Ooghe and Waeyaert (2004)
Mellahi (2005)
Sheth (2005)
Turner (2005)
Ooghe and De Prijcker (2006)
No info
–
–
No info
No info
No info
–
–
–
Small US firms – – –
–
Large Australian firms in Insurance “Good” companies that failed – –
Organizational Organizational + financial Organizational Organizational
Organizational
Organizational Organizational Organizational Organizational
Early stages Review of different phases
Global
Early stages
Early stages
Global
Early stages
Early stages
Early stages
Research object
Main approach – Organizational – Financial
Studied phase of the failure process
Review of Business and Economics
Sullivan et al. (1999)
Time horizon (dynamic)
Statistical technique
–
Documents analysis and phone surveys
Analysis technique
Deductive reasoning (Test of Argenti’s hypotheses)
State of the art
Inductive reasoning
Inductive reasoning
Case studies
–
Document analyses, interviews, case studies
Case studies
State of the art and the- – oretical deduction of a model
State of the art and pro- – position of an integrative framework
Deductive reasoning
State of the art
Inductive reasoning
Main methodology
2008 / 3
Type of studies – Static – Dynamic
The Business Failure Process: An Integrative Model of the Literature
297
Review of Business and Economics
2008 / 3
298
N. Crutzen & D. Van Caillie
1. The limits related to the dynamics of the results proposed
Static studies (Peterson et al., 1983; Sheth, 2005) appear to be able to anticipate and to prevent failure less accurately than dynamic ones because they do not describe the sequence in which causes and symptoms of failure may follow each other and because they do not consequently forecast how events should evolve over time (Argenti, 1976). Furthermore, the causes of failure are often inventoried subjectively in static studies (Jaminon, 1986). These subjective and diverse presentations may mislead readers when they interpret the data (Sharma and Mahajan, 1980; Jaminon, 1986). Dynamic studies present important limitations too. Firstly, most researchers analyse the business failure process in general (Sharma and Mahajan, 1980; Hambrick and D’Aveni, 1988; Sheppard and Chowdhury, 2005): fundamentally, they stress important steps through which all failing firms should evolve but they do not present different specific combinations of factors whereas the literature demonstrates that diverse specific failure paths have to be distinguished (Argenti, 1976; Daubie and Meskens, 2001). Secondly dynamic studies do rarely take the time dimension into account: researchers generally do not insert the time dimension as an obvious variable in their research and, when a time scale is proposed, it differs from one research to another. For example, Argenti (1976) proposes a time horizon of 20 years to study business failure whereas Sheppard (1994) proposes a time horizon of 5 years.
2. The limits related to the failure stage(s) taken into consideration
Few researchers explicitly analyse all the different stages of the failure process (Weitzel and Jonsson, 1989). Most of them generally focus on one particular phase of the process: the early stages (Koenig, 1985; Cromie, 1991) or the last stages (Ooghe and Van Wymeersch, 1986; Laitinen, 1991; Balcaen and Ooghe, 2006b)). Furthermore, even if some authors present their research as global, few of them actually concentrate on the succession of the different stages of the failure process. For example, Argenti (1976), Sharma and Mahajan (1980), Newton (1985) and Marco (1989) concentrate more specifically on the early stages of the failure process whereas Malecot (1981) does not clearly and explicitly reconcile the early and last stages.
3. The limits related to the conceptual approach chosen
Even if the organizational and the financial approaches appear to be complementary for understanding and explaining completely the phenomenon of business failure, they have still not been really combined in the considered studies so far (Bilderbeek, 1973; Van Caillie and Dighaye, 2002). This third limitation is strongly linked with the previous one: because researchers generally concentrate either on the early stages or on the last stages of the failure process and because the concep-
The Business Failure Process: An Integrative Model of the Literature
299
4. The limits related to the research object
2008 / 3
tual approach they adopt generally depends on the failure stages taken into consideration in the study. It is methodologically logical that the organizational and financial approaches have still not been reconciled and are considered separately.
Most existing studies tend to focus on specific types of companies according to their age, size and/or industry. More specifically, Thornhill and Amit (2003) consider that most of the attention has been paid to early failures (namely failure of new and adolescent small firms) as well as to large, dramatic corporate collapses (Argenti, 1976; Hambrick and D’Aveni, 1988; D’Aveni, 1989) while many firms exit between these extreme positions. As a result, there is a considerable gap in the understanding of how and why less extreme types of firms fail (Thornhill and Amit, 2003). In addition, some authors as Singh et al. (1986a/b), Cameron et al. (1987a/b), Levinthal (1991), Van Wymeersch and Wolfs (1996), Carter et al. (1997), Henderson (1999) or Mellahi (2005) focus on specific industries: retail, newspaper, personal computer, insurance industries, etc.
5. The limits related to the methodology used
Review of Business and Economics
Business failure prevention models tend to be derived on an ad-hoc basis with little theoretical underpinning, whereas there is an important number of “normative” theories in social sciences (economy, finance, corporate strategy, contracting or agency theory) which may usefully contribute to a global and general understanding of corporate failure and of its potential causes (Morris, 1997). Studies on business failure prevention are generally constructed on the basis of inductive reasoning (against empirical or survey evidence) with little or no reference to any theoretical framework (Dimitras et al., 1996). So, as Morris argues, most researchers have chosen to gather empirical evidence and to rationalise inductively about the potential causes or process(es) of business failure (Peterson et al., 1983; Sullivan et al., 1999) while less contributions are based on deductive reasoning, namely a test of theoretical hypotheses (Sharma and Mahajan, 1980; D’Aveni, 1989; Thornhill and Amit, 2003). This inductive methodological trend notably explains why the business failure literature may be considered as highly fragmented. Indeed, most researchers propose their own results without really taking previous research into consideration: they choose to study some particular aspects or stages of the business failure process according to their experience or interests and their results are inferred from the analysis of empirical data they have collected on specific samples. So, disparate results are proposed in the literature and this is mainly caused by the fact that a large part of the authors have used inductive reasoning to elaborate their model with little or no reference to previous research (Morris, 1997).
2008 / 3
300
N. Crutzen & D. Van Caillie
To date, nobody has really suggested reconciling and integrating the fragmented results in order to propose a unifying conceptual framework. Nevertheless, as several researchers have pointed out (Dimitras et al., 1996), if we want to make further progress in the study of business failure (Weitzel and Jonsson, 1989), there is a need for convergence and thus for a unifying conceptual framework of the business failure process.
Review of Business and Economics
To sum up, an in-depth analysis of studies focusing on failure prevention (and their limits) leads to the finding that even if different authors have studied business failure in this perspective, their contributions are very different because the various authors have approached this dynamic phenomenon with little or no reference to a theoretical framework (inductive reasoning) and with complementary points of view (organizational or financial view, for example) which have still not been combined (Van Caillie and Dighaye, 2002). So, up to now, as Dimitras et al. argue, there seems to be no unifying theory in the literature (no unifying theory on how (description of the dynamic failure process(es)) and why (causes of business failure) companies fail while several authors emphasise the need of a complete understanding and explanation of business failure). Indeed, Dimitras et al. (1996) state that “a unifying theory of business failure has not been developed, in spite of few notable efforts” and Altman et al. (1994) underline that “we know many things about how companies fall into economic distress, about crisis processes and company decline, but we do not have a complete theory”. A deep examination of research on business failure prevention shows that, even if it is fragmented, a solid theoretical basis on why and how companies fail exists in the literature. Indeed the disparate contributions are in fact similar, complementary or supplementary: on the basis of inductive reasoning mainly, the different researchers have suggested elements that can be combined or they have differently presented analogous results. A solid theoretical basis on this thematic exists thus in the literature but, up to now, nobody has propounded to really integrate the different points of view and the underlying results. Considering this gap in the literature, we propose an integration of the fragmented literature in the next section.
III.
The business failure process: an integrative model
A. Theoretical framework
As mentioned above, the model integrates, in an interpretative perspective (Thiétart, 2003), the similar, complementary or supplementary findings of representative
The Business Failure Process: An Integrative Model of the Literature
301
In order to clearly organise all this diverse information, we have translated the findings of the previous preventive studies according to the principles of the resource-based view of the firm and we position thus our model of the business failure process within this perspective (Wernerfelt, 1984; Rumelt, 1991; Barney, 1991; Thornhill and Amit, 2003)
2008 / 3
studies from the literature on business failure prevention. It presents and combines the business failure factors that are the most often evoked in this literature. That is why it can be described as subjective and non exhaustive (Jaminon, 1986).
The resource-based view of the firm depicts firms as heterogeneous bundles of idiosyncratic, hard-to-imitate resources and capabilities (Wernerfelt, 1984; Barney, 1991). Amit and Schoenmaker (1993) define resources as “stocks of available factors that are owned or controlled by the firm” and capabilities as “informationbased, tangible or intangible processes that are firm-specific and are developed over time through complex interactions among the firm’s resources”. In addition, resources and capabilities encompass a broad array of assets, tangible and intangible, as well as numerous ways of deploying such assets and determine the firm competitive advantage (Barney, 1991). In the same sense, Teece et al. (1997) argue a firm’s competitive position derive from its managerial and organizational processes, shaped by its assets position, and the path available to it.
B. Integrative model on how and why companies fail
To be consistent with the principles of the resource-based view, the resources of the firm and the way they are deployed in order to reach the company’s goals are at the centre of our model. From the most fundamental failure factors, i.e. causes of the failure, to the most symptomatic ones, i.e. financial warning signals that appear some weeks before the potential bankruptcy of the firm, this integrative model presents, in a global and dynamic way, why (causes) and how (sequence of events) companies fail. It is
Review of Business and Economics
As Thornill and Amit (2003) state, even if the resource-based view has principally been used in the study of above-normal firm performance, it is interesting to apply its principles in the context of below-normal performance. According to this view of the firm, firm performance is a reflection of the firm’s resources, capabilities and their deployment to establish a viable competitive position. Thus, while superior performance is more likely when resources and capabilities are aligned with strategic industry factors, i.e. characteristics of the competitive environment that are determinants of firm-level profitability (Amit and Schoenmaker, 1993), business failure occurs when resources and capabilities are not (no longer) aligned with these factors.
Review of Business and Economics
2008 / 3
302
N. Crutzen & D. Van Caillie
concretely organised into four chronological phases: the origin of the failure (TA), the appearance and the deterioration of the failure symptoms (TS), the phase at which critical warning signals appear (TX) and finally, the bankruptcy of the failing
Table 2. Integrative model of the business failure process.
The Business Failure Process: An Integrative Model of the Literature
303
The integrative model represents the dynamics of the failure process in a deterministic perspective. More precisely, bankruptcy is presented as the unique possible exit of the process. Nevertheless, failing firms have other (positive or negative) potentialities to exit the failure process. On the one hand, firms may recover if the adequate remedies are taken. However, as problems become more and more obvious when firms enter into later stages of the process, it is more and more difficult to implement corrective actions that would lead to a durable recovery of the firm’s situation (Weitzel and Jonsson, 1989; Liefhooghe, 1997). On the other hand, bankruptcy is not the only possible negative exit of this process: firms may also disappear, i.e. lose their juridical identity, after a voluntary liquidation, a merger or a demerger. (I don’t like scission) These other negative exits do not necessarily take place at the fourth phase of the process
2008 / 3
firm (TZ). Specific failure factors, as well as their potential chronological combinations, have been associated with each phase: causes, symptoms and warning signals.
C. Articulation of the model 1. Introduction
More specifically, a firm enters into a failing process when its resource base and its deployment are not adequate to respond to internal and external pressures. These internal or external pressures are challenges to the continuation of the firm and come respectively from within the firm (internal organizational crises due to expansion, for example) or from outside (changes in customer behaviour, for example) (Weitzel and Jonsson, 1989). These inadequacies in the resource base and in its deployment prove to be an obstacle when dealing with internal and external pressures. The outcome is a misalignment between the firm (its resources and their deployment) and its environment. In short, the failure process starts when a firm’s set of resources and its deployment are inadequate and not adapted to the require-
Review of Business and Economics
An examination of the literature shows that a large number of researchers consider the business failure process to be related to a misalignment between, on the one hand, the resources of the firm and their deployment in order to reach the company’s goals AND, on the other hand, its environment (Tushman et al., 1986; Levinthal, 1991; Thornhill and Amit, 2003, Turner, 2005). In other words, failing firms do not succeed in “obtaining sustained competitive advantages by implementing strategies that exploit their internal strengths, through responding to environmental opportunities, while neutralizing external threats and avoiding internal weaknesses” (Barney, 1991).
Review of Business and Economics
2008 / 3
304
N. Crutzen & D. Van Caillie
ments of its environment. Under these circumstances, the firm does not succeed in creating or in sustaining a valuable strategic position (Barney, 1991; Thornill and Amit, 2003). In this sense, Hambrick and D’Aveni (1988) argue that four elements may have an impact on the entrance of the firm into a failure process: its capacity to anticipate and to respond to the requirements and changes of its environment, its domain initiative, its resources and its performance (that is related to the way its resources are deployed). Before explaining how each phase of the model is articulated, it is important to clarify the concepts of “external environment” and of “resources of the firm”. The external environment of the firm can be divided into two groups: its macroeconomic and its competitive environments. On the one hand, the macroeconomic environment is also commonly called the “PESTEL” environment: the Political, Economic, Social, Technological, Environmental and Legal environment of the firm (Johnson et al., 2005, p. 75-78). On the other hand, the competitive environment of the firm can be depicted according to Porter’s five forces (1980): the pressures exerted by incumbent rivals, new entrants, buyers, suppliers or substitutes on the firm. Both groups of external factors may have an impact on the firm: technological changes, economic downturn, political pressures, changes in the need of buyers or pressures of competitors are examples of potential external failure factors. If the firm is not aware of the current requirement of its external environment and if it does not anticipate its potential future changes, the risk of a firm entering into a failure process is higher. As no generally accepted classification of firm resources has yet emerged, we have decided to classify the firm’s resources according to the distinctions made by De Wit and Meyer (2004). The resource base of the firm consists of two main categories of resources: tangible and intangible resources. On the one hand, tangible resources are resources that can be physically observed. Technical, i.e. buildings, machines or materials, and financial resources are the main types of tangible resources identified. On the other hand, intangible resources are resources that cannot be touched. Among intangible resources, relational resources and competences can be distinguished. Firstly, relational resources are all the means available to the firm derived from the firm’s interaction with its environment. Secondly, competences refer to the firm’s fitness to perform in a particular field. A distinction between technical and managerial competences is judicious. The technical competences are the skills, knowledge or abilities related to a technical activity as industry-specific competences or activity-specific competences (Van Beirendonck, 2004).
The Business Failure Process: An Integrative Model of the Literature
305
2008 / 3
The managerial competences are the skills, knowledge or abilities to manage a business. Different managerial competences are evoked in the literature: the operational, financial, commercial and strategic competences (Kaplan and Norton, 1992; Liefhooghe, 1997); the ability to recognize pertinent information, to anticipate events and to adapt the firm to internal or external pressures (Argenti, 1976; Weitzel and Jonsson, 1989); the ability to control and to delegate; the ability to organize and to coordinate elements; the ability to deal with human beings (leadership, communication skills, team management) (Liefhooghe, 1997; Ooghe and Waeyaert, 2004), etc.
2. Four chronological failure phases
a. Phase 1 – Origin of the failure (TA) a. Deficiencies in the resource base of the firm are at the origin of its failure. They limit its ability to respond adequately to internal or external pressures.
Deficiencies may be observed within the different categories of resources that are generally present in a company. An old-fashioned (Brilman, 1985) or insufficiently specialized machinery (Van Caillie et al., 2006), insufficient financial capital (Liefhooghe, 1997), managerial competences restricted to an area of management such as marketing or finance for instance (Argenti, 1976), a lack of technical competences (Ooghe and Waeyaert, 2004), the lack of ability to anticipate events and the lack of flexibility (Tushman et al., 1986; Hambrick and D’Aveni, 1988, Thornhill and Amit, 2003) or insufficient external relationships (Sheppard, 1994) are examples4 of deficiencies in the resources that are frequently presented as potential failure causes in the literature. Two remarks are necessary. Firstly, the quality and the availability of the firm’s resources depend on three factors: intra-individual factors such as the characteristics of the individuals who are involved in the firm (owners/managers and workers), inter-individual factors such as the atmosphere within the firm and external factors such as pressures from its macroeconomic and competitive environment.
Review of Business and Economics
Two hypotheses have to be considered (Thornhill and Amit, 2003). The firm may be engaged in a failure process since its creation (T0 = TA). In this case, deficiencies in its resource base are present since its origin and the firm’s situation deteriorates over time if no corrective actions are taken. Alternatively, deficiencies in the resources of the firm may occur at different stages of its lifecycle. Once again, these deficiencies may be at the origin of the increasing deterioration of its situation if no corrective actions are taken.
Review of Business and Economics
2008 / 3
306
N. Crutzen & D. Van Caillie
With reference to this model, stakeholders are perceived as members of the external environment of the firm who place a part of their own resources at the firm’s disposal in prospect of future returns (money, goods, services, etc.). The deficiencies in a company’s resources can thus be linked to the profiles of the individuals who are involved in it (their experiences, their business backgrounds or expertise, their personalities, their motivations), inter-individual factors, its external environment (scarcity of some raw materials, regulations, interest rate changes) (Hambrick and D’Aveni, 1988; Sheppard, 1994; Ooghe and Waeyaert, 2004; Van Caillie et al., 2006). Secondly, business failure is generally not only caused by one deficiency in the resources of the firm (Newton, 1985; Ooghe and Waeyaert, 2004). In the main, it is the result of a series of inadequacies in the firm’s resources and in their deployment. b. The set of heterogeneous resources is deployed to reach the firm’s goals (Wernerfelt, 1984, Barney, 1991, Thornhill and Amit, 2003; Van Caillie et al., 2006). That is what we have called the resource management or deployment of the firm. It is the way the resources are organized to enable the firm to operate in its environment. In order to organize clearly the dimension “Resource management – deployment” we have chosen to present the elements according to the four perspectives of the Balanced Score Card model (BSC) proposed by Kaplan and Norton (1992). According to them, the performance of a firm depends on four dimensions that should be taken into account: the innovation and learning perspective, the customer perspective, the production perspective and the financial perspective.5 The key assumption explaining this model (Kaplan and Norton, 1992) is then that the performance of any firm derives from a double equilibrium: a balance between a long-term orientation (focused on innovation and learning as the main leverages to ensure business continuity in the long run) and shortterm orientation (focused on a financial vision of the firm, especially from a shareholder point of view) and a balance between a strategic inside-out perspective (focused on a permanent search for efficiency in the daily operations organized in a “Value-Chain” perspective) and a strategic outside-in perspective (focused on a permanent search for proximity and satisfaction of customers) (De Wit and Meyer, 2004). A firm will thus be well managed (and its resources will consequently be well deployed) if there is an equilibrium between short and long term concerns, financial and non-financial concerns as well as internal and external concerns. Therefore, the deployment of the resources of a firm should take into account the firm’s internal and external requirements and the four key dimensions of the management of a firm in an adequate way (equilibrium).
The Business Failure Process: An Integrative Model of the Literature
307
c. The resource deployment of the firm (its managerial and organizational processes), shaped by its (specific) resource position and by its environment determines its competitive position and more globally its strategic position (Barney, 1991; Thornhill and Amit, 2003, De Wit and Meyer, 2004). More specifically, the strategic position of the firm is determined by the value of its activities6 and by its competitive position, i.e. its position on the market compared to its competitors (related to key success factors under the control of the firm that may confer it a competitive advantage) (Atamer and Calori, 1993). Thus, if the resources of the firm and their deployment are inadequate to respond to internal and external pressures, the firm cannot create or sustain a
Review of Business and Economics
To summarize points a and b, the origin of the failure of a firm is related to deficiencies in its resource base and in its deployment. These deficiencies limit the ability of the firm to respond to internal or external pressures and this leads to a misalignment between the firm (its resources and its management) and its environment.
2008 / 3
According to the four dimensions of the BSC, here are some examples of deficiencies in the resources deployment of a firm that might be at the origin of its failure. Concerning the strategy and development policy of the firm, some of the failure factors frequently evoked in the literature are the absence of strategic planning (Sharma and Mahajan, 1980), an insufficient or an excessive diversification strategy (Sheppard, 1994), bad management of the firm’s growth (Malecot, 1981; Laitinen, 1992), an inadequate investment policy (Ooghe and Waeyaert, 2004) or corporate governance problems (Sheppard, 1994; Dehaene et al., 2001). Concerning deficiencies in the commercial policy of the firm, the excessive dependence of the firm towards a limited number of clients (Van Caillie et al., 2006), the difficulty to determine an adequate sales price (Brilman, 1985), the inadequacy of the “product-market” couple (Slatter, 1984), the absence/insufficiency of promotion (Ooghe and Waeyaert, 2004) or a poor management of the recovery of account receivables (Newton, 1985) are examples often cited in the literature. Concerning the operational policy, deficiencies regularly presented in the literature are a problematic localisation of the firm (Newton, 1985), the absence or the inadequacy of organizational routines (Stinchcombe, 1965) or a deficient organization of the internal/external value chain of the firm (Marco, 1989; Van Caillie et al., 2006). Concerning inadequacies in the financial policy of the firm, a deficient accounting system, a poor administrative management (Ooghe and Waeyaert, 2004) or the absence of previsions or budgets (Argenti, 1976; Sheppard, 1994) are often mentioned as causes of failure in the literature.
Review of Business and Economics
2008 / 3
308
N. Crutzen & D. Van Caillie
valuable strategic position (Martinet, 1988; Thornhill and Amit, 2003). It cannot develop or sustain valuable activities and/or a viable or sustainable competitive position. d. Then, if no corrective actions are taken to (re)align the resources of the firm and their deployment with the environment requirements, the failing firm enters into an organizational downward spiral (Argenti, 1976; Marco, 1989; Van Caillie, 1992). More precisely, the inadequate way in which the firm is organized and managed as well as its non viable or non sustainable strategic position have a negative impact on the value and onto the size of its set of resources (Newton, 1985): its resources deteriorate more and more (Newton, 1985; Gazengel, 1991 in Liefhooghe, 1997). The increasing deficiencies in the resource base lead to an increasing inadequacy in their deployment. For example, when a firm is failing, the needed strategic changes are not taken (strategic inertia) or a series of strategic decisions, which are not adapted to the situation, are made (Hambrick and D’Aveni, 1988; Sheppard, 1994). These growing inadequacies in the resources of the firm and in their deployment generate more and more difficulties to create or maintain a valuable strategic position (Martinet, 1988; De Wit and Meyer, 2004). Etc. The sequence of interrelationships presented above happens again and again.7 As the company is evolving in this downward spiral, the problems and the inadequacies worsen (Newton, 1985). The relationships between various dimensions become more pronounced and the opportunities to recover, i.e. to come out of the downward spiral and the failure process in general, progressively whittle down. (Weitzel and Jonsson, 1989; Liefhooghe, 1997).
b. Phase 2 – Appearance of failure symptoms (TS)
After a time lag that varies from one failing firm to another, failure symptoms appear. During the first phase of the failure process, one or a number of unfavourable conditions can be quietly developing without being immediately recognizable by outsiders or even by the management. Alternatively, at this second stage of the failure process, the increasing deterioration of the firm’s situation begins to be externally observable, namely in the deterioration of its financial indicators. Indeed, the failure symptoms, which are mainly financial ones, are the first failure signals that are recognizable by outsiders. It is worth noting that the failing firm continues to evolve in the organizational failure downward spiral until the end of the failure process. Its resources are more and more deficient, deployment is more and more inadequate and its strategic position
The Business Failure Process: An Integrative Model of the Literature
309
The chronological sequence of failure symptoms can be explained as follows:
2008 / 3
is weaker and weaker. Actually, the deterioration of (financial) failure symptoms is only the translation of the deteriorating organizational situation of the firm into observable elements.
a. As stated by Thornhill and Amit (2003), two types of failing firms may be distinguished: the (younger) failing firms that never succeed in obtaining a viable strategic position or the (older) failing firms that did succeed in having a good strategic position but that cannot sustain it. The poor position of the firm or the weakening of its position leads to insufficient sales compared to the amount of charges it has to support (Ooghe and Van Wymeersch, 1986; Marco, 1989; Liefhooghe, 1997; Thornhill and Amit, 2003). b. This leads to a lack or a decrease in the profitability of the firm (Ooghe and Van Wymeersch, 1986, Marco, 1989; Liefhooghe, 1997) and in its competitiveness (Marco, 1989; Liefhooghe, 1997). These problems can respectively be observed in the poor (or decreasing) firm’s profitability ratios and in the poor (or declining) market share held by the firm. As the first failure signals are observable by outsiders, the partners of the firm begin to entertain some fears regarding the situation of the firm. For example, the bankers ask for some additional guarantees before a loan is conceded and the suppliers want to be paid more rapidly (Malecot, 1981; Marco, 1989).
d. This lack of cash flow, eventually combined with excessive investments (Ooghe and Van Wymeersch, 1986), leads to a lack of liquidity. e. As the firm needs cash, the leaders of the firm are obliged to look for external financing (Newton, 1985; Ooghe and Van Wymeersch, 1986; Marco, 1989; Laitinen, 1992; Liefhooghe, 1997). Indeed, as the situation of the firm has deteriorated and as the partners begin to have some doubts about it, it is difficult to attract new investors in order to increase the equity of the firm. The leaders are thus forced to increase the level of external debts (banks, public institutions, etc.) and, as a result, the firm’s solvency decreases. f. The increase in the level of external debts of the firm finally leads to an increase in its financial charges (Ooghe and Van Wymeersch, 1986; Marco, 1989; Liefhooghe, 1997). g. The increase in the financial charge of the firm combined with the increasing deterioration of its organizational situation leads to an additional decrease in the firm’s profitability and competitiveness. Subsequently, there is a more substantial lack of cash flow and more significant liquidity problems. That is why
Review of Business and Economics
c. The insufficient or decreasing profitability of the firm leads to a lack of cash flow (Ooghe and Van Wymeersch, 1986; Laitinen, 1992; Liefhooghe, 1997).
2008 / 3
310
N. Crutzen & D. Van Caillie
the level of external debts increases and the financial charges too... etc. (Marco, 1989; Liefhooghe, 1997).
c. Phase 3 – Appearance of critical warning signals (TX) After a variable period of time (Laitinen, 1991), the increasing deterioration of the firm’s organizational and financial situation results in the appearance of critical warning signals. More specifically, the financial situation of the firm has significantly deteriorated (its solvency and its liquidity are critically low) and the firm’s partners start to distrust the firm. At this third stage of the failure process, the failing firm can be considered as “theoretically bankrupt”: it meets the criteria of bankruptcy (critical insolvency and illiquidity) but, practically, the legal and official procedure of bankruptcy in the Court of Commerce has not been activated yet. As soon as the firm has entered this third phase of the failing process, the events follow each other quite rapidly (Marion, 1993 in Liefhooghe, 1997). Thus, the firm may go bankrupt within a few months and it is generally too late to reverse the trend (Newton, 1985; Weitzel and Jonsson, 1989).
Review of Business and Economics
d. Phase 4 – Bankruptcy of the firm (TZ) In this integrative model, the combination of events proposed in the phases 1, 2 and 3 ends with the bankruptcy of the firm. However, as explained before, other negative outcomes of a failure process are possible. In Belgium, the juridical procedure of bankruptcy starts when the leaders of the distressed firm make a confession of bankruptcy because they are conscious the firm’s situation is irrecoverable. Nevertheless, if the leaders do not confess, the unpaid creditors may ask the Court of Commerce to pronounce the firm bankrupt (Gerard et al., 1998). Legally a firm is considered to be bankrupt on the basis of the two following criteria: serious insolvency and illiquidity. More globally, according to Martinet (1988), a firm is (economically) bankrupt when: Its resources have strongly deteriorated: they are insufficient and inadequate. The deployment of these resources is deficient. Similarly, Levinthal (1991) states “a failing firm disappears when its organizational capital, i.e. its resource base and the way it is deployed, is below the minimum viable threshold”. The strategic position of the firm is weak and not adequate. For example, the firm has lost market share and its reputation has collapsed (Weitzel and Jonsson, 1989).
The Business Failure Process: An Integrative Model of the Literature
Summary and implications
An analysis of the literature on business failure prevention leads to the identification of gaps in this literature. Particularly a unifying conceptual framework on how and why companies fail is currently lacking. In order to fill in this particular gap, this paper proposes, from an interpretative perspective, an original and integrative model on how and why companies fail. This unifying model counters a large part of the limits of previous contributions: it presents all the different phases of the failure process, it combines the organizational and financial approaches to business failure and it is based on a deductive reasoning rather that on an inductive one.
2008 / 3
IV.
311
From a conceptual point of view, this integrative model is a first step towards a consensus on how and why companies fail. As lots of researchers argue (Dimitras et al. 1996; Mellahi and Wilkinson, 2004), a unifying conceptual framework is necessary to make further progress in the research on business failure and, particularly, in the prevention of this phenomenon. From a managerial point of view, it offers owners, managers or even external advisors, who have access to internal data, crucial information about the potential causes and consequences of the failure of their firm. This model allows them to go back to the early stages of failure, understand them and perhaps find remedies to its causes. So, it helps organizational leaders to identify strategies to prevent or reverse failure.
Review of Business and Economics
To date, the present research has some theoretical and empirical limitations. Firstly, up to now, we have decided to exclude from our research all the contributions that belong to the business failure prediction literature because we consider their objective is specific, i.e. to discriminate between firms in a credit risk management perspective without highlighting why and how some firms have a significant risk of failure, compared to our aim, i.e. to identify why and how firms enter and evolve in a failure process in order to anticipate and to prevent their potential failure or bankruptcy. However, it would be interesting to consider, in a future research, the (non) financial variables used in predictive models. More specifically, it would be interesting to identify which ones are most related to the failure risk of firms and to compare them with the factors/variables included in our integrative model. Secondly, the time dimension is still not incorporated in the integrative model of the business failure process. Thirdly, the passages from one phase of the failure process to another should be refined. Indeed clear dynamic links have still not been traced between the different phases and, particularly, between organizational and financial elements. Finally, the integrative model has still not been empirically validated.
2008 / 3
312
N. Crutzen & D. Van Caillie
NOTES 1. Researcher, Centre d’Etude de la Performance des Entreprises, HEC Ecole de Gestion, University of Liège – Contact:
[email protected]. 2. Professor, Centre d’Etude de la Performance des Entreprises, HEC Ecole de Gestion, University of Liège – Contact:
[email protected]. 3. For a clarification of this concept, see Crutzen and Van Caillie (2007a; 2007b). 4. Of course a lot of other resources’ deficiencies may be at the origin of the failure of a firm. A review of all the potential failure causes is not the aim of the present paper. For more information on this point, see, for example, Argenti (1976), Malecot (1981), Newton (1985), Koenig (1985), Ooghe and Waeyaert (2004). 5. In our model, we refer ourselves to these 4 dimensions through, respectively, the notions of strategy and development policy, commercial policy, operational policy and financial policy. 6. In an Activity-Based Management perspective, the activities of the firm are the different interfaces consuming/using up resources in the firm (Lorino, 1991). 7. The number of time this sequence of events happens again is variable and, to date, we are not able to be more accurate about that point.
Review of Business and Economics
REFERENCES Altman, E.I., 1968, Financial Ratios, Discriminant Analysis and the Prediction of Corporate Bankruptcy, Journal of Finance 23, 589-609. Altman, E.I., Marco, G. and Varetti, F., 1994, Corporate Distress Diagnosis: Comparison using linear discriminant analysis and neutral networks, Journal of Banking and Finance 18, 505-529. Amburgey, T.L., Kelly, D. and Barnett, W.P., 1993, Resetting the clock: The dynamics of Organizational Change and Failure, Administrative Science Quarterly 38, 51-73. Amit, R. and Schoemaker, P.J.H., 1993, Strategic assets and organizational rent, Strategic Management Journal 14, 33-46. Argenti, J., 1976, Corporate Collapse: the causes and symptoms (Holsted Press, McGrawHill, London). Atamer, T. and Calori, R., 1993, Diagnostic et décisions stratégiques (Dunod, Paris). Balcaen, S. and Ooghe, H., 2006a, 35 years studies on business failure: an overview of the classical statistical methodologies and their related problems, The British Accounting Review 38, 63-93. Balcaen, S. and Ooghe, H., 2006b, The exit path of failing firms: Which factors determine the time to exit?, Working Paper Series, Vlerick Leuven Gent Management School. Barney, J.B., 1991, Firm Resources and Sustained Competitive Advantage, Journal of Management 17, 99-120. Bates, T. and Nucci, A., 1989, An analysis of small business size and rate of discontinuance, Journal of Small Business Management 27 (4), 1-7. Beaver, W., 1966, Financial Ratios as Predictors of Failure, Empirical Research in Accounting: Selected Studies, Supplement to the Journal of Accounting Research 6, 71-87. Bercovitz, J. and Mitchell, W., 2007, When is more better? The impact of business scale and scope on long-term business survival, while controlling for profitability, Strategic Management Journal 28, 61-79.
The Business Failure Process: An Integrative Model of the Literature
313
2008 / 3 Review of Business and Economics
Bilderbeek, J., 1973, Financieel-economische indices ten behoeve van de bedrijfsbeoordeling, Economisch en Sociaal Tijdschrift 27, 141-155. Brilman, J., 1985, Gestion de crise et redressement d’entreprises (Editions Hommes et Techniques, Paris). Burderl, J. and Schussler, R., 1990, Organizational mortality: The liability of newness and adolescence, Administrative Science Quarterly 35, 530-547. Cameron, K.S., Whetten, D.A. and Kim, M.U., 1987, Organizational dysfunctions of decline, Academy of Management Journal 30, 126-138. Cameron, K.S., Kim, M.U. and Whetten, D.A., 1987b, Organizational effects of decline and turbulence, Administrative Science Quarterly 32, 222-240. Carling, K., Jacobson, T., Linde, J. and Roszbach, K., 2007, Corporate credit risk modelling and the macroeconomy, Journal of Banking and Finance 31 (1), 845-868. Carroll, G. and Delacroix, J., 1982, Organizational mortality in the newspaper industries of Argentina and Ireland: an ecological approach, Administrative Science Quarterly 27, 169198. Carter, N.M., Williams, M. and Reynolds, P.D., 1997, Discontinuances among new firms in retail: The influence of initial resources, strategy and gender, Journal of Business Venturing 12, 125-145. Charest, G., Lusztig, P. and Schwab, B., 1990, Gestion financière (Editions du renouveau pédagogique, Québec). Cochran, A.B., 1981, Small business mortality rates: A review of the literature, Journal of Small Business Management 19 (4), 50-59. Cromie, S., 1991, The problems experienced by young firms, International Small Business Journal 9 (3), 43-61. Crutzen, N. and Van Caillie, D., 2007a, The prevention of business failure: a state of the art, submitted for publication in Accounting History, October 2007. Crutzen, N. and Van Caillie, D., 2007b, The business failure process: towards an integrative model of the literature, EIASM Workshop on Default Risk and Financial Distress, Rennes (France), September 13-14. Daubie, M. and Meskens, N., 2001, Bankruptcy Prediction: Literature Survey of the Last Ten Years, Belgian Journal of Operations Research, Statistics and Computer Science 41. D’Aveni, R., 1989, The Aftermath of organizational decline: a longitudinal study of the strategic and managerial characteristics of declining firms, Academy of Management Journal 32, 577-605. Dehaene, A., De Vuyst, V. and Ooghe, H., 2001, Corporate Performance and Board Structure in Belgian Companies, Long Range Planning 34, 383-398. De Wit, B. and Meyer, R., 2004, Strategy: Process, Content, Context: an International Perspective (Thomson Learning, United Kingdom). Dimitras, A., Zanakis, S. and Zopoudinis, C., 1996, A survey of business failures with an emphasis on failure prediction methods and industrial applications, European Journal of Operational Research 90, 487-513. Fichman, M. and Levinthal, D.A., 1991, Honeymoons and the liability of adolescence: A new perspective on duration dependence in social and organizational relationships, The Academy of Management Review 16, 442-486. Fitz Patrick, P., 1932, A comparison of ratios of successful industrial enterprises with those of failed firms, Certified Public Accountant, 598-605. Gazengel, A. and Thomas, P., 1991, Les défaillances d’entreprises, E.S.C.P., n° 93-105, pp. 135 in Liefhooghe, B. (1997), Causes et mécanismes des faillites d’entreprises: une synthèse bibliographique, Cahiers de la Faculté des Sciences Economiques, Sociales et de Gestion 189, 1-45.
Review of Business and Economics
2008 / 3
314
N. Crutzen & D. Van Caillie
Gerard, P., Windey, J. and Gregoire, M., 1998, Le concordat judiciaire et la faillite – Lois des 17 juillet et 8 août 1997 (De Boeck et Larcier, Bruxelles). Guilhot, B., 2000, Défaillance d’entreprise: Soixante-dix ans d’analyses théoriques et empiriques, Revue Française de Gestion, 52-67. Hambrick, D. and D’Aveni, R., 1988, Large Corporate Failures as Downward Spirals, Administrative Science Quarterly 33, 1-23. Henderson, A.D., 1999, Firm strategy and age dependence: A contingent view of the liabilities of newness, adolescence and obsolescence, Administrative Science Quarterly 44, 281314. Jaminon, R., 1986, Facteurs explicatifs de faillites, Annales de Droit de Liège 3, 197-207. Johnson, G., Scholes, K., Whittington, R. and Frery, F., 2005, Stratégique (Pearson Education, Paris). Kaplan, R.S. and Norton, D.P., 1992, The Balanced Score Card – Measures that drive performance, Harvard Business School Press, 71-79. Keasey, K. and Watson, R., 1991, The State of the Art of Small Firm Failure Prediction: Achievements and Prognosis, International Small Business Journal 9, 11-28. Koenig, G., 1985, Entreprises en difficultés: des symptômes aux remèdes, Revue Française de Gestion, 84-92. Laitinen, E.K., 1991, Financial ratios and different failure processes, Journal of Business, Finance and Accounting, 649-673. Laitinen, E.K., 1992, Prediction of Failure of a Newly Founded Firm, Journal of Business Venturing 7, 323-340. Larson, C.M. and Clute, R.C., 1979, The failure syndrome, American Journal of Small Business 4 (2), 35-43. Levinthal, D.A., 1991, Random walks and organizational mortality, Administrative Science Quarterly 36, 397-420. Liefhooghe, B., 1997, Causes et mécanismes des faillites d’entreprises: une synthèse bibliographique, Cahiers de la Faculté des Sciences Economiques, Sociales et de Gestion 189, 145. Lorino, P., 1991, Contrôle de gestion stratégique: la gestion par les activités (Dunod, Paris). Luoma, M. and Laitinen, E.K., 1991, Survival analysis as a tool for company failure prediction, Omega International Journal of Management Science 19, 673-678. Marco, L., 1989, La montée des faillites en France: 19ème et 20ème siècle (L’Harmattan, Paris). Martinet, A.C., 1988, Diagnostic stratégique (Vuibert, Paris). Malecot, J-F, 1981, Les défaillances: un essai d’explication, Revue Française de Gestion, 1019. Marion, A., 1993, Le diagnostic d’entreprise (Economica, Paris). Mellahi, K and Wilkinson, A., 2004, Organizational failure: a critique of recent research and a proposed integrative framework, International Journal of Management Reviews 5/6, 2141. Mellahi K., 2005, The Dynamics of Boards of Directors in Failing Organizations, Long Range Planning 38, 261-279. Miller, D and Friesen, P.H., 1983, Successful and unsuccessful phases of corporate life cycle, Organizational Studies 4, 339-356. Morris, R., 1997, Early Warning Indicators of Corporate Failure: A critical review of previous research and further empirical evidence (Ashgate Publishing Ltd, Great Britain). Newton, G.W., 1985, Bankruptcy and Insolvency Accounting: Practice and procedure (John Wiley and Sons, New York).
The Business Failure Process: An Integrative Model of the Literature
315
O’Neill, H.M. and Duker, J., 1986, Survival and failure in small business, Journal of Small Business Management 24 (1), 125-135. Ooghe, H. and Van Wymeersch, C., 1986, Modèles prévisonnels de la faillite, Annales de Droit de Liège 3, 183-196.
2008 / 3
Nystrom, P.C. and Starbuck, W.H., 1984, To avoid organizational crisis, Unlearn, Organizational Dynamics 12(4), 53-65.
Ooghe, H. and Van Wymeersch, C., 2006, Traité d’analyse financière (Kluwer, Antwerpen). Ooghe, H. and Waeyaert, N., 2004, Oorzaken van faling en falingspaden: Literatuur overzicht en conceptueel verklaringsmodel, Economisch en Sociaal Tijdschrift 57, 367-393. Ooghe, H. and De Prijcker, S., 2006, Failure processes and causes of company bankruptcy: a typology, Working Paper Series, Vlerick Leuven Gent Management School. Peterson, R.A., Kozmetsky, G., Ridgway, N.M., 1983, Perceived Causes of Small Business Failure: a Research Note, American Journal of Small Business 8, 15-19. Porter, M.E., 1980, Competitive Strategy: techniques for analyzing industries and competitors (Free Press, New York). Rumelt, R.P., 1991, How much does industry matter?, Strategic Management Journal 12, 167185. Sharma, S. and Mahajan, V., 1980, Early Warning Indicators of business failure, Journal of Marketing 44, 80-89. Sheppard, J.P., 1994, Strategy and Bankruptcy: an Exploration into Organizational Death, Journal of Management 20, 795-833. Sheppard, J.P. and Chowdhury, S., 2005, Riding the Wrong Wave: Organizational Failure as a Failed Turnaround, Long Range Planning 38, 239-260. Sheth, J., 2005, Why good companies fail, European Business Forum, Issue 22, 24-30. Shumway, T., 2001, Forecasting bankruptcy more accurately: A simple hazard model, Journal of Business, 74 (1), 101-124.
Singh, J.V., House, R.J. and Tucker, D.J., 1986b, Organizational change and organizational mortality, Administrative Science Quarterly 31, 587-61. Slatter, S., 1984, Corporate recovery: successful turnaround strategies and their implementation (Penguin Books, New York). Stinchcombe, A.L., 1965, Social structure and organisations in J.G. March, Handbook of Organisations, (Rand McNally, Chicago) 142-193. Sullivan, T.A., Warren, E. and Westbrook, J., 1999, Financial Difficulties for Small Businesses and Reasons for Their Failure, US Small Business Administration Office of Advocacy Research Study N° 188. Sutton, R.I., 1987, The process of organizational death: disbanding and reconnecting, Administrative Science Quarterly 32, 542-569. Teece, D.J., Pisano, G. and Shuen, A., 1997, Dynamic capabilities and strategic management, Strategic Management Journal 18, 509-533. Thornhill, S. and Amit, R., 2003, Learning about Failure: Bankruptcy, Firm Age and the Resource-Based View, Organization Science 15, 497-509. Thietart, R. A., 2003, Méthodes de recherche en management (Dunod, Paris). Tock, T., 1985, Modèle intégré de gestion d’un portefeuille de débiteurs commerciaux sur base d’informations non bilantaire: Application au cas d’une compagnie d’assurancecrédit in Jaminon (1986). Turner, I., 2005, Corporate failure, Henley Manager Update 17, 33-42.
Review of Business and Economics
Singh, J.V., Tucker, D.J. and House, R.J., 1986a, Organizational legitimacy and the liability of newness, Administrative Science Quarterly 31, 171-193.
Review of Business and Economics
2008 / 3
316
N. Crutzen & D. Van Caillie
Tushman, M.L., Newman, W.H. and Romanelli, E., 1986, Convergence and Upheaval: Managing the Unsteady Pace of Organizational Evolution, California Management Review 29, 29-44. Van Beirendonck, L., 2004, Management des compétences: Evaluation, développement et gestion (De Boeck, Bruxelles). Van Caillie, D., 1992, La prévention des faillites des petites et moyennes entreprises débutantes: Propositions de modèles empiriques, financiers et organisationnels, utilisables par un analyste externe en vue de détecter les forces et les faiblesses d’une PME débutante, Thèse présentée en vue de l’obtention du grade de docteur en sciences de gestion (Université de Liège). Van Caillie, D. and Dighaye, A., 2002, The concept of “Economic Added Result”, a new tool to prevent bankruptcy?, European Accounting Association Congress, 25th-27th April. Van Caillie, D., 2004, La faillite d’entreprise, champ d’application privilégié des méthodes quantitatives en gestion, in Blair, J. and Henry, V., Regards croisés sur l’usage des méthodes quantitatives en gestion (Editions de l’Université de Liège). Van Caillie, D., Santin, S., Crutzen, N. and Kabwigiri, C., 2006, L’analyse équilibrée des symptômes de déséquilibre de la PME à reprendre, facteur-clé du succès du processus de reprise: légitimation théorique et première validation empirique, Communication aux Premières Journées G. Doriot, Normandie, mars. Van Wymeersch, C. and Wolfs, A., 1996, La “trajectoire de faillite” des entreprises: une analyse chronologique sur base des comptes annuels, Cahiers de la Faculté des Sciences Economiques et Sociales 172. Weitzel, W. and Jonsson, E., 1989, Decline in Organizations: A literature Integration and Extension, Administrative Science Quarterly 34, 91-109. Wernerfelt, B., 1984, A resource-based view of the firm, Strategic Management Journal 5, pp. 171-180. Wichman, H., 1983, Accounting and marketing – Key small business problems, American Journal of Small Business 7, 19-26. Whetten, D.A., 1980, Organizational Decline: A neglected Topic in Organizational Science, Academy of Management Review 5, 577-588.
G.L.L. Reniers & A. Audenaert
2008 / 3
Innovatief kosten/baten model voor de optimalisatie van beslissingen aangaande veiligheidsinvesteringen in een chemische onderneming 1
ABSTRACT
JEL CODES
SLEUTELWOORDEN
I.
D0, L0, L2, L5 kosten/baten analyse, optimalisatie van veiligheidsinvesteringen, preventiekosten in de chemische nijverheid.
Inleiding
De baten die veiligheidsinvesteringen met zich kunnen meebrengen worden bij de grote meerderheid van bedrijven niet optimaal in rekening gebracht. Doorgaans worden investeringen in veiligheidsmaatregelen nog teveel gezien als kosten voor de onderneming die weinig bijdragen tot de bedrijfsresultaten en tot de financiële perfor-
Review of Business and Economics
Dit artikel gaat na in welke mate een investering in veiligheid voor het plaatsen van een nieuwe scheikundige installatie rendabel is. De auteurs stellen een management-ondersteunend kosten/baten model voor dat de veiligheidsinvesteringen van een chemische installatie identificeert en evalueert. In het innovatieve model wordt onderscheid gemaakt tussen zware ongevallen en niet-zware ongevallen wat toelaat een reëel beeld te bekomen van kosten en baten van preventie. In klassieke kosten/baten analyses waarbij deze differentiatie niet wordt gemaakt, bekomt men immers slechts een rudimentair beeld van de eventuele rendabiliteit van veiligheidsinvesteringen waar slechts beperkte managementconclusies uit kunnen worden getrokken. Het voorgestelde model wordt toegepast in een reële praktijkstudie waarbij de veiligheidsinvesteringen voor een te plaatsen chemische installatie worden afgewogen t.o.v. de geschatte hypothetische baten van de te implementeren preventieve maatregelen op de installatie. In de beschouwde casus blijkt dat de voorgestelde preventie-investeringen goed verantwoordbaar zijn.
Review of Business and Economics
2008 / 3
318
G.L.L. Reniers & A. Audenaert
mantie van de onderneming en worden de baten van veiligheidsinvesteringen onderschat. Nochtans leidt een verhoging van bedrijfsveiligheid niet alleen tot onmiddellijke vermindering van financieel verlies wanneer een incident kan worden vermeden, maar tevens tot indirecte winst door de betere sociale aanvaarding van de onderneming door de maatschappij. Dit staat nog los van het verbeterde klimaat binnen het bedrijf waardoor bv. looneisen kunnen getemperd worden. Hoewel de financiële impact van het imago van een onderneming een zeer moeilijk te kwantificeren begrip is, is algemeen bekend dat een negatieve gebeurtenis als gevolg van gebrekkige veiligheidsmaatregelen een zeer ernstige nefaste impact kan hebben op de bedrijfsresultaten. Als voorbeeld (Reisch, 2004) van zo een negatieve gebeurtenis kan de grootste ramp in de geschiedenis van de industriële chemie aangehaald worden, namelijk het ongeval van Bhopal in Indië (1984). Bij dit ongeval vonden ongeveer 3.800 mensen rechtstreeks de dood, terwijl vele tienduizenden tot op de dag van vandaag lijden aan ernstige gezondheidsproblemen. De negatieve effecten van deze ramp hadden zowel economische als sociale gevolgen voor Union Carbide, het chemische bedrijf dat de vestiging in Bhopal uitbaatte. Hun aandelenprijs kende een onmiddellijke drastische terugval van 27,9% hetgeen zich vertaalde in een totaal verlies aan marktwaarde van meer dan 1 miljard dollar (Blacconiere en Patten, 1994). De organisatie heeft meer dan twintig jaar na de ramp nog steeds te kampen met het negatieve publieke imago. Bewijs daarvan is de daling in aandelenprijs van Dow bij de overname van het resterende Union Carbide in 2001 (Wilkins, 1987). De vermindering van financieel verlies bij het vermijden van incidenten door het implementeren van preventiemaatregelen is, net zoals de financiële impact van het bedrijfsimago, een heel theoretisch gegeven, en is dus heel moeilijk te kwantificeren. In een enquête uitgevoerd in 2003 (Reniers et al., 2003) werd gepeild naar de inspanningen die bedrijven leveren om deze economische oefening te maken. Op de 24 ondervraagde chemische bedrijven bleek geen enkele onderneming de hypothetische baten van haar geïmplementeerde veiligheidsmaatregelen nauwkeurig te berekenen. Ondanks de grote onzekerheid van de gegevens betreffende sommige batenposten, is het toch nuttig om de theoretische vermindering aan verlies door het vermijden van incidenten, te bepalen. Het bedrijfsmanagement krijgt zo immers een beter inzicht in veiligheidsinvesteringen en kan met meer accurate cijfers beslissingen nemen betreffende veiligheid en het aanvaarden van de bedrijfsrisicograad. De volgende sectie geeft een theoretische uitwerking van het voorgestelde kosten/baten model dat door het bedrijfsmanagement kan worden gebruikt om veiligheidsinvesteringen binnen de onderneming te analyseren, te evalueren en desgewenst te communiceren. Sectie II bespreekt de kosten en baten voor het praktijkgeval van een nieuw te plaatsen chemische installatie en analyseert en evalueert de resultaten van de studie. Sectie IV besluit dit artikel met een samenvatting van het voorgestelde model en haar concrete toepassing.
Innovatief kosten/baten model: veiligheidsinvesteringen in een chemische onderneming
Kosten/baten model voor veiligheidsinvesteringen
A. Achtergrond
Aangezien de veiligheidsgraad van een bedrijf niet rechtstreeks kan worden gekwantificeerd, gebruikt men zogenaamde “veiligheidsprestatie indicatoren” om de graad te meten. Deze indicatoren leiden tot de implementatie van maatregelen waarvan men de kosten kent. De veiligheidsindicatoren (en dus ook de kosten van preventieve maatregelen) kunnen worden opgedeeld in verschillende categorieën zoals brand- en explosieveiligheid, procesveiligheid, contractorveiligheid, transportveiligheid, arbeidsveiligheid, etc. De ontwikkeling van het bedrijfsveiligheidsbeleid bevat naast directe kosten als het aanwenden en implementeren van veiligheidsmateriaal, ook indirecte kosten als training en ontwikkeling van de werknemers en onderhoud van het bedrijfsveiligheidsmanagementsysteem.
De niet-kwantificeerbare ongevalkosten kunnen niet gedetailleerd in een algemeen model (dat toepasbaar is op een arbitraire chemische onderneming) worden opgenomen, aangezien zulke kosten sterk afhankelijk zijn van niet-generieke data zoals persoonsgebonden kenmerken, de bedrijfscultuur en/of de maatschappij in haar geheel. Ze treden echter wel op, waardoor de werkelijke kosten de kwantificeerbare kosten overstijgen. Om niet-kwantificeerbare kosten te bepalen maakt men in de economie (bv. bij milieu-gerelateerde probleemstellingen) vaak gebruik van monetaire waarderingstechnieken zoals de contingente waarderingsmethode, conjoint analyse of hedonische methoden (Hanley en Spash, 1993). In het geval van niet-kwantificeerbare ongevalkosten wijzen verschillende studies uit dat deze een veelvoud vormen van de kwantificeerbare kosten (Mossink en Licher, 1997). Wetenschappelijk betrouwbare cijfers bestaan echter niet. De kwantificeerbare socio-economische ongevalkosten (zie tabel 1) kunnen worden opgedeeld in directe en indirecte kosten. De directe kosten geven bij een toestand waar er zich geen ongevallen voordoen aanleiding tot directe hypothetische baten,3 terwijl de indirecte kosten aanleiding geven tot indirecte hypothetische baten. Indirecte hypothetische baten ontstaan bijvoorbeeld als gevolg van niet-ziekteverzuim of niet-afwezigheid, niet-personeelsverloop, niet-arbeidsongeschiktheid en niet-veranderingen in de werksfeer.
Review of Business and Economics
Een ongeval gaat gepaard met allerhande directe en indirecte kosten. We kunnen dus stellen dat door de implementatie van een degelijk veiligheidsbeleid en door het nemen van preventieve veiligheidsmaatregelen2 er kosten worden vermeden, namelijk de kosten van ongevallen die nooit hebben plaatsgevonden. Op die manier worden ongewenste gevolgen vermeden op socio-economisch vlak, maar ook naar het milieu en de omgeving toe. De ‘hypothetische baten’ worden echter door de bedrijven niet of zeer onvoldoende in kaart gebracht omwille van hun complexiteit (Reniers et al., 2003). In Tabel 1 worden de mogelijke sociale kosten die ongevallen met zich meebrengen, zichtbaar (De Greef & Van den Broek, 2004).
2008 / 3
II.
319
Review of Business and Economics
2008 / 3
320
G.L.L. Reniers & A. Audenaert
Tabel 1. Kwantificeerbare en niet-kwantificeerbare socio-economische gevolgen van ongevallen. Belanghebbenden: Niet-kwantificeerbare gevolgen van ongevallen
Kwantificeerbare gevolgen van ongevallen
Slachtoffer(s)
Pijn en lijden Moreel en psychisch lijden Medische en familiale lasten Verminderde arbeidsmotivatie Angst Stress
Verlies aan loon en premies Beperking van beroepsvaardigheden Tijdverlies (medische behandeling) Financieel verlies Extra kosten
Collega's
Slecht gevoel Ongerustheid of paniek Verminderde arbeidsmotivatie Angst Stress
Tijdverlies Eventueel verlies aan premies Meer werkbelasting Opleiding en begeleiding van tijdelijke werknemers
Organisatie
Verslechtering van het sociale klimaat Slecht imago
Interne doorlichting Schade aan uitrusting en materieel Productieverlies Verlies aan kwaliteit Nieuwe opleiding van personeel Technische storingen Organisatorische kosten Hogere productiekosten Hogere verzekeringspremie Administratieve kosten Sancties van moederbedrijf Sancties van overheid
Bron: gebaseerd op De Greef, M. & Van den Broek, K., 2004.
Een schatting van de grootte van de input en output van een veiligheidsbeleid is dus duidelijk allerminst eenvoudig. Het is onmogelijk om de kosten of de baten te bepalen van één afzonderlijke maatregel. De effectiviteit (en de kosten en baten) van het veiligheidsbeleid moet als een geheel bekeken worden. We dienen bijgevolg een model op te stellen dat zowel rekening houdt met de geaccumuleerde veiligheidsinvesteringen (kosten) als met de mogelijke gesommeerde (hypothetische) baten die deze investeringen met zich meebrengen. Op die manier bekomen we een kosten/baten model dat de degelijkheid van managementbeslissingen aangaande veiligheidsinvesteringen werkelijk kan sturen en duurzaam verbeteren.
Innovatief kosten/baten model: veiligheidsinvesteringen in een chemische onderneming
321
Om een beeld te krijgen van de verschillende soorten kosten zullen we enerzijds een onderverdeling maken tussen vaste en variabele kosten en anderzijds tussen directe en indirecte kosten.
2008 / 3
B. Kostenposten van preventieve maatregelen
Vaste kosten blijven constant, ongeacht de veranderingen van ondernemingsactiviteiten. Een voorbeeld van een vaste kost is bijvoorbeeld de aankoop van een brandwerende deur. Deze aankoop gebeurt slechts eenmalig en de kosten ervan variëren niet naargelang de productie. Variabele kosten wijzigen in tegenstelling tot vaste kosten wél proportioneel naargelang de ondernemingsactiviteiten. Het aankopen van veiligheidshandschoenen is een variabele kost omdat deze sneller aan vervanging zullen toe zijn wanneer de handschoenen veelvuldig of intensiever gebruikt zullen worden bij de toename in productie. Directe kosten houden rechtstreeks verband met de productie. Wat veiligheidskosten betreft, gaat het hier bijvoorbeeld om kosten van een afzuigsysteem. Indirecte kosten daarentegen staan niet rechtstreeks in verband met het productieniveau. Een omgevingsveiligheidsrapport bijvoorbeeld zal aangeven waar de opslag van gevaarlijke stoffen moet gebeuren, maar zal niet rechtsreeks een effect hebben op de productie.
C. Batenfactoren
Een ongeval dat niet optreedt in een onderneming dankzij de aanwezigheid in het bedrijf van een adequaat veiligheidsbeleid, wordt in het vervolg van de tekst kortweg een niet-ongeval genoemd. Het hebben van niet-ongevallen (of m.a.w. het verhinderen van ongevallen) leidt ertoe dat een aantal kosten niet voorkomen en creëert aldus hypothetische baten. Naast het identificeren van de directe hypothetische baten die niet-ongevallen met zich meebrengen, dienen tevens de indirecte hypothetische baten te worden meegerekend die volgen bij een niet-ongeval uit niet-ziekteverzuim of niet-afwezigheid, niet-personeelsverloop, niet-arbeidsongeschiktheid en niet-veranderingen in de werksfeer. Een eerste batenfactor zijn de besparingen die het zich niet voordoen van ongevallen, kortweg ‘niet-ongevallen’, met zich meebrengen. Deze hypothetische baten
Review of Business and Economics
Voor elk van de vijf categorieën waarin veiligheidsindicatoren worden ontwikkeld (op basis van interne bedrijfsdocumenten van de onderneming waar een gevalsstudie werd uitgevoerd: brand-en explosieveiligheid, procesveiligheid, contractorveiligheid, transportveiligheid en arbeidsveiligheid) dienen de vaste en variabele directe kosten en de vaste en variabele indirecte kosten te worden geïdentificeerd voor het deel van het bedrijf waar men de veiligheidsinvesteringen wenst te analyseren en te evalueren. Bedrijven hebben meestal ervaring in het berekenen van de directe en indirecte kosten van preventiemaatregelen.
Review of Business and Economics
2008 / 3
322
G.L.L. Reniers & A. Audenaert
kunnen door de onderneming geïdentificeerd worden door het niet stukgaan van de installatie (hoeveel kost de installatie of een onderdeel ervan?), door het niet moeten betalen van een hogere verzekeringspremie (wat is de meerprijs van de premie indien een specifiek ongeval zich zou voordoen?), etc. Deze hypothetische baten van niet-ongevallen op ondernemingsniveau kunnen in 5 categorieën opgedeeld worden. Een eerste categorie behelst de niet-verloren werktijd. Onder deze categorie valt het niet moeten uitbetalen aan de werknemer die op het ogenblik van een ongeval geen waarde meer creëert voor de onderneming (en waar het blijven uitbetalen dus een pure kost zou zijn). Het niet-verlies van korte termijn activa is een tweede batenpost en kan bijvoorbeeld het niet-verlies van ruwe materialen omvatten. Een derde tak slaat dan weer op de lange termijn activa met mogelijk niet-verlies van machines. Diverse korte termijn baten zoals niet-transportkosten en niet-boetes zijn een vierde categorie. Een vijfde hoofding bevat niet-verloren inkomsten door niet-gemiste contracten of niet-prijsreducties (Rikhardsson, 2006). Het is duidelijk dat de zichtbare of directe baten door het vermijden van kosten t.g.v. niet-ongevallen slechts een klein deel vertegenwoordigen van de totale baten die niet-ongevallen met zich meebrengen (bv. niet slechter imago, vermijden van rechtszaken, etc.). De tweede batenfactor is de analyse van de hypothetische indirecte baten die nietziektedagen en niet-afwezigheid met zich meebrengen. Logischerwijze zijn de nietloonkosten tijdens de niet-ziektedagen hét meetinstrument om de mogelijke baten in kaart te brengen. Nochtans is deze zienswijze niet helemaal correct omdat tal van indirecte baten hiermee gepaard gaan. Vermits dezelfde loonkosten tijdens het aanwezig zijn van het personeelslid moeten betaald worden vertegenwoordigen deze niet-loonkosten geen additionele baten van niet-absenteïsme. De werkelijke baten van niet-absenteïsme moeten daarom gezocht worden in de impact die deze baten hebben op de ondernemingsactiviteiten. Omwille van eventuele afwezigheid zal de totale hoeveelheid aan arbeidsinspanningen dalen, waardoor de productiviteit van de onderneming verstoord wordt. De vraag of er net genoeg of teveel personeel aanwezig is om een bepaalde productiviteit te halen heeft een sterke invloed op de impact van de baten van niet-afwezigheid. Indien de onderneming personeelscapaciteit over heeft dan zal het werk immers toegewezen worden aan andere werknemers, waardoor de baten van niet-afwezigheid minder belangrijk worden. Ook baten van niet-nieuwe arbeid, niet-vervanging, het niet moeten opleiden of geen overwerk moeten verrichten, dienen geïncorporeerd te worden in het kosten/baten model. Het effect van niet-absenteïsme hangt sterk af van hoe de onderneming zichzelf heeft georganiseerd. Bij het niet-vervangen van een personeelslid of niet genoodzaakt toewijzen van de opdrachten aan andere werknemers zullen er eigenlijke baten van management en administratie ontstaan. Additionele baten zullen daarom variëren naargelang de efficiëntie van deze systemen. Hoe de onderneming zich heeft ingedekt tegen gezondheidsrisico’s is een andere factor die de
Innovatief kosten/baten model: veiligheidsinvesteringen in een chemische onderneming
323
2008 / 3
grootte van de additionele baten zal doen fluctueren (Ministry of Social Affairs and Health, 1999). De baten van niet-absenteïsme bevatten dus een hele reeks aan bijkomende baten bovenop de normale niet-loonkosten. De grootte van deze baten verschillen naargelang de structuur van de organisatie. Indien de onderneming het veiligheidsbeleid degelijk uitwerkt zal er bijgevolg een forse besparing op niet-absenteïsme gegenereerd kunnen worden. Een volgende batenfactor van niet-ongevallen is het niet-personeelsverloop. Het verloop van werkkrachten is in een onderneming een goede indicator voor de evaluatie van werkomstandigheden en werktevredenheid. Een hoog percentage kan duiden op lage werktevredenheid. In dergelijk geval zullen werknemers weinig spontane acties ondernemen en zal weinig kennis in de onderneming verspreid worden waardoor dit een negatieve invloed zal hebben op de productiviteit van de processen en de kwaliteit van de producten. Niet elk hoog personeelsverloop is echter resultaat van slechte werkomstandigheden of lage werktevredenheid. Jobhopping of bewust veranderen van werk maakt nu eenmaal deel uit van een huidige carrièreontwikkeling. Ook de uitbreiding van de onderneming kan ervoor zorgen dat er nieuwe functies ontstaan waardoor personeel met verschillende kwalificaties vereist is. Wegens hoge rekruteringskosten van goede werknemers en de kosten voor het trainen van personeel wordt het retentiebeleid wel alsmaar belangrijker in een onderneming. Een imago naar het grote publiek om een goede werkgever te zijn met goede, veilige en gezonde werkomstandigheden speelt een belangrijke rol in het aantrekken van werkkrachten op alle niveaus (ILO, 2001).
Personeels[aantal werknemers die de organisatie verlaten aantal nieuwe werknemers] verloop = [2 het gemiddelde personeelsaantal]
Ondanks de omvang van sommige organisaties dient het management zich ervan bewust te zijn dat de zichtbare baten, zoals het niet moeten betalen van een consultant of het niet moeten plaatsen van een advertentie, slechts een klein deel vertegenwoordigen van de totale baten dankzij een laag personeelsverloop. De onzichtbare baten hebben betrekking op de baten als gevolg van het feit dat een werknemer, dankzij het veiligheidsbeleid en het niet hebben van ongevallen, de organisatie niet
Review of Business and Economics
De onderneming ondervindt hypothetische baten van niet-personeelsverloop wanneer een contract van een werknemer niet stopgezet wordt of wanneer een nieuw contract niet dient te worden opgesteld. Personeelsverloop wordt vaak aangeduid als een percentage van het gemiddelde aantal werknemers. In een onderneming met een groot aantal werknemers zullen de cijfers van personeelsverloop logischerwijze minder grote schommelingen vertegenwoordigen dan een onderneming met een klein aantal werknemers. Het personeelsverloop wordt in de onderneming waarvan in dit artikel een casus wordt beschouwd, als volgt berekend:
Review of Business and Economics
2008 / 3
324
G.L.L. Reniers & A. Audenaert
verlaat. Wanneer een werknemer de organisatie niet verlaat dan is de voornaamste baat de winst aan werkuren waarvoor andere werknemers niet moeten worden ingeschakeld. Managementtijd wordt dan niet besteed aan interviews en routines bij het beëindigen van het contract. De kosten van het rekruteren van nieuwe werknemers kunnen aanzienlijk zijn wanneer de onderneming werknemers met ervaring en ondernemingsspecifieke vaardigheden zou moeten vervangen. Door te onderzoeken welke redenen aan de basis liggen van het rekruteren van nieuw personeel kan een onderscheid gemaakt worden tussen enerzijds het rekruteren van nieuwe werknemers wegens expansie van de organisatie en anderzijds het vervangen van mensen die ontslag zouden genomen hebben wegens een slecht veiligheidsbeleid (Chartered Institute of Personnel and Development, 2005). Een laatste batenfactor zijn de mogelijk verminderde kosten door het niet hebben van een overinvestering in een veranderende werksfeer. Een op de onderneming specifiek toegepast veiligheidsbeleid dat als strategische component actief is in de onderneming zal vanuit het perspectief van veranderingsmanagement financiële voordelen opleveren omdat het beleid geïntegreerd is in alle processen en activiteiten van de onderneming. Het geïntegreerde veiligheidsbeleid is bijgevolg relatief eenvoudig aan te passen per departement en kan dan met weinig moeite bijgestuurd worden. Het vraagt hierdoor ook minder inspanning van het geheel aan werknemers van de onderneming dan het traditionele (niet-geïntegreerde) veiligheidsbeleid (Simons, 2000) waar incidenten en verbeteringen ad hoc worden aangepakt per departement en/of per proces. De leereffecten en de verhoogde efficiëntie creëren de voornaamste financiële voordelen van een geïntegreerd veiligheidsbeleid t.o.v. een niet-geïntegreerd veiligheidsbeleid.
D. Modellering van kosten en baten van veiligheidsinvesteringen
In het te ontwikkelen model dienen de kosten van preventieve maatregelen afgewogen te worden ten aanzien van de hypothetische baten als gevolg van het zich niet voordoen van ongevallen of als gevolg van een degelijk uitgebouwd veiligheids- en gezondheidsbeleid in een onderneming. Vanuit een financieel en economisch standpunt enerzijds en vanuit een beleidsstandpunt anderzijds kan men zich de vraag stellen welke veiligheidsgraad wenselijk is in een organisatie om schade aan mens en infrastructuur van de onderneming te voorkomen. Het uitwerken van een goed preventiebeleid en de opvolging ervan brengen kosten met zich mee, maar het vermijden van ongevallen en schade leidt dan weer tot een hypothetische opbrengst. Bijgevolg moet een onderneming met aandacht voor veiligheid een optimum proberen te vinden tussen enerzijds de investering (kost) in veiligheid en de daardoor gerealiseerde veiligheidsgraad en an-
Innovatief kosten/baten model: veiligheidsinvesteringen in een chemische onderneming
325
Een onderneming die beslist om enkel te investeren in veiligheid op basis van de wettelijke minimumeisen loopt een hoger risico op ongevallen en rampen dan een onderneming waarbij de preventie-investeringen het louter wettelijke overstijgen. Een onderneming die zich focust op wettelijke vereisten zal bij een ongeval dan ook zeer vermoedelijk aan hogere kosten onderhevig zijn. Bovendien is de wettelijke grens op zich ook afhankelijk van interpretaties en types oplossingen voor veiligheidsproblemen en bestaat er geen mogelijkheid om een eenduidige monetaire waardebepaling te maken van de minimumeisen n.a.v. wetgeving. Dit is een gevolg van de evolutie (onder impuls van de Europese wetgever) van prescriptieve regelgeving naar meer performantiegebaseerde of resultaatsgebonden wetgeving (Hale et al., 2002).
Daarnaast krijgen deze ondernemingen ook een preventieadviseur over de vloer die hun moet begeleiden met het uitwerken van een degelijk plan om de arbeidsongevallen drastisch te doen dalen. Er wordt bovendien een bonus-malussysteem ingevoerd voor de arbeidsongevallen-verzekering (Kluwer, 2006) vermits de huidige methode van verzekeren veilige én onveilige bedrijven met minder dan 200 werknemers op een gelijke manier behandelt. Ondernemingen binnen éénzelfde sector met minder dan 200 werknemers betalen immers dezelfde verzekeringspremie. Op deze manier worden de kosten van onveiligheid door de meer onveilige bedrijven deels afgewenteld op de veiligere bedrijven. Via het voorgestelde bonus-malussysteem is het niet enkel de bedoeling om de onveilige bedrijven te bestraffen, maar ook om de ondernemingen die extra inspanningen leveren voor de preventie van arbeidsongevallen, (indirect) te belonen. De bedrijven die een verplichte preventiebijdrage zullen moeten betalen, zijn de ondernemingen die gedurende drie opeenvolgende jaren opmerkelijk slechter scoren op gebied van aantal arbeidsongevallen dan het gemiddelde van hun sector (gemeten op basis van de zogenaamde frequentiegraad6). Deze bijdrage zal tussen
Review of Business and Economics
De Belgische Wetgever is wel bezig met de verdere uitbouw van allerlei concepten en projecten inzake de verdere verbetering van het welzijn op het werk. Het meest recente initiatief is de verplichte preventiebijdrage. Het wetsontwerp met betrekking tot de strengere aanpak van onveilige bedrijven is gepubliceerd in de wet van 13 juli 2006 (Belgisch Staatsblad 1.9.2006), artikelen 54 tot 57. Deze wet vormt een onderdeel van het zogenaamde FARAO-plan3.4 De bedrijven waar aanzienlijk meer ongevallen gebeuren in vergelijking met de sector, betalen vanaf 1 januari 2008 een forfaitaire preventiebijdrage. Er is sprake van verzwaard risico dat gedefinieerd wordt als “wanneer, met uitsluiting van het risico voor de ongevallen op de weg naar of van het werk, gedurende de observatieperiode de frequentie en de ernst van de ongevallen de grens overschrijden”.5
2008 / 3
derzijds het vermijden van schade onder welke vorm dan ook (hypothetische opbrengst).
2008 / 3
326
G.L.L. Reniers & A. Audenaert
3.000 € en 15.000 € liggen, afhankelijk van de grootte van het bedrijf en het aantal ongevallen (Gorle, 2006; Kluwer, 2006). Aangezien bedrijven worden afgerekend op een onvoldoend veiligheidsbeleid en bovendien de hypothetische baten van vermeden ongevallen aanzienlijk kunnen zijn, kan men redelijkerwijze aannemen dat de “optimale investeringsgrens” zich boven de wettelijke minimumgrens bevindt. Anderzijds kan een onderneming ook overinvesteren in veiligheid waardoor de preventie van schade eerder een beperkende functie dan een motiverende functie zal vormen in de productie of in de reële opbrengst met een negatieve impact op de rendabiliteit tot gevolg.
Review of Business and Economics
Elke onderneming dient dus te streven naar een zo goed mogelijke afweging tussen veiligheidskosten en hypothetische baten. Men beoogt een afweging te maken waarbij de investering in veiligheid eveneens de hypothetische baten laat stijgen omwille van een daling in potentiële schadekosten. Hierbij dalen de totale potentiële kosten ondanks de veiligheidsinvesteringen. Kosten en baten kunnen worden uitgezet in functie van de ‘veiligheidsgraad’ (V). De veiligheidsgraad is de ‘mate van veiligheid’ van het bedrijf in kwestie. De theoretische veiligheidgraad kan dus variëren tussen een grote mate van onveiligheid (V = (0 + )%) en een grote mate van veiligheid (V = (100 – )%), waarbij een (kleine) waarde veronderstelt die aanduidt dat ‘volledige onveiligheid’ of ‘volledige veiligheid’ in een onderneming in realiteit niet mogelijk is. Het economische break~ even veiligheidspunt ( V), d.i. het punt waarbij de kosten van veiligheidsmaatregelen (Kv) gelijk worden aan de baten ervan (Bv), kan dan grafisch worden voorgesteld (figuur 1). We onderscheiden hier twee gevallen: zware ongevallen en nietzware ongevallen. Zware ongevallen dienen in deze studie te voldoen aan vier voorwaarden: (i) bij het ongeval zijn één of meer gevaarlijke stoffen betrokken,7 (ii) het ongeval houdt verband met een ongewoon voorval bij een industriële activiteit,8 (iii) de gebeurtenis levert een ernstig gevaar op voor de mens en/of voor het milieu en (iv) één van de kenmerken van het ongeval is het zich voordoen van een explosie, een hevige brand en/of de emissie van een giftige gaswolk. Deze definitie voor een zwaar ongeval is gebaseerd op informatie die ter beschikking gesteld wordt op de officiële website van de Federale Overheidsdienst Werkgelegenheid, Arbeid en Sociaal Overleg (ES1). Alle ongevallen waarbij één van deze criteria niet voldaan is, beschouwen we als niet-zware ongevallen. Dit onderscheid dient gemaakt te worden omdat de kosten en baten van de twee soorten ongevallen wezenlijk verschillen. Bij niet-zware ongevallen zijn de hypothetische baten als gevolg van het zich niet voordoen van ongevallen aanzienlijk lager dan in het geval van zware ongevallen. Dit betekent dat de onderneming in principe veel meer potentiële financiële baten heeft bij investeringen ter preventie van zware ongevallen dan ter preventie van niet-zware ongevallen. De gesommeerde kosten van preventiemaatregelen voor zware ongevallen liggen over het algemeen ook aanzienlijk hoger dan deze voor niet-zware ongevallen. Zware ongevallen worden immers
Innovatief kosten/baten model: veiligheidsinvesteringen in een chemische onderneming
327
Men kan verwachten dat naarmate de onderneming meer investeert in veiligheid, de veiligheidsgraad ook zal toenemen. Men heeft dus een stijgende Kv-curve naarmate V toeneemt. Bovendien is het zo dat bij een hogere veiligheidsgraad het steeds moeilijker wordt om deze nog verder te verbeteren (i.e., te verhogen) en de curve voor de veiligheidsinvesteringen bijgevolg een asymptotisch gedrag vertonen. Men kan ook verwachten dat, naarmate er vanuit een V = (0 + )%, meer in veiligheid wordt geïnvesteerd, er hogere baten kunnen verwacht worden als gevolg van het zich niet manifesteren van ongevallen. Deze curven zullen een veel vlakker verloop vertonen vanwege het feit dat marginale investeringen niet leiden tot grote
Review of Business and Economics
Figuur 1. Grafische voorstelling van de economische break-even veiligheidspunten voor een zwaar ongevalscenario (met index 1) en voor een niet-zwaar ongevalscenario (met index 2).
2008 / 3
voornamelijk vermeden via (dure) technische studies en het plaatsen van (dure) technische apparatuur in het kader van procesveiligheid. Niet-zware ongevallen gaan dan weer meer gepaard met persoonlijke bescherming, EHBO, het dagelijkse beheer van veiligheidsgedrag, etc.
2008 / 3
328
G.L.L. Reniers & A. Audenaert
extra baten. De batencurve en de kostencurve zullen snijden in een break-even vei~ ligheidspunt V. Indien men na dit punt nog meer in veiligheid investeert, zullen de baten niet meer opwegen tegen de kosten. Voor beide ongevalscenario’s (zware ongevallen en niet-zware ongevallen) heeft men een verschillende ligging van de kos~ ten- en batencurves en dus een verschillende ligging van V. Figuur 1 schetst de kwalitatieve batencurven9 voor beide ongevalscenario’s die aangeduid zijn met Bv1 en Bv2. De batencurve met index ‘1’ stellen zware ongevallen voor (bv. domino ongevallen, ongevallen met meerdere slachtoffers, etc.) en deze met index ‘2’ stellen niet-zware ongevallen voor (bv. een aantal dagen werkonbekwaam, EHBO-ongevallen, etc.). Het is duidelijk dat bij niet-zware ongevallen de baten van veiligheidsmaatregelen (dus het vermijden van kosten dankzij niet-ongevallen) Bv2 en de kosten van veiligheidsmaatregelen Kv2 aanzienlijk lager liggen dan bij zware ongevallen (curves Bv1 en Kv1). Het kosten/baten break-even veiligheids~ punt voor niet-zware ongevallen ( V2) ligt eveneens lager dan dit voor zware onge~ vallen ( V1 ).
Review of Business and Economics
Figuur 1 geeft ook aan dat bij zware ongevallen waarbij het bedrijf zeer zware financiële schade oploopt, (i) de hypothetische baten bij het vermijden van zulke ongevallen steeds zeer hoog zijn, en (ii) het break-even punt gesitueerd moet worden bij de (100 – )% veiligheidsgraad grens. Dit houdt in dat bij zulke ongevallen bijna steeds alle noodzakelijke investeringen gerechtvaardigd zijn, ongeacht de grootte van de kostenposten. De vermeden kosten (voor het geval zulke ongevallen zich niet manifesteren) of hypothetische baten zijn immers (bijna) altijd groter dan
Figuur 2. Frequentiegraad van een onderneming in functie van de veiligheidsgraad.
Innovatief kosten/baten model: veiligheidsinvesteringen in een chemische onderneming
329
2008 / 3
de kosten van de nodige veiligheidsmaatregelen. Voor de zware ongevallen kunnen we dus stellen dat er geen echte kosten-baten afweging dient te gebeuren ten opzichte van een bepaalde “veiligheidsgraad van zware ongevallen”, maar dat er een veiligheidsgraad van zware ongevallen van (100 – )% moet gegarandeerd worden, ongeacht de kosten van de veiligheidsmaatregelen hiervoor. Uiteraard dienen de veiligheidskosten voor het vermijden van zware ongevallen eveneens gedeeltelijk voor het vermijden van niet-zware ongevallen. Ook de hypothetische baten van zware en niet-zware ongevallen kunnen niet los van elkaar worden gezien. Daarom houden we bij de bepaling van de totale kosten curve en totale baten curve bij nietzware ongevallen ook rekening met de investeringen naar aanleiding van zware ongevallen. Voorts is het ook zo dat bij niet-zware ongevallen wel concreet een breakeven veiligheidspunt kan worden bepaald per bedrijf. We dienen hiertoe eerst de veiligheidsgraad te definiëren van een onderneming. We stellen dat een veiligheidsgraad van 50% overeenkomt met de gemiddelde frequentiegraad van de industriële sector waartoe het bedrijf behoort en dat met een maximale veiligheidsgraad van (100 – )% een bedrijfsfrequentiegraad van 0 overeenkomt. Figuur 2 schetst deze gedefinieerde relatie tussen veiligheidsgraad en frequentiegraad. Dit verband kan worden uitgedrukt als volgt: Vbedrijf
2
bedrijfsfrequentiegraad gemiddelde frequentiegraad van de industriële sector
1
Idealiter zal een bedrijf de impact van het veiligheidsbeleid optimaliseren, m.a.w. de netto-baten van haar veiligheidsbeleid maximaliseren, gegeven de beperking dat aan de ondernemings- en wettelijke minimumvereisten moet worden voldaan. Echter, hiervoor zijn meer accurate gegevens nodig. Indien dit minimum niet accuraat kan worden vastgelegd, kan de optimalisatie ook niet degelijk worden uitgevoerd. Uit een enquête verricht in 2003 door Reniers en uit Mossink en Licher (1997) blijkt dat bedrijven in het beste geval enkel schattingen van kosten en baten van veiligheidsinvesteringen bezitten en hanteren om beslissingen te nemen. Meestal is het zelfs zo dat kosten en baten niet worden beschouwd voor de preventie-investeringen, maar dat deze plaatsvinden op basis van ervaring. Het model dat in dit artikel wordt voorgesteld dient in praktijk te kunnen worden gebruikt om objectieve investeringsbeslissingen te kunnen nemen via voorhanden zijnde – of relatief eenvoudig te achterhalen – gegevens. We opteren er dus voor om de kosten-en batencurves uit figuur 1 als totale kosten en totale baten te beschouwen (en niet als marginale kosten en marginale baten). Om managementbeslissingen een objectieve input te kunnen geven, dient het model aan te geven of een veiligheidsinvestering bij de veiligheidsgraad die het bedrijf
Review of Business and Economics
waarbij we een minimum veiligheidsgraad gelijk aan (0 + )% veronderstellen (dus Vbedrijf 0) en waarbij we verwaarloosbaar klein veronderstellen.
2008 / 3
330
G.L.L. Reniers & A. Audenaert
karakteriseert, de gewenste baten oplevert voor het bedrijf. Hiervoor stelt elk be~ drijf voor zich een gewenst break-even veiligheidspunt van Vx voorop (bijvoorbeeld 15% hoger dan de gemiddelde frequentiegraad voor de gehele industriële sector waartoe het bedrijf behoort). Op die manier kan een onderneming een voortdurende streefwaarde van de veiligheidsgraad voor niet-zware ongevallen in de hand werken. We berekenen Kv en Bv aan de hand van de in de vorige secties voorgestelde kostenposten en batenposten. Vervolgens dienen we de berekende veiligheidsinvesteringen met de hypothetische baten met elkaar af te wegen. Om de afweging vanuit bedrijfsperspectief op een objectieve manier te kunnen maken, voeren we, voor de duidelijkheid van het model, twee voorwaarden in die bepalen of er al dan niet extra investeringen vereist zijn. Voorwaarde (i) de berekende veiligheidsgraad van het bedrijf dient hoger te liggen ~ dan Vx ; Voorwaarde (ii) de verhouding van de gemiddelde hypothetische veiligheidsbaten over de totale veiligheidsinvesteringen dient te liggen tussen 1,5 en 3. Indien minstens één van beide voorwaarden niet vervuld is, kunnen de veiligheidsinvesteringen best worden gewijzigd.
Review of Business and Economics
~
Indien het bedrijf onder de vooropgestelde Vx drempel zit, dienen er bijkomende investeringen te gebeuren om de veiligheidsgraad te doen stijgen. Na het bepalen van de soort en de grootte van de bijkomende investeringen dient ook gereflecteerd te worden over of de nieuwe investeringen de vooropgestelde (hogere) veiligheidsgraad zullen realiseren. Per vastgestelde periode (bv. driejaarlijks) wordt de veiligheidsgraad geëvalueerd. Bij de tweede voorwaarde wordt verondersteld dat de gemiddelde hypothetische baten anderhalf à driemaal hoger dienen te liggen dan de preventiekosten, wat een redelijke factor is naar socio-economische maatstaf. Dit voorgestelde model geeft input aan het bedrijfsmanagement om meer gegronde beslissingenaangaande veiligheidsinvesteringen te maken. De kosten/baten oefening dient bij een herformulering van de veiligheidsinvesteringen naar aanleiding van de resultaten van het model,volledig opnieuw te gebeuren.
III.
Gevalstudie
A. Achtergrondschets
In de toepassing van dit kosten-baten model worden de veiligheidsinvesteringen van een nieuw te plaatsen installatie geëvalueerd. De levensduur van een nieuwe chemische installatie waarbij geen grondige aanpassingen dienen te gebeuren,
Innovatief kosten/baten model: veiligheidsinvesteringen in een chemische onderneming
331
2008 / 3
wordt geschat op 10 tot 20 jaar. In dit artikel veronderstellen we daarom een gemiddelde levensduur van 15 jaar. De levensduur per investering en per hypothetische baat werd in tabel 2, respectievelijk tabel 3, vermeld. Een ‘eenmalige’ investering/baat komt overeen met een investering/baat die geldt voor 15 jaar en wordt dus afgeschreven op 15 jaar. De levensduur van alle investeringen worden in dit model herleid tot drie jaar operatie van de chemische installatie. Drie jaar is namelijk de duur waarmee wordt rekening gehouden voor het bepalen van de gemiddelde frequentiegraad in het FARAO-plan voor het eventueel vastleggen van de preventiecontributie. De discontovoet waarmee werd gewerkt in deze paper, bedraagt 6% op jaarbasis. Het onderzochte bedrijf waarvan de nieuw te plaatsen installatie wordt onderzocht, is een grote multinational met ruime expertise op gebied van arbeids-en procesveiligheid. Het bedrijf heeft op de plaats van vestiging van de onderzochte installatie 110 werknemers. De gemiddelde frequentiegraad van het bedrijf voor de jaren 20042005-2006 bedroeg 2,8. De frequentiegraad van de industriële sector waarvan de onderneming deel uitmaakt, bedroeg voor 2004 10,63. Voor de jaren 2005 en 2006 zijn nog geen gegevens van de gehele industriële sector gekend (Fonds voor Arbeidsongevallen, ES2, 2008). Omwille van strikte confidentialiteit worden geen verdere gegevens over de onderneming in kwestie vermeld.
In sectie II.B wordt een volledig overzicht gegeven van alle geschatte veiligheidskosten voor de deze eenheid. In sectie II.C worden de hypothetische baten als gevolg van verminderde ongevallen bij deze installatie besproken. In sectie II.D worden de kosten en de baten met elkaar afgewogen volgens het voorgestelde theoretische model van sectie II.D.
Review of Business and Economics
De onderzochte installatie binnen het bedrijf is een nieuwe eenheid die geplaatst dient te worden binnen de bestaande bedrijfsterreinen. Een diepgaande uitleg over het proces zoals het wordt uitgevoerd in de onderneming kan om confidentiële redenen niet opgenomen worden in dit artikel en is bovendien ook weinig relevant. De verschillende kosten die voor deze unit van toepassing zijn om de veiligheid binnen de installatie en (dus) ook binnen het gehele bedrijf te garanderen, worden wel besproken. De belangrijkste gevaren die bij het proces ontstaan, zijn het oververhitten van de installatie (sterk exotherme reactie), gevaren voor de werknemers zoals o.a. blootstelling aan chemicaliën en gevaar voor het ontstaan van explosieve atmosferen. Verschillende fysische metingen zijn belangrijke indicatoren om na te gaan of het proces optimaal verloopt. Wanneer vooropgestelde limieten overschreden worden, dient de installatie automatisch te worden uitgeschakeld. Op basis van twee studies werden alle mogelijke rampscenario’s onderzocht. Naast de wettelijk verplichte procesveiligheidsstudie, werd tevens een HAZOP analyse10 uitgevoerd. Deze HAZOP studie houdt niet alleen rekening met het proces zelf maar ook met andere invloedsfactoren.
2008 / 3
332
G.L.L. Reniers & A. Audenaert
B. Veiligheidskosten voor de onderzochte installatie
Analoog als in sectie II.B waar de veiligheidskosten theoretisch besproken werden, komen deze veiligheidskosten in dit praktijkstuk aan bod. Vooreerst vormen de conceptuele ontwerpkosten opgeteld bij de procesuitrustingskosten, samen de procesveiligheidskosten. Daarna worden de brand-en explosieveiligheidskosten geëvalueerd. De operationele kosten zijn vergelijkbaar met de kosten die onder de rubriek arbeidsveiligheid besproken werden. Ten slotte worden de transportveiligheid-en contractorveiligheidskosten in kaart gebracht voor de eenheid.
1. Procesveiligheidskosten
Review of Business and Economics
a. Conceptuele ontwerpkosten Vooraleer de installatie daadwerkelijk gebouwd kan worden, dienen een aantal risicoanalyses opgemaakt te worden. Dit kan ondergebracht worden onder de term conceptuele ontwerpkosten. Deze conceptuele ontwerpkosten zijn vaste en indirecte kosten. Ze worden éénmalig gemaakt (tenzij een aanpassing van de installatie nodig is, waardoor nieuwe ontwerpkosten nodig zijn) en veranderen niet naargelang de activiteit op de installatie. Ze staan ook niet rechtstreeks in relatie met de productiekosten. Deze conceptuele fase is nodig om na te gaan of het project haalbaar is en welke veiligheidsmaatregelen en -inspanningen daarvoor nodig zijn. Om dit te weten is eerst en vooral een studie nodig om de mogelijke risico’s en gevaren die het proces met zich meebrengt, te identificeren. De procesveiligheidsstudie onderzoekt welke risico’s gepaard gaan met het proces zelf. Factoren zoals temperatuursafhankelijkheid, gasvorming, drukvereisten en -limieten en mengsels zullen op een correcte wijze onder controle moeten worden gehouden om het risico op escalatie van het proces, en dus de installatie, zo klein mogelijk te houden. Escalatie van het proces zou tot een zwaar ongeval kunnen leiden. Onder conceptuele ontwerpkosten valt eveneens de mogelijke vertraging omwille van kwaliteit-, welzijn-, en milieuvereisten en ook het zoeken en ter beschikking stellen van competent en gekwalificeerd personeel. Het volledige proces moet geanalyseerd worden waarbij de onderzoeksmedewerkers, de operatoren, de preventieadviseurs, de explosiespecialisten en de procesveiligheidsspecialisten betrokken dienen te zijn. Er worden t.o.v. de wettelijke minimumeisen extra maatregelen genomen door het bedrijf om het conceptuele ontwerp te realiseren. Enkele voorbeelden zijn bijkomstige overdrukventielen, sturingaanpassingen aan pompen en verhoogde afzuigdebieten. De uitbatingsvergunningskost (permit cost), is de kost om de eenheid te kunnen uitbaten binnen de wettelijke richtlijnen. Deze kost bestaat uit de studies en aanpassingen nodig om binnen de wettelijke emissies te blijven en uit de opmaak van bouw- en milieuvergunningen. Het uitwerken van een omgevingsveiligheidsrapport besluit de conceptuele fase. Dit rapport maakt een studie van de mogelijke risico’s (opstellen van iso-risicocontouren) die kunnen ontstaan op de perceelsgrens van de uitbater. Deze studie berekent
Innovatief kosten/baten model: veiligheidsinvesteringen in een chemische onderneming
333
2008 / 3
de restrisico’s buiten de perceelsgrenzen van het bedrijf voor de omwonenden en eventueel zwakkere groepen (rusthuizen, kinderopvangcentra, scholen). Voor de eenheid dient bijgevolg extra aandacht besteed te worden aan de analyse van de gevaren die sommige chemische stoffen met zich meebrengen. De opslag gebeurt in een dichtbevolkte woonzone die het bedrijf omsluiten. Potentiële risico’s op rampen met de gevaarlijke stoffen en de daarbij gepaard gaande effecten op de omgeving dienen geanalyseerd en geëvalueerd te worden.
b. Procesuitrustingskosten
2. Brand-en explosiekosten
De installatie zal gebouwd worden in een bestaande ruimte op het complex van het bedrijf. Hierdoor zullen een aantal veiligheidsaanpassingen aan de ruimte en het materiaal moeten worden uitgevoerd. De kosten met betrekking tot deze aanpassing kunnen voornamelijk beschouwd worden ter bestrijding en vermijding van brand en explosie. De installatie dient uitgerust te worden met een volledig sprinklersysteem met meer dan voldoende watervoorraad en met voldoende detectoren. De installatie wordt uitgerust met gasdetectoren om gevaarlijke gassen tijdig te kunnen detecteren en ook om de nodige detectie te verrichten teneinde het risico om een explosieve atmosfeer tijdig te kunnen opsporen. De ruimte is ook voorzien van rookdetectoren, alarmen en
Review of Business and Economics
Eenmaal het ontwerp van de installatie voldoende veiligheidsgaranties biedt, kan overgegaan worden tot het opzetten van specifieke procesvereisten. De bovenvermelde risicoanalyses en veiligheidsstudies geven aan welke maatregelen er moeten worden getroffen en welke materiële veiligheidsvereisten aan de installatie dienen te worden opgelegd. Hoewel de installatie dan naar wettelijke en industriële normen al veilig geacht wordt, kiest dit bedrijf ervoor om extra materiële beperkingen op te leggen aan de eenheid en aan het proces om alles zo veilig mogelijk te laten verlopen. De kosten om dit te realiseren kunnen aangeduid worden als bijkomende procesuitrustingskosten. In dit artikel worden enkel de bijkomende procesuitrustingskosten naar aanleiding van veiligheidsaspecten bedoeld, dus bv. niet de kosten die louter dienen om het proces te monitoren zonder effect op de procesveiligheid. Deze kosten zijn eveneens eenmalige investeringen en zullen niet variëren naargelang het productieniveau, maar staan wel rechtstreeks in verband met het proces. Deze procesuitrustingskosten zijn dan ook vaste, directe kosten. De bijkomende uitrusting heeft betrekking op drukmeters, thermometers, hydrometers, overdrukbeschermingspompen, etc. Daarnaast zijn bijkomstige noodstoppen op praktische plaatsen op de installatie gemonteerd om in geval van een noodsituatie de installatie volledig te kunnen stilleggen. Om de processpecifieke factoren zoals de temperatuur, druk, vochtigheid, etc. te kunnen beoordelen naar hun kritieke eigenschappen zijn computergestuurde programma’s en controledisplays ontwikkeld om de procesveiligheid niet in het gedrang te brengen en zo geen aanleiding te geven tot een zwaar ongeval met grote schade.
2008 / 3
334
G.L.L. Reniers & A. Audenaert
een sprinklerinstallatie. Het sprinklerwater dient omwille van wettelijke milieurichtlijnen ook opgevangen te kunnen worden, waardoor in de ruimte opvangvoorzieningen dienen aangebracht te worden. Bovendien worden er binnen het veiligheidsprogramma van het bedrijf standaard collectieve beschermingsmaatregelen voorzien zoals oogwasfonteinen en nooddouches. Daarnaast is een zoneringstudie en -rapport vereist om na te gaan op welke plaatsen en momenten tijdens het proces explosieve atmosferen kunnen ontstaan of vrijkomen, teneinde maatregelen te nemen door onderdelen aan te passen en beperkingen op te leggen aan de operatoren.
3. Operationele kosten
Review of Business and Economics
De operationele kosten staan in rechtstreeks verband met het productieniveau en variëren naargelang de intensiteit van het gebruik van de installatie. De voornaamste kosten voor deze installatie zijn het nazicht en onderhoud van de procesveiligheidsvoorzieningen zoals overdrukventielen, gasdetectoren, noodstoppen, etc. Wegens de vele veiligheidsinspanningen zal het opstarten van de installatie 20% langer duren, wat zich vertaalt in een extra productiekost. Daarnaast dient ook de opslag van gevaarlijk materiaal verrekend te worden. De milieuvereisten stellen dat op geen enkel ogenblik iets uit de tanks mag lekken. Bovendien dient er ook heel wat tijd besteed te worden aan het schrijven en periodiek nakijken van operationele richtlijnen en het geven van (her)training aan de operatoren en onderzoekers.
4. Transport veiligheidskosten
Er wordt veel belang gehecht aan risico’s buiten de ondernemingsgrenzen (perceelsgrenzen) van de onderneming. Een belangrijke factor hier is het aan-en afvoeren van de gasflessen met gevaarlijke stoffen, die een variabele en indirecte kost vertegenwoordigt. Bovendien is het laden en ontladen ook belangrijk, daar statistisch gezien de meeste incidenten gebeuren tijdens dit proces. De leveranciers dienen strikte wettelijke vereisten na te leven tijdens het transport (ADR18-wetgeving). Ze dienen ook procedures te volgen voor het ontladen en opslaan van deze toxische gassen. Dit heeft een rechtstreekse impact op het uitbaten van de installatie. Daarnaast dienen de verantwoordelijken controlelijsten op te stellen en de veiligheidsdocumenten nauwkeurig in te vullen.
5. Contractorveiligheidskosten
In contractorveiligheid kostenpost bestaande uit variabele indirecte kosten, worden de kosten vermeld die nodig zijn om contractorfirma’s te selecteren voor veiligheid. Indien de contractorfirma’s weerhouden worden, dienen de medewerkers nog een veiligheidstraining te krijgen, te slagen in een kwalificatie en daarna de nodige instructies te krijgen voor het veilig werken aan de installaties. Dit geheel aan uit te
Innovatief kosten/baten model: veiligheidsinvesteringen in een chemische onderneming
335
2008 / 3
voeren taken maakt dat er zowel voor het bedrijf als voor de onderneming van buitenaf heel wat extra kosten voor veiligheid bijkomen voor het opstellen van de installatie. Het is wel zo dat in praktijk dikwijls met dezelfde contractoren wordt gewerkt en er sprake is van een langetermijnbeleid.
6. Samenvattende tabel voor kostenposten
Tabel 2 geeft een overzicht van alle kostenposten voor zware ongevallen en voor nietzware ongevallen en maakt een totaalsom voor de veiligheidsinvesteringen voor de nieuw te plaatsen eenheid. Bij de toewijzing tot welke categorie een kost behoort (‘zware ongevalsinvesteringen’ of ‘niet-zware ongevalsinvesteringen’) hebben we (conform de bedrijfsstrategie) verondersteld dat de kosten in eerste instantie gemaakt worden om zware ongevallen te vermijden (sommige van deze investeringen kunnen uiteraard ook leiden tot het vermijden van niet-zware ongevallen, maar bedrijven passen zulke maatregelen in eerste instantie toe i.v.m. de preventie van zware ongevallen). Bovenop de zware ongevalinvesteringen gebeuren er door het bedrijf (vanzelfsprekend) ook aanvullende investeringen om niet-zware ongevallen te vermijden (deze investeringen leiden evenwel nooit tot het vermijden van zware ongevallen). Veiligheidsinvesteringen worden dus ofwel als kosten voor de preventie van zware ongevallen beschouwd, ofwel als kosten voor de preventie van niet-zware ongevallen. Dit reflecteert zich in Tabel 2 in het feit dat de subtotalen van de twee soorten kosten mogen opgeteld worden tot een totale veiligheidsinvesteringskost.
Veiligheidskostenposten
Bedrag van de investering (de levensduur van de kostenposten werd herrekend op een operatie van de installatie gedurende drie jaar) (€) Niet-zware ongevalsinvesteringen
Zware ongevalsinvesteringen
Procesveiligheidskosten (originele levensduur van de kostenpost) Conceptuele ontwerpkosten Procesveiligheidsstudie (eenmalig) Atex zoneringstudie en -rapport (eenmalig) Kosten van zoeken van gekwalificeerd personeel en middelen (eenmalig) Team veiligheidsinput ter ontwikkeling (operatoren, preventieadviseurs, seminaries, reisonkosten, etc.) (eenmalig)
8420,5 2074,9 419,3 3834,4
Review of Business and Economics
Tabel 2. Veiligheidsinvesteringen voor de nieuwe eenheid (cijfers op basis van interne bedrijfsdocumenten).
2008 / 3
336
G.L.L. Reniers & A. Audenaert
Veiligheidskostenposten
Bedrag van de investering (de levensduur van de kostenposten werd herrekend op een operatie van de installatie gedurende drie jaar) (€) Niet-zware ongevalsinvesteringen
Kost van wettelijke vereisten bovenop algemene industriestandaarden (eenmalig) Uitbatingsvergunning (eenmalig) Uitwerken van omgevingsveiligheidsrapport (eenmalig)
2636,2 5991,3 7189,6
Procesuitrustingskosten Design veiligheidskosten (drukmeters, thermo- en hydrometers, overdrukbeschermingspompen, etc.) (eenmalig) Noodstoppen en automatische afsluitsystemen (eenmalig) Visuele inspectiesystemen (niveau-indicatoren, displays, etc.) (eenmalig)
Review of Business and Economics
Brand- en explosiekosten Gasdetectiesystemen (eenmalig) Brandbestrijdingsmiddelen (sprinklers, brandblussers, etc.) (eenmalig) Collectie- of opvangvoorzieningen (eenmalig) Collectieve beschermingssystemen (oogwasfonteinen, nooddouches, etc.) (eenmalig) Operationele kosten Onderhoud veiligheidstoestellen (jaarlijks) Additionele kost bij het opstarten/stilleggen (5-jaarlijks) Operationele veiligheidsprocedures (jaarlijks) Opslag van gevaarlijk materiaal en substanties (eenmalig) Noodsystemen (eenmalig)
10784,5 757,3 4638,7
6591,9 4074,1 2396,5 608,7
4272,7 1003,6 566,6 4888,9 2324,6
Transportveiligheidskosten Verplaatsing flessen toxische gassen (jaarlijks) Opstellen controlelijsten (eenmalig) Veiligheidsdocumenten (jaarlijks)
5100,1 455,3 2125
Contractorveiligheidskosten Selectieproces veilige contractoren (eenmalig) Training contractoren (eenmalig) Verlies aan werktijd door training (eenmalig)
23,9 89,8 89,8
Subtotalen voor niet-zware ongevallen en voor zware ongevallen Totaal bedrag voor veiligheidsinvesteringen:
Zware ongevalsinvesteringen
18.450,6 €
62.907,6 €
81.358,2 €
Innovatief kosten/baten model: veiligheidsinvesteringen in een chemische onderneming
337
In deze sectie worden de verschillende veiligheidsbatenposten bepaald voor het zich niet voordoen van ongevallen bij de eenheid.
2008 / 3
C. Hypothetische baten voor de installatie
1. Baten m.b.t. het ontwerp van de installatie
2. Operationele baten
Een onderneming is geneigd enkel de directe kosten, d.w.z. de loonkosten van de werknemer, van absenteïsme in rekening te brengen. Nochtans is het duidelijk gebleken uit sectie II.C dat de baten die een onderneming ondervindt dankzij niet-absenteïsme hoofdzakelijk verborgen indirecte baten zijn. Deze baten vertegenwoordigen een aanzienlijk groter deel dan de directe baten. Volgens de Nederlandse Gezondheidsraad komen de totale baten van niet-afwezigheid per werknemer gemiddeld op 400 € per dag. Dit bedrag omvat de sociale zekerheidsbijdragen, de bijkomende administratieve en managementkosten alsook de baten van de medische dienst en andere personeelsbaten. Het veiligheidsbeleid betreffende de eenheid
Review of Business and Economics
Doordat er in de conceptuele fase van de eenheid veel aandacht besteed werd aan de identificatie van mogelijke risico’s, zijn het ontwerp en de systeemvereisten van het proces zeer goed gekend. Deze kennis bevordert de productiviteit van de installatie omdat de werknemers op elk moment weten wat de installatie aankan en wat niet. Zonder de kennis van de risicofactoren zou de installatie bij bepaalde omstandigheden onnodig stilgelegd kunnen worden omdat de werknemers de installatie onterecht als onveilig zouden beschouwen. In het andere geval zouden de werknemers bij gebrek aan risicoanalyses de installatie laten opereren terwijl ze misschien een potentieel risico zou vormen op incident, mogelijk escalerend tot een zwaar ongeval met beduidende schade. Het tot in de puntjes kennen van de risicofactoren van de installatie zal dus leiden tot grote operationele activiteit. Gemiddeld zal de operatie 5% efficiënter werken (gebaseerd op proces engineering design studies), wat dus naar personeelskost een verhoging van 5% efficiëntie of 30 minuten per operator per werkdag met zich meebrengt. De verhoging in efficiëntie brengt een opbrengst van 3125 € per jaar met zich mee. Bovendien zijn de extra inspanningen om het ontwerp zo veilig mogelijk te maken terug te verdienen in verminderde inspectiekosten. Door het veilige ontwerp zal de installatie op tijd aanduiden wanneer een mogelijk probleem zich voordoet. Bovendien werd getracht inspecties zo visueel mogelijk te maken. Dit heeft als grote voordeel dat de installatie niet regelmatig zal stilgelegd moeten worden om er een controle op uit te voeren. Deze controle wordt uitgevoerd door de gekwalificeerde veiligheidsinspecteurs en door deze preventieve manier van werken kan de inspectietijd met 50% gereduceerd worden, wat overeenkomt met 754 € aan extra opbrengsten per jaar.
Review of Business and Economics
2008 / 3
338
G.L.L. Reniers & A. Audenaert
zorgt ervoor dat op korte termijn (stel afwezigheidreductie van 20 dagen per jaar) de baten 8.000 € bedragen. Op lange termijn (stel afwezigheidsreductie van 6 maanden of 120 dagen) loopt deze opbrengst op tot 48.000 €. De baten door niet-ongevallen zijn gelijkaardig aan de indirecte baten van niet-absenteïsme. Daarnaast kan een bijkomende besparing op niet-verlies aan eigendom gegenereerd worden. Potentiële materiële schade vormt een grote batenpost aangezien de eenheid uit een groot aantal vaten, motoren, pompen, warmtewisselaars etc. bestaat. Er kan schade zijn aan de machines zelf en/of er kan een grondstofverlies zijn. In het extreme geval van een zwaar ongeval is er zelfs schade aan de inrichting of gebouw waardoor alle herstellingskosten in acht moeten genomen worden. De totale infrastructuur wordt op 300.000 € geschat. Het vermijden van verliezen m.b.t. infrastructuur bedraagt makkelijk 10% bij (vermeden) zware ongevallen, en 1% bij (vermeden) niet-zware ongevallen. Materiële schade die niet plaatsvindt, heeft bovendien een impact op de werknemer die al dan niet gewond zou worden door het ongeval. Hierdoor moet niet-tijdverlies van het slachtoffer aan de onderneming gerekend worden. Logischerwijze beïnvloedt dit tevens de werktijd van medewerkers. Zij zouden immers bij een ongeval hulp dienen te bieden aan het slachtoffer en meehelpen aan het opstellen van een ongevallenverslag. In sommige gevallen zou de werkplaats gereorganiseerd moeten worden. Het opsporen van de ongevallenoorzaak of het verrichten van een incidentenanalyse vergt ook tijd en onderzoek van het veiligheidsteam. Voornaamste tijdsbesteding is het opstellen van documenten en analyseren van de gegevens. Bij de onderneming die we beschouwen worden ongevalonderzoeken, hoe klein ook, uitgevoerd door de leden van de preventiedienst. De gemiddelde duur van een onderzoek inclusief verslaggeving bedraagt 1 uur voor alle betrokkenen. Bij ernstige arbeidsongevallen dient een volledig onderzoeksteam opgesteld te worden. Dit team bestaat uit de preventieadviseurs, de hiërarchische lijn, de verantwoordelijke van de installatie en externe deskundigen. Een onderzoeksverslag dient opgemaakt en verstuurd te worden naar de Federale Overheidsdienst Toezicht op het Welzijn op het Werk. De kosten van een externe deskundige bedraagt rond de 1.000 € per dag. De interne personeelskosten worden geschat op 825 € per dag. Door de implementatie van de veiligheidsmaatregelen is een reductie van 6 onderzoeksdagen per jaar voor ernstige ongevallen geschat. Deze opbrengst komt neer op 10.950 €. Voor niet-zware ongevallen kan één onderzoeksdag worden aangerekend, hetzij 1.825 €. Al veroorzaakt een ongeval geen materiële schade, toch zal het een productieverlies veroorzaken. Een gewonde werknemer zorgt voor een vermindering aan productiviteit en een beschadigde machine zorgt voor outputverlies. Het is duidelijk dat wanneer de machine niet meer operatief is, de kosten hoog zouden oplopen. Bij explosie of brand loopt de productie ongeveer 1 jaar vertraging. Een kwantificering van deze gevolgen is vrijwel onmogelijk, maar het is duidelijk dat de maatregelen om de veiligheid te verzekeren van enorm belang zijn en enorme baten kunnen vertegenwoor-
Innovatief kosten/baten model: veiligheidsinvesteringen in een chemische onderneming
339
2008 / 3
digen. Een andere hypothetische opbrengst die de onderneming kan realiseren door veilig te werken en het aantal ongevallen te beperken is via een verlaging van de verzekeringspremie. Het jaar waarin de studie plaatsvond (2007) werd er bv. een aanzienlijke daling van de premie vastgesteld. Indien er een ernstig incident zou gebeuren met de nieuwe installatie, zou deze premie de volgende drie jaar vastgelegd worden aan een veel hoger bedrag. De baten als gevolg van vermeden ongevallen worden door ondernemingsexperten (voorzichtig) geschat op minimum 25.000 € bij zware ongevallen en 7.500 € bij niet-zware ongevallen.
3. Baten i.v.m. niet-personeelsverloop
Wanneer de omstandigheden binnen de installatie ongezond en onveilig geacht zouden worden door de werknemers, bestaat de mogelijkheid tot personeelsverloop. Managementbaten die gepaard gaan met niet-personeelsverloop worden geschat op 1.000 €. Op korte termijn zou er een bijkomende hypothetische baat gerealiseerd worden van 5.000 €. Bij personeelsverloop kunnen vervangingen intern en extern gebeuren. De grootste kost is indien de medewerker extern dient gezocht te worden. Het zoeken naar een nieuwe medewerker bedraagt ongeveer 1.600 € per kandidaat. Door de noodzakelijke opleiding van de nieuwe operator wordt een verlies aan productiviteit veroorzaakt (werknemer eerste zes maanden intensieve training: 12.500 € aan loonkosten) en zal de tijd van de coach en het opstellen van de cursussen moeten meegerekend worden, wat ongeveer 1 dag per week en dit gedurende zes maanden bedraagt (= 3.600 €).
In de onderneming waar de eenheid wordt geplaatst vormt veiligheid een inherent onderdeel van het beleid. Het geïntegreerd veiligheidsbeleid is zo nauw verbonden met de strategie en de activiteiten van de onderneming dat het makkelijk aangepast en gestuurd kan worden indien omgevingsfactoren veranderen. Indien de organisatie bij ieder incident telkens opnieuw een nieuw veiligheidsbeleid zou moeten uitstippelen, zou de kost van extra training van personeel op meer dan 36.667 € (110 werknemers aan halve dag per maand) komen. Gezien de grote impact op de volledige organisatie, vertegenwoordigt de flexibiliteit van de onderneming een grote opbrengst. Voor factoren zoals vermeden productieverlies, gunstig imago, motivatie en tevredenheid van de werknemers, is een kwantificatie zeer moeilijk. Nochtans kunnen de baten als gevolg van deze factoren naargelang de ernst van een ongeval aanzienlijk zijn. Het bedrijf in kwestie is zich hiervan bewust en treft alle maatregelen om zware incidenten zoveel mogelijk te vermijden. De kosten die met deze factoren gepaard gaan kunnen immers aangroeien tot enorme proporties voor de onderneming. Met voorbeelden van rampen als Bhopal in het achterhoofd (zie inleiding) is het zelfs niet onwaarschijnlijk dat de impact zo groot zou kunnen zijn
Review of Business and Economics
4. Baten n.a.v. de impact op de veranderende omgeving
2008 / 3
340
G.L.L. Reniers & A. Audenaert
dat de moederonderneming in financiële moeilijkheden komt. Bij de meest recente ramp in de VS, de ontploffing van een raffinaderij in Texas City in 2005, heeft BP bijvoorbeeld een zeer groot imagoverlies geleden. De baten in deze categorie worden niet geschat, maar er wordt bij de eindanalyse van de kosten/baten studie wel rekening mee gehouden.
5. Samenvattende tabel voor hypothetische batenposten
Review of Business and Economics
Tabel 3 geeft een overzicht van alle batenposten voor zware ongevallen en nietzware ongevallen en maakt een totaalsom voor de gemiddelde hypothetische baten voor de eenheid. In het geval van het bepalen van de totale hypothetische batenposten dienen we er rekening mee te houden dat sommige baten zowel voor zware ongevallen als voor niet-zware ongevallen worden meegeteld. We mogen dus niet zomaar de baten n.a.v. zware ongevallen en niet-zware ongevallen sommeren, anders krijgen we dubbeltellingen. Vermits het ontwerp van de installatie in het bedrijf volgens algemene standaarden gebeurt en onafhankelijk is van het soort ongeval dat zich zou voordoen, zijn de hypothetische baten voor deze post dezelfde voor de twee soorten ongevallen. Ook bij het vermijden van personeelsverloop kan men voor het bedrijf in kwestie stellen dat de sfeer op het werk en de werkomstandigheden geacht worden niet of nauwelijks afhankelijk te zijn van de preventie-inspanningen, en dat de gerealiseerde baten voor de twee types ongevallen dus ongeveer dezelfde zijn. Ook het geïntegreerd zijn van het veiligheidsbeleid in de onderneming en de ermee gepaard gaande hypothetische baten hangen niet af van het zich eventueel voordoen van zware ongevallen of niet-zware ongevallen. De hypothetische operationele baten kunnen bij beide soorten ongevallen dan weer wel aanzienlijk verschillen: het zich voordoen van een zwaar ongeval zal een aanzienlijk grotere impact hebben op de voortgang van de operaties dan wanneer een niet-zwaar ongeval plaatsvindt. Het is natuurlijk zo dat de kans op een niet-zwaar ongeval vele malen groter is dan de kans op een zwaar ongeval. We zouden dus kunnen stellen dat we enkel rekening dienen te houden met de baten van niet-zware ongevallen wat de kansen betreft. Echter, de gevolgen van een zwaar ongeval kunnen veel aanzienlijker zijn dan deze van een niet-zwaar ongeval. De baten voor het vermijden van een zwaar ongeval zijn dus vele malen hoger voor wat de gevolgen betreft. Klassiek wordt in risicomanagement aan kans en gevolgen voor risico’s een evenredig belang toegekend (Claes, 2004), vandaar de keuze voor een 50/50-verdeling voor de baten van zware ongevallen en niet-zware ongevallen en de uiteindelijke berekening van een gemiddelde voor de totale hypothetische baten. Er dient nogmaals benadrukt te worden dat de (hypothetische) opbrengst die door de veiligheidsinvestering gerealiseerd is, niet tastbaar is en dus ook niet terug te vin-
Innovatief kosten/baten model: veiligheidsinvesteringen in een chemische onderneming
341
2008 / 3
den is in boekhouding van de onderneming. De hypothetische baten zijn niet reëel, maar de kosten die ermee overeenkomen worden wel werkelijkheid indien er zich incidenten zouden voordoen met de installatie en/of haar operatoren.
Tabel 3. Batenfactoren voor de eenheid (cijfers op basis van interne bedrijfsdocumenten). Batenfactoren
Bedrag van de hypothetische baat (de levensduur van de batenposten werd herrekend op een operatie van de installatie gedurende drie jaar) (€) Niet-zware ongevalsbaten
Zware ongevalsbaten
1497,8
1497,8
361,4
361,4
22667,1
136002,8
8500,1
85001,7
Ontwerp van de installatie Vermindering in stilstand dankzij conceptuele en ontwerpfase (eenmalig) Verminderde inspectiekost (eenmalig) Operationele baten Niet-absenteïsme (KT: 8.000 €; LT: 48.000 €) (jaarlijks) directe en indirecte vermeden kosten aangaande infrastructuur (jaarlijks) vermeden kosten (allerlei) (jaarlijks)
5170,9
31025,6
21250,4
70834,8
Niet-personeelsverloop Management-en lijnmanagementbaten (eenmalig)
479,3
479,3
KT-impact (eenmalig)
2396,5
2396,5
Rekruteringsbaten (2 werknemers op de nieuwe eenheid) (eenmalig)
1533,7
1533,7
Aanleren professionele vaardigheden (eenmalig)
5991,3
5991,3
Baat trainingsoperator (eenmalig)
1725,5
1725,5
17574,9
17574,9
89.148,9 €
354.425,3 €
Impact op veranderende omgeving Werkuren veiligheids-en gezondheidspersoneel, trainingen veiligheidspersoneel, werkuren management en lijnmanagement, baat van medische dienst (capaciteit), onderzoek en metingen, baten van externe diensten, baten van training en ontwikkeling, organisatiekosten (eenmalig) Subtotalen voor niet-zware ongevallen en voor zware ongevallen Totaal gemiddeld bedrag voor veiligheidsbaten:
221.787,1 €
Review of Business and Economics
vermeden ongevalsonderzoekskosten (jaarlijks)
Review of Business and Economics
2008 / 3
342
G.L.L. Reniers & A. Audenaert
D. Kosten/baten analyse voor de nieuwe eenheid
De kosten/baten analyse leert dat de veiligheidskost van de investeringen voor het vermijden van niet-zware ongevallen 18.450,6 € bedraagt. De baten voor vermeden niet-zware ongevallen bedragen 89.148,9 €. Bij zware ongevallen bedragen de veiligheidsinvesteringen 62.907,6 € en de veiligheidsbaten 345.425,3 € in het geval van de beschouwde installatie. Voor beide types van ongevallen liggen de hypothetische baten dus aanzienlijk hoger dan de reële kosten aangaande veiligheidsinvesteringen. Sommige baten worden echter tweemaal geteld: eenmaal voor het vermijden van zware ongevallen en eenmaal voor het vermijden van niet-zware ongevallen. Vandaar de bepaling van de gemiddelde baten als gevolg van de veiligheidsinvesteringen. Rekening houdend met dit cijfer, dient er opgemerkt te worden dat de totale veiligheidsinvesteringen (81.358,2 €) eveneens aanzienlijk lager liggen dan het gemiddelde bedrag van veiligheidsbaten (221.787,1 €). Dit is voornamelijk te wijten aan preventie-investeringen voor zware ongevallen. De hypothetische baten n.a.v. vermeden niet-zware ongevallen (89.148,9 €) liggen immers slechts iets hoger dan de totale veiligheidsinvesteringen (81.358,2 €). Hieruit kunnen we concluderen dat de investeringen voornamelijk opbrengen (i.e., hypothetische baten met zich meebrengen) ingeval vermeden zware ongevallen. Indien zich de eerste drie jaren van operatie geen zwaar ongeval voordoet, hebben we een theoretische netto-opbrengst van ongeveer 140.429 € (221.787 € – 81.358 €). Indien er zich wel een zwaar ongeval zou manifesteren, zou dit een reële netto-kost betekenen van ongeveer 354.425 €. Door de investeringen in de preventie van zware ongevallen hebben we door een zwaar ongeval te vermijden dus een hypothetische baat van 494.854 € gerealiseerd voor de eerste drie jaren van operatie van de installatie. Indien baten zoals publiek imago, motivatie en tevredenheid van de werknemers en vermeden productieverlies mee beschouwd zouden worden in de analyse, slaat de balans al gauw nog veel sterker door naar de baten. We gaan nu aan de hand van het voorgestelde model na of we binnen de ‘goede zone’ zitten qua investering en of we de grootte van de veiligheidsinvesteringen niet dienen bij te sturen.11 We bepalen hiervoor eerst de veiligheidsgraad van de onderneming. Deze bedraagt 0,8566. Het bedrijf in kwestie heeft zich tot doel gesteld dit cijfer nog verder te verhogen, maar stelt dat het streefcijfer t.o.v. het gemiddelde cijfer van de industriële sector, gehaald is. De eerste (bedrijfsbeleidafhankelijke) voorwaarde voor het mo~ del is dus vervuld (we zitten boven Vx ). Voorts stellen we vast dat de verhouding van de gemiddelde hypothetische baten (221.787,1 €) t.o.v. de totale veiligheidskosten (81.358,2 €) 2,73 bedraagt en dus binnen het vooropgestelde interval [1,5;3] ligt. Er dient bijgevolg geen verdere studie te gebeuren naar een verdere optimalisatie van de veiligheidsinvesteringen.
Innovatief kosten/baten model: veiligheidsinvesteringen in een chemische onderneming
343
IV.
2008 / 3
Men kan opmerken dat t.o.v. de gemiddelde industriële veiligheidsgraad zeer veilig gewerkt wordt in de onderneming waar de installatie zal worden geplaatst. De totale veiligheidsinvesteringen liggen lager dan de kwantificeerbare hypothetische baten. Indien rekening wordt gehouden met de niet-kwantificeerbare baten zullen de gemiddelde baten veel hoger liggen dan de totale kosten van veiligheidsmaatregelen. Hieruit kunnen we besluiten dat de veiligheidsinvestering in casu dient te worden aanbevolen op basis van de kosten/baten berekening. Er wordt noch overgeïnvesteerd (wat economisch niet aanvaardbaar zou zijn voor het bedrijf), noch ondergeïnvesteerd (wat sociaal niet aanvaardbaar zou zijn voor de maatschappij).
Besluiten
Met de uitwerking van een algemeen theoretisch kosten/baten analysemodel opgesplitst in zware ongevallen en niet-zware ongevallen en waarbij de veiligheidsinvesteringen worden afgewogen tegen de eruit voortvloeiende (hypothetische) baten, wordt in dit artikel een innovatieve methode ontwikkeld om de baten van veiligheid beter te kunnen begrijpen en om deze baten ook te kunnen kwantificeren. De voorgestelde methodologie is gebruiksvriendelijk en door ondernemingen eenvoudig praktisch toepasbaar om beslissingen aangaande preventie-investeringen te optimaliseren. Het model werd onderzocht aan de hand van een praktijkstudie waarbij een kosten/baten analyse werd uitgevoerd op een nieuwe installatie met verschillende risico’s. Uit de totale te kwantificeren baten mag men besluiten dat de voorgestelde veiligheidsinvesteringen een hoog rendement vertonen, zowel naar zware ongevallen als naar niet-zware ongevallen toe. Het is ook duidelijk dat zodra er gevaren (bv. toxiciteit, brand, explosie, etc.) zijn die zeer ernstige risico’s met zich mee-
Review of Business and Economics
De investering in veiligheid behelst verschillende onderdelen in een organisatie. De afweging van kosten t.o.v. baten naar aanleiding van de procesveiligheid, de branden explosieveiligheid, de transportveiligheid, de contractorveiligheid en de arbeidsveiligheid moet geanalyseerd worden om een veiligheidsgraad te bereiken waarbij de kans op zowel zware als niet-zware ongevallen tot een minimum beperkt wordt. Veiligheid heeft zijn prijs, maar de gevolgen die kunnen voortvloeien uit onveiligheid vertonen eveneens een kostenplaatje. Het hebben van niet-ongevallen veroorzaakt immers het hebben van niet-kosten. Het vermijden van ongevallen kan dus beschouwd worden als het bereiken van hypothetische baten. Naast de baten van niet-ongevallen, zijn er ook baten van niet-afwezigheid of niet-absenteïsme, baten van niet-personeelsverloop en baten van niet te moeten investeren in een veranderende werksfeer. De grootte van deze baten is vaak aanzienlijker dan bedrijven op het eerste gezicht vermoeden.
2008 / 3
344
G.L.L. Reniers & A. Audenaert
brengen, de hypothetische baten van het niet hebben van de kosten van zeer grote schade (financieel, intern en extern imago, etc.) het sowieso toelaten de noodzakelijke investeringen te doen op procesveiligheidsgebied.
ACKNOWLEDGEMENTS
We willen in eerste instantie graag onze dank betuigen aan een inspecteur van de Federale Overheidsdienst Werkgelegenheid, Arbeid en Sociaal Overleg voor het kritische nalezen van dit artikel en de ermee gepaard gaande opmerkingen. Voorts willen we ook de preventiedienst van het samenwerkende bedrijf bedanken voor de informatie en de bereidwillige medewerking bij het totstandkomen van deze paper. Tot slot danken we de anonieme reviewer voor de vele nuttige opmerkingen en suggesties die de leesbaarheid en het begrip van dit artikel aanzienlijk hebben vergroot.
Review of Business and Economics
NOTEN 1. ARGoSS, Antwerp Research Group on Safety and Security, Universiteit Antwerpen, Prinsstraat 13, 2000 Antwerpen, tel: +32-(0)3-220.41.82, e-mail:
[email protected] b CEDON, Centrum voor Duurzaam Ondernemen, Hogeschool-Universiteit Brussel, Stormstraat 2, 1000 Brussel, tel: +32-(0)2.609.18.07. 2. Met preventieve maatregelen worden maatregelen bedoeld die aangewend worden teneinde een (hypothetisch) ongeval vermijden (en dus ook de kosten ervan te vermijden). 3. Hypothetische baten zijn baten als gevolg van het zich niet voordoen van een ongeval en bijgevolg het vermijden van de ermee gepaard gaande kosten. 4. FARAO staat voor ‘Federaal Actieplan voor de Reductie van Arbeid Ongevallen’. 5. Aangaande de bijdrage dienen volgende punten nog verder gespecifieerd te worden door de wetgever: (i) de frequentie, de ernst, de grens en de exacte duurtijd van de observatieperiode, (ii) de berekening, de periode van aanrekening en de toepassingsmodaliteiten van de forfaitaire preventiecontributie, alsook het maximum-en minimumbedrag ervan, (iii) voor welke ondernemingen of bedrijfssectoren de forfaitaire preventiecontributie kan geïnd worden, (iv) de nadere regels van vaststelling en kennisgeving aan de verzekeringsonderneming door het Fonds voor Arbeidsongevallen en van rapportering aan het Fonds voor Arbeidsongevallen, en (v) de nadere regels van kennisgeving aan de werkgever. 6. De frequentiegraad wordt gedefinieerd als het aantal ongevallen van een groep werknemers (bv van één bedrijf) in functie van het totaal aantal gepresteerde uren door deze werknemers gedurende één jaar; fg = het aantal ongevallen * 1.000.000/het totaal aantal uren blootstelling aan het risico (Codex Welzijn op het Werk, 1996). 7. Onder gevaarlijke stoffen worden hetzij stoffen bedoeld die voorkomen op de lijsten van bijlage II of III van artikel723quinquies van het Algemeen Reglement op de Arbeidsbescherming (ARAB), hetzij (zeer) giftige, ontvlambare,ontplofbare of oxyderende stoffen
Innovatief kosten/baten model: veiligheidsinvesteringen in een chemische onderneming
9.
10.
11.
die beantwoorden aan de indicatieve criteria van bijlage IV van artikel723quinquies van het ARAB. Met een ‘ongewoon voorval’ wordt een abnormale, ongecontroleerde (escalerende) ontwikkeling in het verloop vande activiteit (of opslag) bedoeld. Met kwalitatieve batencurven wordt bedoeld dat de curven op basis van semi-kwalitatieve gegevens werden afgeleid en dat ze kunnen verschillen van situatie tot situatie en van bedrijf tot bedrijf. Een HAZOP analyse is een semi-kwalitatieve gestructureerde methodologie om gevaren te identificeren. Het is de meest populaire risicoanalysemethodiek die in de chemische procesindustrie wordt gebruikt (Reniers, 2005). De HAZOP methode wordt uitgevoerd door een multidisciplinair team en maakt gebruik van ontwerpdocumenten van de te bestuderen eenheid om via een systematisch onderzoek naar deviaties van parameters in het proces, de gevaren van de eenheid in kaart te brengen. Merk op dat de veiligheidsinvesteringen voor zware ongevallen sowieso verantwoordbaar zijn, cfr. sectie II.D.
2008 / 3
8.
345
REFERENTIES
Review of Business and Economics
Baker, J., 2005, The Report of the BP U.S. Refineries Independent Safety Review Panel (U.S. Chemical Safety and Hazard Investigation Board, Texas City, Texas). Blacconiere, W. en Patten, D., 1994, Environmental Disclosures, Regulatory Costs and Changes in Firm Value, Journal of Accounting and Economics 18. Chartered Institute of Personnel and Development, 2005, Recruitment, retention and turnover (CIPD, London, England). Claes, P.F., 2004, Risicomanagement (Wolters-Noordhoff, Groningen/Houten, Nederland). Dahlgren, K., 2001, Safety Performance Indicators, Topical Issues in Nuclear Safety (Vienna, Austria). De Greef, M. en Van den Broeck, K., 2004, Making the case for workplace health promotion (European network for workplace health promotion, Essen, Germany). ES1: http://www.werk.belgie.be/home.aspx. ES2: http://socialsecurity.fgov.be/faofat/site_nl/stats_etudes/stats_etudes.html. Hanley, N. en Spash, C.L., 1993, Cost-Benefit Analysis and the Environment, Edward Elgar Publishing. ILO (International Labour Organisation), 2001, The economics of health, safety and wellbeing: assessing the economic value of developing a healthy work environment (Geneva, Switzerland). Ministry of Social Affairs and Health, 1999, The Tyta model – implement for evaluating the company’s working environment costs (Department for Occupational Safety and Health, Tampere, Finland). Mossink, J. en Licher, F. (ed.), 1997, Costs and Benefits of Occupational Safety and Health, Proceedings of the European Conference on Costs and Benefits of Occupational Safety and Health 1997, The Hague, The Netherlands. Reisch, M.S., 2004, Twenty years after Bhopal, Smokescreen or true reform? Has the chemical industry changed enough to make another massive accident unlike?, Chemical & Engineering News 23. Reniers, G., Soudan, K., 2003, Risicoanalyse procedures in de scheikundige nijverheid: resultaten van kwalitatief onderzoek bij 24 chemische plants, Economisch en Sociaal Tijdschrift 57(3).
Review of Business and Economics
2008 / 3
346
G.L.L. Reniers & A. Audenaert
Reniers Genserik, Dullaert Wout, Soudan Karel, Ale Ben, 2005, The use of current risk analysis tools for preventing external domino accidents, Journal of loss prevention in the process industries, 18(3). Rikhardson, P., 2006, Corporate cost of occupational accidents: an activity-based analysis, Accident Analysis and Prevention 36(2). Simons, R., 2000, Performance measurement and control systems for implementing strategy (Prentice Hall, New Jersey). Wilkins, L., 1987, Shared vulnerability: the media and American perceptions of the Bhopal disaster (Greenwood Publishing Group, New York, New York).
2008 / 3
Book reviews
Recently two very interesting books have been published at Oxford University Press regarding the international trade system: Mavroidis, P.C., 2007, Trade in Goods: The GATT and Other Agreements regulating Trade in Goods (Oxford University Press, Oxford). Delimatsis, P., 2007, International Trade in Services and Domestic Regulations: Necessity, Transparency, and Regulatory Diversity (Oxford University Press, Oxford).
Review of Business and Economics
Recent negotiations in Geneva at the World Trade Organisation failed to successfully conclude the Doha Development Round. Further multilateral liberalization of world trade is hampered by both conflicting economic interests and the complexity of the international trade system. These two books shed a light on the general context of these issues. Both books are written by distinguished international law scholars, respectively Peter C. Mavroidis (Columbia Law School, University of Neuchâtel and CEPR) and Panagiotis Delimatsis (Universities of Tilburg and Bern). Although they take a legal perspective on the topic, they contain very interesting economic insights as well, which may undoubtedly be helpful for future economic analysis in the field of international trade. Examples include the welfare effects of trade, the interaction between multilateral and regional trade integration, antidumping and safeguards. Moreover these books perfectly fit into the increasing interest in multidisciplinary research regarding international trade policy. Mavroidis’ book contains an extensive overview of the history and principles of the General Agreement on Tariffs and Trade (GATT) and the World Trade Organisation (WTO). Undoubtedly there are many books doing so, but Mavroidis succeeds in combining law and economic insights in a clearly written and well-structured text. The book is also very critical regarding the current ‘rules of the game’ as well as about the achievements of the past. Finally Mavroidis makes clear suggestions for improving the trade system in the future. Apart from the general overview, the book contains detailed insights regarding the principles of ‘national treatment’,
Review of Business and Economics
2008 / 3
348
Book reviews
‘government procurement’, ‘state contingencies’ and the ‘dispute settlement mechanism’. These issues are all involved in the recent attempts of future liberalization of world trade. Moreover international trade policy in the coming decades will increasingly touch on these issues (e.g., active trade policy by major developing countries, increasing number of dispute settlement cases). Mavroidis focuses (deliberately) on trade in goods only, since trade in services is even more complex. Delimatsis’ book is one of the first and most complete overviews of the progress in the liberalization of services trade. In particular, Delimatsis focuses on the interaction between international services liberalization (like the General Agreement on Trade in Services – GATS) and domestic regulation. The latter is mostly used instead of formal trade barriers, which are hard to implement in the case of services, to shield the domestic economy from foreign competition. Contrary to trade in goods, the economic insights regarding the liberalization of trade in services are scarce. Hence, on the one hand, legal insights clearly dominate the current economic insights. On the other hand, legal research may be helpful for future economic research. However, as the author clearly points out, the legal value of existing law is often uncertain too. The GATS itself is an incomplete contract and a substantial part of legislation is based on sector-specific case-law. Hence there is a clear need to improve the trading system and international trade policy in the future. The author recognizes nevertheless the merits of regulatory diversity across countries. Hence any reform, aiming at successful liberalization of trade in services, should be a balance between useful and acceptable domestic regulation and a fairer international set of rules (either through increased mutual recognition, international harmonization of legislation or domestic treatment). In particular, the author calls for a horizontal test, across all services sectors, in order to limit the number of unnecessary and trade-distorting barriers. This is undoubtedly a very good suggestion. Given the limited size of current trade flows in comparison with the overall importance of services in our economy, services liberalization will undoubtedly continue to be on the agenda of international negotiations. The ideas put forward by Delimatsis are definitely worth being taken into account. To conclude, both books are highly recommendable to any reader interested in international trade policy. They contain not only complete overviews of the current international legislation, but also critical and original insights into potential reforms in the international trading system. Economic research will undoubtedly benefit from these extensive legal insights. Jan Van Hove (H.U.Brussel and K.U.Leuven) Transformational Changes in Higher Education, 2007, ed. Madeleine B. d’Ambrosio and Ronald G. Ehrenberg (Edward Elgar Publishing, Cheltenham). Disclaimer: Many people were involved in the writing of this book and in an attempt to maintain the summary readable and concise I refer interested readers to
Book reviews
349
Personal Impression: This book portrays the problematic facing higher education in the United States today and constitutes an informative summary of the challenges ahead. The authors (and guest speakers) raise interesting questions and sometimes offer ideas as to what the solutions may be. Some chapters include research data which makes the problematic clearer and the argument more convincing. Each chapter of the book being written by a different author, the overall writing style is somewhat chaotic and the information is sometimes redundant, but the content remains and is presented in a logical order.
Catherine Haeck (K.U.Leuven)
Review of Business and Economics
Summary: Higher education in the United States is under pressure to undergo major transformational changes to be able to efficiently produce one of the key components to sustain economic growth, namely high skilled human capital. Leading a university in this environment is a demanding and stressful mandate and the support of the board members during the transition between presidents and after is essential. The role of the president being so important, the choice of the president is also extremely important. Financing higher education is a central theme. Universities have different sources of funds and with the share coming from the state going down they need to proactively reduce their costs and look for other ways to finance their activities. The complexity of the means by which they can finance their activities calls for a greater cooperation between the universities to understand them. To reconcile the expectations of the different stakeholders of higher education also constitutes a challenge. At the core of these challenges lies the balancing act between efficiency and equality. In higher education consumers are sensitive to costs and quality, such that competition between universities may lead to higher costs through increased quality. With all this in mind, tuition and aid policies need more and more to be designed in a creative way to achieve at the same time increased revenues and increased access from the students from low income family background. In order to design these policies the use of empirical models is highly recommended. Economic research (by Catharine bond Hill and Gordon Winston) revealed that high profile students with low income family background were under represented in higher education. As a result, many colleges and universities designed new aid program to palliate to this. In the name of greater efficiency, colleges and universities also need to respond to the demand for greater accountability but the measures used to assess their efficiency needs to be designed cautiously as these measure may have perverse effects. In sum, higher education faces complex challenges that require carefully designed solutions based on sound research involving the input of the different stakeholders.
2008 / 3
the book to obtain the appropriate references. The opinion of the author was kept silent in the section labelled summary.
Geert Van Hootegem, Pierre van Amelsvoort, Gert Van Beek en Rik Huys
Anders organiseren & beter werken
Review of Business and Economics
Handboek sociale innovatie en verandermanagement
op hun onderlinge relaties daardoor zin en betekenis krijgen. Een hersteloperatie die tegelijkertijd het product- en technologisch innovatief vermogen van organisaties verhoogt. Met dit boek hopen we tevens instrumenten en tools aan te reiken om met deze ideeën aan de slag te gaan. Dit boek is bedoeld voor al diegenen die bij arbeidsorganisatorische of sociale innovatie betrokken zijn: kader- en directieleden, P&O- of HR-professionals, project- en verandermanagers, interne en externe consultants, leidinggevenden, teamcoördinatoren, … Door de vele voorbeelden, tools en instrumenten is het boek ook geschikt voor studenten. Door de ontgroening en de vergrijzing op de West-Europese arbeidsmarkten zal het arbeidsaanbod gevoelig krimpen de komende decennia. Om meer mensen (langer) aan het werk te houden, zullen we moeten zoeken naar modellen die het werk werkbaar maken (beter werken). Vandaag ondervinden we voortdurend aan den lijve hoe verstikkend, hoe ineffectief en hoe onwerkbaar de bureaucratische manier van organiseren is. En toch raast die bureaucratische organisatievorm maar door. Met dit boek willen we een alternatief aanbieden door organisaties voor te stellen om te kantelen. Om opnieuw de logica van de processen op te zoeken, om te ontregelen en te decentraliseren. Kortom, om zichzelf anders te organiseren. Met dit voorstel schrijven we ons in in de brede beweging die vandaag onder de noemer sociale innovatie opgang maakt. Als complement van technologische en productinnovatie willen wij sociale innovatie zien als vernieuwingen in de sfeer van de wijze waarop mensen met elkaar (moeten) samenwerken en de wijze waar-
Dit boek is deel 1 uit de Synergie-‘trilogie’. Het tweede boek Woorden wekken, voorbeelden strekken (Van Beek & Van Hootegem, 2008) bevat casebeschrijvingen van twaalf innovatieprojecten. Het derde boek In het land van Flanders synergy (Van Hootegem, Huys, Van Beek & Beens, 2008) schetst de demografische en maatschappelijke context waarin het overheidsbeleid gestalte moet krijgen.
ISBN 978-90-334-7033-2 208 blz. € 21,45 (excl. verzendingskosten)
Uitgeverij Acco Blijde Inkomststraat 22, 3000 Leuven tel. 016/62 80 00 fax 016/62 80 01 e-mail:
[email protected] www.uitgeverijacco.be
Gert Van Beek & Geert Van Hootegem (eds.)
Woorden wekken, voorbeelden strekken Een dozijn lessen organisatieverandering in woord en beeld
Vlaanderen en de Beschutte werkplaats Klein-Brabant aan hun arbeidsdeling. Deceuninck ontwikkelde een gezondheidsbeleid en Roularta integreerde verschillende deelredacties tot één integrated newsroom.
ISBN 978-90-334-7043-1 120 blz. € 17,80 (excl. verzendingskosten)
Uitgeverij Acco Blijde Inkomststraat 22, 3000 Leuven tel. 016/62 80 00 fax 016/62 80 01 e-mail:
[email protected] www.uitgeverijacco.be
Review of Business and Economics
Tussen juni en september 2006 startten 12 projecten, onder de vleugels van het Europees Sociaal Fonds, met als doel de werkbaarheid te verhogen via ingrepen in de arbeidsorganisatie. Hun verhalen zijn gebundeld in dit boek, dat zowel een inspiratieals leerboek wil zijn. De gebundelde ervaringen van de pilootprojecten zullen ondernemers, leidinggevenden en werknemers inspireren om samen werk te maken van sociale innovatie. Aan de projectpromotoren is tevens gevraagd de hindernissen, lessen, tips & tricks, do’s en don’ts neer te schrijven. Het boek bevat de innovatieprojecten van ArcelorMittal Gent, Bosch Tienen en EADS rond competentiemanagement. Deloitte en Tandtechnisch Labo Camerlynck initieerden en Philips Turnhout evalueerde zelfsturing. Ook BnS Engineering maakte de kanteling. Binnen de sociale economie sleutelden De Oesterbank, de Kringwinkel Midden West-
Dit boek is deel 2 uit de Synergie-‘trilogie’. Het eerste boek Anders organiseren & beter werken (Van Hootegem, van Amelsvoort, Van Beek & Huys, 2008) bevat de theorie en praktijk van sociale innovatie en verandermanagement. Het derde boek In het land van Flanders synergy (Van Hootegem, Huys, Van Beek & Beens, 2008) schetst de demografische en maatschappelijke context waarin het overheidsbeleid gestalte moet krijgen.
Geert Van Hootegem, Rik Huys, Gert Van Beek en Ellen Beems
In het land van Flanders synergy Werken en ondernemen in een innovatieve economie
Review of Business and Economics
wanneer de markt tegelijkertijd verschillende eisen stelt: kostprijs, kwaliteit, flexibiliteit, innovatie, duurzaamheid, ... Om de innovatie van organisaties te ondersteunen, is dus de toepassing van een geschikte arbeidsdeling nodig. Een arbeidsdeling die organisaties ondermeer toerust met (zelfsturende) teams. Deze arbeidsdeling resulteert tegelijk in een betere kwaliteit van de arbeid. En met deze verbeterde arbeidskwaliteit kan de (vervroegde) uittrede uit de arbeidsmarkt worden vermeden. Maar dan moet er wel een aangepast beleid worden gevoerd. Een beleid dat de innovatieve arbeidsorganisatie stimuleert en ondersteunt. In de tekst komen de voorstellen van de taskforce ‘Innovatieve arbeidsorganisatie’ aan bod. De activiteitsgraad in Vlaanderen stijgt langzaam, maar de vooropgestelde doelstelling van 70% is nog lang niet behaald, laat staan in zicht. Met 32% werkzame 55-plussers in 2005 behoort Vlaanderen tot de staart van het Europese peloton. Gelet op de veroudering van de beroepsbevolking plaatst dit een steeds grotere rem op de globale toename van de activiteitsgraad. Ongetwijfeld spelen een reeks gunstige uitstapregelingen de rol van ‘pullfactoren’ die werknemers uit de arbeidsmarkt trekken. Maar er zijn ook ‘pushfactoren’ die verbonden zijn met de kwaliteit van de arbeid. Het boek is daarom een pleidooi om, conform wetenschappelijke inzichten, jobs te ‘activeren’. Om meer ‘actieve’ jobs te verkrijgen moeten de regelmogelijkheden toenemen en dat vereist dat het takenpakket op een andere wijze wordt samengesteld. De sleutel voor het verhogen van de werkbaarheid van jobs, ligt dus in de arbeidsdeling. Studies tonen aan dat er ook in Vlaanderen nog klassieke of traditionele organisaties bestaan. Zij krijgen het echter moeilijk
Dit boek is deel 3 uit de Synergie-‘trilogie’. Het eerste boek Anders organiseren & beter werken (Van Hootegem, van Amelsvoort, Van Beek & Huys, 2008) bevat de theorie en praktijk van sociale innovatie en verandermanagement. Het tweede boek Woorden wekken, voorbeelden strekken (Van Beek & Van Hootegem, 2008) bevat casebeschrijvingen van twaalf innovatieprojecten.
ISBN 978-90-334-7044-8 56 blz. € 12,00 (excl. verzendingskosten)
Uitgeverij Acco Blijde Inkomststraat 22, 3000 Leuven tel. 016/62 80 00 fax 016/62 80 01 e-mail:
[email protected] www.uitgeverijacco.be