MAXIMIZATION, INCOMPARABILITY, AND MANAGERIAL CHOICE Nien-he Hsieh Abstract: According to one prominent view of rationality, for the choice of alternative to be jusdfied, it must be at least as good as other alternatives. Michael lensen has recently invoked this view to argue that managers should act exclusively to maximize the long-run market value of economic enterprises. According to lensen, alternative accounts of managerial responsibility, sucb as stakeholder theory, are to be rejected because they lack a single measure to compare alternatives as better or worse. Against lensen's account, this paper argues that choosing the alternative that is at least as good as other alternatives need not preclude managers from respecting considerations in addition to long-run market value. The paper argues that such considerations may be incorporated into managerial decision-making by introducing constraints and priorities into the process of maximizing long-run market value and by allowing for "clumpy" values. According to one prominent view of rationality, for the choice of an alternative to be jusdfied, it must be at least as good as the other altematives. Michael Jensen has recently invoked this view of rationality to provide a novel defense of the thesis that managers should act exclusively to maximize the long-run market value of economic enterprises (Jensen, 2002: 236). According to Jensen, compedng theories of corporate accountability, such as stakeholder theory, are to be rejected because they fail to provide a single measure against which alternadves can be evaluated as being better or worse, thereby precluding the possibility of jusfified choice on the part of managers. Managers, in other words, ought exclusively to maximize long-run market value not only as a matter of morality, but also as a requirement of rationality.' By focusing on the requirements of rationality, Jensen presents a strong challenge to theories of corporate accountability that call upon managers to respect considerations in addifion to the long-run market value of economic enterprises. Following Jensen, call choosing an altemadve on the grounds that is at least as good as all other altemadves, maximization.^ The view that jusdfied choice requires maximizadon is held by many and reflects the more general view that jusdfied choice requires the comparabihty of altemadves—a view not only explicit in radonal choice and decision theory according to Ruth Chang, but also implicit in "most forms of © 2007. Business Ethics Quarterly, Volume 17, Issue 3. ISSN 1052-150X. pp. 497-513 BUSINESS ETHICS QUARTERLY consequentialism, some versions of virtue theory, and, arguably, certain forms of deontology" (Chang, 1998:1577-78). According to Jensen, one risks rejecting this view of justified choice if one holds that mangers ought to respect considerations in addition to long-mn market value. To be certain, one need not accept the view of justified choice upon which Jensen's account relies. Indeed, whether justified choice requires maximization is a subject of much debate.^ However, given the prevalence of the view that justified choice requires maximization and the novelty of Jensen's challenge, this paper aims to address Jensen's account on its own terms. Jensen's account suggests both a practical challenge and a theoretical challenge."* The practical challenge is to articulate a workable guide for decision-making that incorporates considerations in addition to long-mn market value, while remaining consistent with the view that rationality requires maximization. The theoretical challenge is that the specification of such a guide is, as a matter of principle, not possible. Articulating a workable guide for managerial decision-making is an important goal. However, there is little reason to consider this pracfical challenge if it is theoretically not possible to respect considerations other than long-mn market value in a manner consistent with maximization. Accordingly, as afirst step toward addressing the practical challenge of articulating a workable guide for managerial decision-making, this paper takes as its focus the theoretical challenge suggested by Jensen's account. Against Jensen's account, the paper argues that viewing justified choice as maximization does not preclude respecting considerations in addition to long-mn market value within an account of corporate accountability. The argument proceeds as follows. The first section summarizes Jensen's account and outlines the practical and theoretical challenges that his account poses for theories of corporate accountability such as stakeholder theory. The second section develops the role of constraints and priorities as ways to incorporate considerations other than long-mn market value into an account of managerial choice as maximization. The third section argues that an account of managerial choice as maximization can further incorporate values in addition to the long-mn market value of economic enterprises. By drawing on an account developed elsewhere (Hsieh, 2005a), the section argues that this is to be done by way of recognizing the possibility of maximizing with plural values. With the aim of developing a workable guide for managerial decision-making, the fourth section illustrates an application of the theoreticalfi'amework advanced in this paper. L Maximization and Incomparability According to Jensen, "managers should make all decisions so as to increase the total long-mn market value of the firm" (Jensen, 2002: 236). Call this thesis of managerial responsibility, the market value thesis.^ This thesis about managerial responsibility, or some variation thereof, is shared by a number of authors (Friedman, 1982, 1987; Machan, 1999; Stemberg, 2000). What is particularly salient about Jensen's account is the nature of his defense of the thesis. In contrast to au MAXIMIZATION, INCOMPARABILITY, AND MANAGERIAL CHOICE 499 thors who invoke moral grounds, such as libertarianism, to defend the market value thesis, Jensen focuses also on the requirements of radonality.^ By focusing on the requirements of radonality, Jensen's account presents a strong challenge to theories of managerial responsibility that cedl upon managers to incorporate consideradons in addidon to the long-run market value of economic enterprises.^ In his defense of the market value thesis, Jensen argues for the rejecdon of stakeholder theory, which he takes to be the main altemadve to the market value thesis.* According to Jensen, "stakeholder theory is incomplete as a specificadon for the corporate purpose of objecdve funcdon, and therefore cannot logically fulfill that role" (Jensen, 2002:236). Because stakeholder theory calls upon managers "to make all decisions so as to take account of the interest of all the stakeholders in a firm" (2002: 236), stakeholder theory requires managers to consider compedng interests without providing a principled basis on which to make trade-offs among them (2002: 238). Without a principled basis on which to make trade-offs among compedng interests, stakeholder theory does not provide managers with a single-valued objecfive funcdon, and "a decision-maker cannot make radonal choices without some overall single dimensional objecdve to be maximized" (2002: 248). As a soludon to this problem, Jensen argues that managers ought to act exclusively to maximize the long-run market value of business enterprises. He argues that the way for managers to address the need to make trade-offs among compedng interests is to measure their respecdve contribudons to the long-mn market value of the business enterprise. This allows managers to consider muldple interests, while at the same dme providing a single dimensional objecdve to maximize (2002: 248). Although there may be situadons in which a manager is not able to determine a global maximum, what the single dimensional objecdve does provide her is a definidon of "at least as good" which is required at a minimum, on Jensen's account of jusdfied choice.' One way in which to interpret Jensen's rejecdon of stakeholder theory is as a pracdcal challenge. For example, Jensen writes "because the advocates of stakeholder theory refuse to specify how to make the necessary tradeoffs among these compefing interests they leave managers with a theory that makes it impossible for them to make purposeful decisions" (Jensen, 2001: abstract). In other words, if advocates of stakeholder theory were to specify how to make such trade-offs, then it would provide managers with a workable guide to decision-making that is consistent with viewing radonal choice as maximizadon. Consider, for example, wage negodadons between management and workers. Although there is the potendal for conflict between the maximizadon of long-run market value and the interests of workers, the interests of workers can be expressed in the same monetary terms as market value. With regard to decisions such as these, an account of corporate accountability would be disdnguishedfi-om the market value thesis by assigning some weight greater than zero to the interests of workers in reladon to the long-run market value of the economic enterprise. The difference between the market value thesis and some other account, such as stakeholder theory, would be in terms of BUSINESS ETHICS QUARTERLY their distribution of benefits and burdens as measured in monetary terms. Competing interests could be incorporated as a consideration for managerial responsibility in a way that need not require rejection of the view that justified choice requires maximization. The problem with stakeholder theory, on one reading of Jensen, is that its proponents have yet to articulate a workable guide that specifies the relevant distribution of benefits and burdens. Competing interests, however, are not the only way in which accounts of managerial responsibility call upon managers to incorporate considerations other than long-mn market value into their decision-making. Accounts of managerial responsibility also require managers to incorporate plural values. As I wiU discuss below, this raises the possibility that Jensen's account advances not only a practical challenge, but also a theoretical challenge—namely, that in theory, one cannot incorporate considerations in addition to long-run market value in a manner consistent with maximization. Consider, for example, the choice between two possible routes for an oil pipeline. Route A is the shorter of the two and passes through a unique ecosystem that is home to a number of rare bird species. Route B avoids this ecosystem by passing around it through a vast expanse of barren desert. Given the difference in cost. Route A maximizes the long-mn market value of the business enterprise. At the same time, in following Route A, there is the risk that the pipeline will result in irreparable harm to the rare ecosystem through which it passes. If it were possible to translate the value of the ecosystem into market value, then the choice between Route A and Route B might be accommodated in the same way that competing interests were incorporated into managerial choice as in the example of wage negotiations. As has been argued by Elizabeth Anderson and others, however, to measure the value of the environment in purely market terms is to misconstme the value of the environment (Anderson, 1993). The reason for this is not that the value of the environment is difficult to measure in monetary terms. Nor is the problem that harm to the environment is an instance of market failure. Instead the problem is that part of what makes the environment valuable is an altogether different kind of value than that measured by tbe market. One way in which to see this is that the market value of a good captures only its use value, but as Anderson writes, "people value environmental goods in ways other than use: we admire many wild animals, feel wonder and awe at spectacular storms and volcanic emptions, demand consideration for delicate ecosystems, appreciate mountains and seascapes for their beauty" (1993: 205). These evaluative attitudes are not about use value, but rather about intrinsic value. To try to capture this aspect of the value of the environment in market terms is to misconstme the nature of the value of the environment and more generally to ignore the plurality of values. The problem that plural values are thought to pose for the possibility of maximization can be described in terms of incomparability.^° Following Ruth Chang, two altematives are said to be comparable if and only if they are related by a positive value relation (Chang, 1997: 6; Chang, 2002: 663). A positive value relation de MAXIMIZATION, INCOMPARABILITY, AND MANAGERIAL CHOICE 501 scribes the way in which two altemadves are related in virtue of some set of respects. For example, "worse than" is a posifive value reladon. Comparisons are intelligible only if they are made in reladon to a set of respects. Following Chang, call the set of respects in virtue of which the value relation holds between two altemadves the covering consideration (2002: 666). Two altematives are then incomparable if for every positive value relation in virtue of some covering consideration, it is not true that a posidve value reladon holds between the two items (2002: 663)." To see the problem that incomparability poses for the possibility of maximizadon, let us continue with the above example. Because the value of the environment cannot be completely captured in terms of market value, in comparing Route A and Route B there are at least two respects on which the altemadves can be appropriately compared: market value and intrinsic environmental value. Taking these two respects together to be the covering consideradon against which judgments of "better than" or "worse than" are to be made, the altematives appear to be incomparable. Route A is better than Route B with respect to market value, but Route B is better than Route A with respect to intrinsic environmental value. Neither altemadve appears to be better or worse than the other with respect to the covering consideradon. At the samefime, it seems mistaken to hold that the two altemadves are equally good. It would appear to be the case that there is no way to rank the altemadves in the manner that is required for maximization to succeed.'^ The possibility of plural values raises a different challenge for proponents of stakeholder theory than the problem of conflicdng interests. The challenge is not simply that proponents of stakeholder theory have yet to specify the trade-offs to be made among compedng interests in a manner consistent with maximizafion, but that maximizafion with considerafions in addidon to long-run market value is inherendy impossible. If this challenge is correct, given that long-run market value is an important consideration on any account of managerial responsibility, it would appear that managers should focus exclusively on maximizing the long-mn market value of economic enterprises. In other words, Jensen's account suggests not only a pracdcal challenge, but also a more fundamental theoredcal challenge against theories of corporate accountability such as stakeholder theory. If it is theoredcally not possible to respect consideradons in addidon to long-run market value in a manner consistent with maximizadon, then there is little reason to consider Jensen's practical challenge to articulate a workable altemadve to the market value thesis. Accordingly, it is to addressing this theoredcal challenge that I now tum. 2. Constraints and Priorities This secdon begins by oudining one way in which to address the theoredcal challenge presented by Jensen's account. The secdon outlines one way in which compedng interests and plural values can be incorporated into an account of managerial responsibility in a manner that is consistent with the view of managerial choice as maximizadon. In particular, the secdon argues that recognizing constraints and BUSINESS ETHICS QUARTERLY priorities in the process of maximizing long-mn market value provides one way in which to incorporate competing interests and plural values into theories of corporate accountability. The place to begin is to recognize that most theories of managerial responsibility committed to some version of the market value thesis hold that managers ought to respect certain constraints in their pursuit of maximizing the long-mn market value of economic enterprises. For example, libertarian theories of corporate responsibility, which include accounts by Milton Friedman, Tibor Machan, and Elaine Stemberg, call upon managers to maximize profits for shareholders, while holding at the same time that principles of faimess or human rights place constraints on what managers are permitted to do in that pursuit (Amold, 2004). These constraints specify courses of action that that it would be impermissible for managers to follow in the pursuit of maximizing the long-mn market value of economic enterprises. As long as these constraints are met, managers need only pursue the maximization of long-mn market value. To the extent that respect for specific interests of stakeholders as well as respect for certain values can be embodied in the form of constraints, then incorporating competing interests and plural values into account of managerial responsibility need not require rejecting tbe view of managerial choice as maximization. In tum, just as specifying constraints need not undermine the basic framework of maximizing, neither too might recognizing priorities that specify hmited courses of action under well-defined conditions. To illustrate the way in which specifying priorities is consistent with the basic conception of managerial responsibility as the maximization of long-mn market value, consider the Rescue Principle. The Rescue Principle holds that when a person encounters a situation in which another person is at risk of suffering grave harm, she is under a duty to prevent that grave harm if she is able to do so at a moderate cost to herself or even to others (Scanlon, 1998; Singer, 1972). The Rescue Principle has been invoked to argue, for example, that pharmaceutical companies have an obligation to donate medications to persons suffering from HIV/AIDS in developing countries (Dunfee, 2006; Hsieh 2005b). Although the Rescue Principle requires managers to forgo maximizing the long-term market value of economic enterprises, the Rescue Principle places a limit on the amount of assistance required. Under the Rescue Principle, once that limit is met, managers are free to maximize the long-term market value of business enterprises. As such, to the extent that competing interests and plural values can be specified as well-defined and limited priorities, then incorporadng competing interests and plural values into a theory of managerial responsibility need not require rejecfing the view that justified choice requires maximization. At this point, two objections may be raised. Thefirst objection is that recognition of constraints and priorities in the process of maximizing long-mn market value is distinct from engaging in maximization. Because constraints and priorities are themselves not objects of maximization, it may be objected that incorporation of competing interests and plural values by way of recognition of constraints and priori MAXIMIZATION, INCOMPARABILITY, AND MANAGERIAL CHOICE 503 des requires moving beyond the view that jusdfied choice requires maximizadon. '^ The second objecdon is that incorporating compedng interests and plural values by way of recognizing constraints and priorifies does not adequately reflect the extent to which accounts of corporate accountability call upon managers to incorporate considerations in addition to long-mn market value into their decision-making. Constraints only preclude certain courses of acdon, and although priorides specify courses of acdon, in order to remain consistent with maximizing long-run market value, the priorides may need to remain limited with regard to the consideradons that they require managers to respect. Consequendy, it may be objected that the challenge remains of articulating an altemadve to the market value thesis that is consistent with the view that jusdfied choice requires maximizadon. In response to the first objection, I raise two points. First, as noted above, most versions of the market value thesis recognize some set of moral consdaints to the maximizadon of long-run market value. As such, this first objection calls into quesdon more generally the extent to which proponents of the market value thesis are in fact committed to the view that justified choice requires maximization. Second, recognition of constraints and priorides need not be the only way in which to incorporate considerations other than long-run market value into an account of corporate accountability. The account that follows argues for another way in which to incorporate compedng interests and plural values without relying on constraints and priorides, and thereby aims to address not only the first objecfion, but the second objection as well. 3. dumpiness In response to the preceding objections, this section puts forward another way in which accounts of corporate accountability can incorporate consideradons other than long-run market value into managerial decision-making without rejecdng the view that jusdfied choice requires maximization. Depending on whether one regards the recognition of constraints and priorities to be consistent with maximizadon, the approach outlined in this secdon may be taken either as a way to supplement the approach oudined in the previous secdon or as an alternative to it. The secdon begins by considering the possibility of maximizing across plural values and then extends consideradon of this possibility to the case of compedng interests. One way in which to make comparisons in the face of plural values is in terms of the concept of dominance. '"• Let us say that an item X dominates another item Y if on each of the relevant values, item X is equally good or better than item Y. To condnue with the example of the oil pipeline, suppose that there is a third alternate route. Route C, that also passes through the rare ecosystem but along a route that is longer dian Route A. Assuming that the cost of building the pipeline is proportional to its length and that therisk of environmental damage is the same for pipelines built along either Route A or Route C, then Route A may be said to dominate Route C. Route A is better in terms of market value and it is just as bad in terms of environ BUSINESS ETHICS QUARTERLY mental damage. If we accept that Route A is at least as good as Route C, in cases of dominance, maximization is possible with plural values. The incorporation of plural values by way of dominance is incomplete. There remains the possibility in some comparisons that neither altemative dominates the other. In the example at hand, neither Route A nor Route B dominates the other. Neither Route A nor Route B would be justified in terms of maximization if the account of corporate accountabihty required managers to consider both long-mn market value and intrinsic environmental value. Dominance, however, is not the only way in which to make comparisons in the face of plural values. Consider the example of grading student papers. In evaluating student papers, what is at stake is not the grade itself, but rather some quality like "goodness as an undergraduate student paper for a given course." This quality is likely to be comprised by a number of respects, such as originality, constmction of argument, effectiveness of prose, and grasp of course material. Although grades are in part relative, they do provide a way to speak about "goodness as an undergraduate student paper for a given course" in more than a relative sense. To say that a paper is a "B" paper is not only to say that it is better than a "C" paper with respect to the covering consideration, but that it is better in virtue of possessing certain qualities with respect to the covering consideration. For example, it might be well written, but not highly original. Grading involves a covering consideration that is clumpy or exhibits dumpiness.'^ A clumpy covering consideration sorts items into classes, or clumps, based upon the degree to which the items possess each of the relevant respects that comprise the covering consideration. B-papers, for example, comprise a clump. For a given covering consideration, comparisons may be more or lessfine-grained. One grading system may function in terms of the clumps, "A," "B," "C," etc. Another grading system with a more fine-grained resolution may function in terms of the clumps, "A-H," "A," "A-," etc. The clumps into which items are sorted, or belong, might be said to reflect the smallest unit of measurement for purposes of comparison against the covering consideration. In tum, comparisons between items are made in terms of the clumps to which they belong. Papers that belong to the "B" clump are related by the comparative relation "better than" to papers that belong to the "C" clump. Papers in the same clump are related by the relation "equally good."'* In contrast to comparisons by way of dominance, comparisons involving clumpy covering considerations operate in terms of holistic comparisons.'^ A comparison of dominance involves a series of individual comparisons between two papers on each of the respects that comprise the covering consideration, such as originality, constmction of argument, effectiveness of prose, and grasp of course material. A holistic comparison similarly involves comparisons of the papers with regard to each of the individual respects. However, a holistic comparison involves an additional step. This additional step is the specification of the requisite degree to which an item must display the various respects that comprise the covering consideration for purposes of evaluative comparisons between items. One way, but not the only way. MAXIMIZATION, INCOMPARABILITY, AND MANAGERIAL CHOICE 505 to conceive of this is to imagine that a holisdc comparison attaches a weight to each of the individual respects on the basis of their reladve importance in determining the overall comparison between two papers. For example, one paper might be better in terms of originality and constmction of argument whereas the other paper might be better in terms of effecdveness of prose and grasp of course material. In this instance, the relation between the two papers is indeterminate if dominance is the only way by which to make comparisons involving plural values. In contrast, a holisdc comparison translates the individual component comparisons into a single evaluadve comparison with regard to the covering consideradon, "goodness as an undergraduate student paper for a given course." Clumpy covering considerations enable maximizafion in the face of plural values. In contrast to maximizadon by way of dominance, maximizadon by way of clumpy covering consideradons does not giverise to the problem of incomparability. In the case of grading papers, for example, even though the comparison involves muldple respects, any two papers can be compared with respect to the covering consideradon, "goodness as an undergraduate student paper for a given course." Of any two papers, they may be in the same clump, in which case the two papers are equally good, or they may be in different clumps, in which case one paper is better than the other. Of any two papers, in other words, one is at least as good as the other. In this manner, clumpy covering consideradons enable maximizadon in the face of plural values. Similarly, if we imagine the possibility of a clumpy covering consideradon involving both long-mn market value and intrinsic environmental value, then a comparison is possible between Route A and Route B in the example of the oil pipeline. Depending on the relative importance assigned to both values in such a covering consideration. Route A and Route B may belong in the same clump, in which case they are equally good altematives, or they may be in different clumps, in which case one route is better than the other. In tum, given that one of the routes is at least as good as the other, recognition of plural values need not preclude managerial choice as maximization. At this juncture, it may be objected that comparing routes for an oil pipeline differs from comparing student papers in an important sense, which is that the former involves conflicdng considerations in the way that the latter typically does not. The clumpy covering consideradon that incorporates both long-mn market value and environmental value must often rank altemadves for which furthering one value comes at the expense of furthering the other. In contrast, there is typically no conflict between the muldple respects that comprise the covering consideradon, "goodness as an undergraduate student paper for a given course." Apart from constraints of dme and effort, for example, originality, constmcdon of argument, effecdveness of prose, and grasp of course material are typically not achieved at the expense of one another. This difference, it may be thought, counts against the possibility of a clumpy covering consideradon that incorporates both long-mn market value and environmental value."* BUSINESS ETHICS QUARTERLY In response, the point may be raised that this difference does not result from features of the covering considerations themselves, but rather is a contingent feature of the altematives being evaluated. For example, if there were an oil pipeline route—call it Route D—that realized long-mn market value to a greater extent than Route A and that realized environmental value to a greater extent than Route B, it is natural to suppose that the clumpy covering consideration at hand would place Route D in a clump that was better than the clump containing Route A or the clump containing Route B. The fact that altematives that are superior on all dimensions, such as Route D, are rarely available to managers is independent of the way in which the clumpy covering consideration is specified. As such, there appears to be no reason to mle out a clumpy covering consideration incorporating both long-run market value and environmental value on grounds that it would need to rank altematives for which furthering one value comes at the expense of another. In tum, this response gives us reason to hold that maximization in the context of competing interests is also possible by way of acknowledging a clumpy covering consideration. Consider the example raised earlier conceming wage negodations between management and workers. The interests of shareholders and the interests of workers may be incorporated as though they were each a different value. The sense in which the interests are competing may be reflected in a covering consideration that ranked as worse any altemative that favored only one interest in comparison to an altemative that favored both interests to some degree. Any two negotiated outcomes would belong to the same clump or belong to different clumps. In tum, this would allow for maximization in the context of competing interests in a manner no different from that of incorporating plural values. There is reason to conclude that viewing rationality as maximization need not preclude the possibility of articulating an account of corporate accountability that guides managers to respect considerations other than long-mn market value in their choices. To close this secdon, it may help to disdnguish between the claim that two altemadves are comparable by way of a clumpy covering consideradon and the claim, common in radonal choice theory, that for any two altemadves, a decision-maker is able to rank one of them as being at least as good as another. The latter claim does not specify in virtue of what the ranking is to be made. It may be said that the altemadve the decision- maker ranks as being at least as good as the other altemadve is, by virtue of that ranking, at least as valuable as the other altemadve. However, in the case of managerial decision-making, theories of corporate accountability are intended to provide managers with guidance in their choices. As such, given the assumpdon that jusdfied choice requires maximizadon, the debate about corporate accountability is the debate over specifying that which managers ought to maximize. In such a context, it is inadequate to respond to Jensen's challenge by claiming that for any two altemadves a decision- maker is able to rank one altemadve as being at least as good as another. To respond to the challenge raised by Jensen, the theory of corporate accountability must specify that in virtue of which an altemadve is at least as good as another altemadve. A clumpy covering consideradon, as has been argued, allows for such a specificadon." MAXIMIZATION, INCOMPARABILITY, AND MANAGERIAL CHOICE 507 4. An Example The paper takes addressing Jensen's theorefical challenge to be important in its own right as well as a required first step in meedng the practical challenge of arficuladng a workable altemadve to the market value thesis. Given the importance of addressing this pracdcal challenge, the remainder of the paper takes up this task. To this end, this secdon illustrates one way in which managers can respect consideradons in addidon to long-run market value in their decisions. Although the illustradon provides a plausible basis for an alternative to the market value thesis, this discussion is not meant to argue for a specific altemadve to the market value thesis. To specify and jusdfy a complete altemadve to the market value thesis is beyond the scope of this paper. Instead, the discussion aims to illustrate the possibility of articulating a workable altemative to the market value thesis in a manner consistent with the theoredcal framework described above.^° In pracdce, managers face a variety of suggesdons about ways in which to respect consideradons in addidon to long-mn market value. For example, there are the mission statements of companies. For managers of muldnadonal enterprises (MNEs), another source of guidance may be found in codes of conduct. These codes specify, either at the level of an individual corporadon or at the level of an industry, the acdons to be taken by an MNE in the event of a conflict between the corporadon and society (Sethi, 2003: 64).^' Furthermore, a well-defined organizadonal culture or the history of previous decisions may provide content to the covering consideradon against which managers are to maximize. As has been noted, organizadons have an influendal role in stmcturing the decisions taken by individuals (Margolis, 2001; Philips and Margolis, 1999).^^ The difficulty appears to be that many of these forms of guidance are open to Jensen's cridcism of stakeholder theory—namely, that they call upon managers "to make all decisions so as to take account of the interest of all the stakeholders in a firm" (Jensen, 2002: 236). As discussed in the preceding sections, however, the interests of stakeholders may differ in ways that allow them to be respected in a manner that need not be incompafible with the general process of maximization. The following example illustrates this possibility.^^ Founded in the early 1950s, SABAF is an ItaHan small muldnadonal enterprise (SME) that produces components for gas cookers and domesdc gas cooking appliances. It is one of the world's leading producers of such components, employing around 500 people. In April 2001, SABAF established a Brazilian manufacturing site under SABAF do Brasil in order to produce components to meet the demand of muldnadonal companies operadng in Brazil. Since then, SABAF do Brasil has come to produce components for other Latin American markets. Like many companies, SABAF has a charter of values, which is to serve as "the tool of govemance with which SABAF's Board of Directors, the company's goveming body, lays down the company's values, principles of conduct and commitments to its stakeholders (employees, investors, customers, suppliers, financiers, public administradon, society and the environment)" (SABAF, 2003a: 2). The company's stated vision is "to combine economic decisions and results with ethical values by going beyond BUSINESS ETHICS QUARTERLY the model of family capitalism in favor of a managerial logic geared not just to the creation of value, but also to respect for values." These values include "a focus on people, faimess and transparency, safety, and innovadon and technology" (SABAF, 2005). They provide guidance in ways that are consistent with the theoretical framework outlined above. Consider the value of safety in the context of SABAF do Brasil's operadons. The value of safety and its role in managerial decision-making at SABAF is explained as follows: Safety is one of the core concems in Sabaf business plan. Safety within the Company, perceived as protection of the physical well-being of employees, is guaranteed through constant modernisation and upgrades made to the equipment used in operations and implementation of the highest quality standards. The safety of end users is ensured by offering products with superior quality and safety characteristics. Sabaf products are submitted to strict controls, in compliance with the toughest market standards, to certify that they are free from defects. (SABAF, 2005) Safety serves as both a priority and a constraint in relation to long-mn market value. In the case of employees, it serves as a priority. For example, SABAF do Brasil complies with the same technological and worker safety standards as applied in Italy (Perrini, 2005). Safety represents a priority in the sense that only after a certain standard of safety has been met are managers to maximize the long-term market value of the enterprise. Safety also serves as a constraint. For example, given the frequent electricity blackouts in Brazil, a company contacted SABAF about transforming electric ovens into gas ovens. SABAF proposed utilizing components with safety features in the ovens. The potendal customer refused S ABAF's proposal and, in tum, SABAF declined to condnue with the project (Perrini, 2005). In this instance, safety represents a constraint in that managers decline to engage in activides that fall below a certain threshold of safety even if they may enhance long-mn market value. The objection might be raised that actions such as these are consistent with maximizing the long-mn market value of the company. To be certain, it would be unreasonable to assume there are no benefits to SABAF in taking safety to be either a priority or a constraint. For example, in 2002 and 2003, the company benefited from the fact that there were no accidents, a likely result of SABAF's emphasis on safety (SABAF, 2003b: 18). At the same time, it seems implausible to assume that SABAF could not have improved its long-mn market value by relaxing its safety standards. One of the reasons that MNEs locate in developing economics is to benefit from labor standards that are less stringent.^* Furthermore, to the extent that companies benefit from respecting stakeholder interests though favorable publicity, it would appear that such benefits apply more readily to large-scale MNEs with high public profiles. More generally, a review of the empirical literature raises doubts about the claim that respecting stakeholder interests maximizes corporate financial performance (Margolis and Walsh, 2003). MAXIMIZATION, INCOMPARABILITY, AND MANAGERIAL CHOICE 509 In illustradng the potendal for managerial decision-making to incorporate features of the theoredcal framework oudined in this paper, this discussion provides one response to the pracdcal challenge raised by Jensen's account—that is, the challenge of specifying a workable altemadve to the market value thesis. On reflecdon, the ability to incorporate consdaints should not come as a surprise. As discussed above, most accounts of corporate accountability, including those consistent with the market value thesis, recognize consdaints on the maximizadon of long-mn market value. This recognidon is consistent with the emphasis placed by many companies on respecting basic human rights, which are often constmed as taking the form of specifying constraints. Once specified, constraints may serve as a kind of "trigger," meaning that in the absence of violating a consdaint, managers ought to pursue the maximizadon of long-mn market value.^' Accordingly, in working toward articuladng a complete altemative to the market value thesis, specifying constraints presents a natural focus and starting point. This would then provide the basis for specifying priorides and broader measures to provide guidance to managers in their decision-making. 5. Conclusion This paper began with Michael Jensen's defense of the market value thesis—the thesis that managers ought to act exclusively to maximize the long-mn market value of economic enterprises. By calling upon managers to incorporate consideradons in addidon to long-run market value into their decision-making, altemative accounts of corporate accountability, such as stakeholder theory, are said to lack a single objecdve for managers to maximize, thereby precluding the possibility of justified choice. Given the prominence of the view that justified choice requires maximizadon, Jensen's account presents a sdong challenge to proponents of altematives to the market value thesis. According to Jensen, altemadve accounts of corporate accountability, such as stakeholder theory, are to be rejected not so much as a matter of morality but more fundamentally as a requirement of radonality. This paper responds largely to the theoretical challenge posed by Jensen's account. The paper has argued that viewing jusdfied choice as requiring maximizadon does not preclude articulating an account of corporate accountability that guides managers to incorporate consideradons other than long-run market value into their choices. The paper outlines two ways in which to articulate such an account. Recognizing consdaints and priorities in the process of maximizing long-mn market value provides one way in which to incorporate competing interests and plural values into theories of corporate accountability. Clumpy covering considerations further allow for the possibility of maximization in the context of plural values and compedng interests. Depending on whether one regards the recognidon of consdaints and priorides to be consistent with maximizadon, reliance on clumpy covering consideradons may be taken either as a way to supplement the approach involving consdaints and priorities or as an altemadve to it. BUSINESS ETHICS QUARTERLY To be clear, the paper leaves open the quesdon of what it is appropriate for managers to maximize and whether radonality even requires managers to maximize at all. That this quesdon remains open is, in one important sense, the point of the paper. The debate over what counts as an appropriate theory of corporate accountability is an important debate to condnue to be had. The point of this paper is that it would be unreasonable to suggest that we can avoid this debate altogether by invoking the requirements of pracdcal reason. Notes This paper benefited from presentation at the Third Biennial Global Conference on Business Ethics held at the Markkula Center for Applied Ethics, Santa Clara University and I thank participants in the conference. For their comments on this occasion and others, I especially thank Shawn Berman, John Boatright, Ryan Burg, Robert Kolb, Alexei Marcoux, John McCall, Lisa Newton, Francesco Perrini, Robert Phillips, Alan Strudler, Sara Toomey, Alec Walen, three anonymous referees for the Society of Business Ethics Annual Meeting, and three anonymous referees for Business Ethics Quarterly. For funding this research and enabling participation in the conference, I thank the Carol and Lawrence ZickJin Center for Business Ethics Research, the Wharton Legal Studies Research Program, and the Markkula Center for Applied Ethics. All errors remain my own. 1. The paper uses the terms "corporate accountability" and "managerial responsibility" interchangeably. 2. "Maximization," as the term is used by Jensen, is to be distinguished from the term as used by Amartya Sen (1997: 746; 2000: 486). Sen uses the term "maximization" to refer to choosing the altemative for which there is no better altemative. Sen uses the term "optimization" to refer to what Jensen refers to as "maximization." This paper follows Jensen's use of the term "maximization." 3. On this debate, see, for example, Byron, 2004; Chang, 1997; Sen, 1997; and Sen, 2000. 4. I thank an anonymous referee for pressing me on this distinction. 5. Jensen calls this thesis, the "value maximization thesis" (Jensen, 2002: 236). 6. For an evaluation of the moral arguments for such accounts, see Amold, 2004. 7. To be clear, Jensen also argues that by maximizing long-run market value of economic enterprises, managers maximize social welfare (Jensen, 2002: 239^0). He also argues that because there is no single dimensional objective in stakeholder theory, managers will be less accountable and better positioned to pursue their own personal interests, thereby destroying the value of economic enterprises (2002: 242). Jensen's argument is similar in form to the argument that has been made in defense of utilitarianism on grounds of rationality. I thank John McCall and Alec Walen for discussion on this point. 8. The literature on stakeholder theory is vast. For helpful discussions, see Donaldson and Preston, 1995 and Phillips, 2003. 9. To be clear, Jensen holds that justified choice requires the ability to evaluate altematives as "better" or "worse" than one another (Jensen, 2002: 238). Taking "better than" to imply "at least as good as" and "worse than" to imply "not at least as good as," the paper simplifies Jensen's requirement for justified choice to be that the chosen altemative is at least as good as other altematives. MAXIMIZATION, INCOMPARABILITY, AND MANAGERIAL CHOICE 511 10. The problem has also been described in terms of incommensurability. For a discussion, see Chang, 1997. The summary in this paragraph draws from Hsieh, 2007. 11. Chang's defmition leaves open what constitutes the range of positive value relations. If the range of positive value relations is thought to be exhausted by the relations, "better than," "worse than," and "equally good," then to say that two altematives are incomparable is to hold that neither is better than the other and they are not equally good. On the possibility of a fourth comparative relation, see Chang, 2002; Griffin, 1986: 80-81, 96-98, 104; and Parfit, 1987: 431. For an argument against such accounts, see Hsieh, 2005a. 12. It might be objected that plural values are not a problem for maximization because, as a matter of fact, people do make choices between apparently incomparable alternatives. For example, people may forgo economic benefits to preserve the wilderness. That people do make such choices is not in doubt. However, to infer from this that plural values pose no problem for justified choice is to assume away the problem. The challenge is to articulate a plausible account of justified choice as maximization that recognizes the possibility of plural values. 1 thank an anonymous referee for raising this objection. I discuss this objection further at the end of section 3 below. For discussion on this point more generally, see for example, Stocker, 1990 and the articles in Chang, 1997. 13. I thank John Boatright for raising this point. 14. The discussion in the following paragraphs draws upon the account in Hsieh, 2005a. 15. I draw inspiration from a teacher who said that she graded in terms of clumps. I thank Alan Strudler for relating this anecdote to me. 16. To be clear, sorting items into clumps at a less fine-grained resolution need not be the result of "satisficing" (Simon, 1955, 1997). It may be the case that the task at hand does not require a more fine-grained resolution even though a more fine-grained resolution may be possible at no additional cost. I thank an anonymous referee for comments on this point. 17. T. K. Seung and Daniel Bonevac make a similar distinction. They distinguish between algorithmic comparisons and nonalgorithmic comparisons, which correspond, roughly, to component comparisons and holistic comparisons. Where their account differs is that the authors consider all nonalgorithmic comparisons to be based on intuition. On the account in this paper, holistic comparisons can be well-defined (Seung and Bonevac, 1992: 801). 18. I thank Alexei Marcoux for raising this objection. 19. I thank Robert Kolb and John McCall for discussion on this point. 20. The prioritization of stakeholders is an area of both normative and empirical debate. For one approach to specifying the dimensions that are important to this prioritization as an empirical matter, see Mitchell, Agle, and Wood, 1997.1 thank Shawn Berman and the anonymous referees for Business Ethics Quarterly for helpful suggestions on this section. 21. For one detailed review of ways in which companies have implemented codes of conduct see Mamie, 2004. 22. I thank an anonymous referee for discussion of this point. 23. For a discussion of SABAF's corporate social responsibility activities, see Perrini, 2005. 24. For one discussion of this point and the possibility of maintaining labor standards, see Elliott and Freeman, 2003. 25. I thank an anonymous referee for raising this point. BUSINESS ETHICS QUARTERLY References Anderson, E. 1993. Value in ethics and economics. Cambridge, Mass.: Harvard University Press. Amold, D. G. 2004. Libertarian theories of the corporadon and global capitalism. Journal of Business Ethics, 48: 155-73. Byron, M. (Ed.). 2004. Satisficing and maximizing: Moral theorists on practical reason. Cambridge: Cambridge University Press. Chang, R. 1997. Introducdon. In R. Chang (Ed.), Incommensurability, incomparability, and practical reason: 1-34. Cambridge, Mass.: Harvard University Press. Chang, R. 1998. Comparison and the jusdficadon of choice. University of Pennsylvania Law Review, 146: 1569-98. Chang, R. 2002. The possibility of parity Ethics, 112: 659-88. Donaldson, T, & Preston, L. 1995. The stakeholder theory of the corporadon: Concepts, evidence, and implications. The Academy of Management Review, 20: 65-91. Dunfee, T. 2006. Do firms with unique competencies for rescuing vicdms of human catastrophes have special obligadons? Corporate responsibility and the AIDS catastrophe in sub-Saharan Africa. Business Ethics Quarterly, 16: 185-210. EUiott, K., & Freeman, R. 2003. Can labor standards improve under globalization? Washington, D.C: Insdtute for Intemadonal Economics. Griffin, J. 1986. Well-Being. Oxford: Oxford University Press. Friedman, M. 1982. Capitalism and freedom. Chicago: University of Chicago Press. Friedman, M. 1987. The social responsibility of business. In K. Leube (Ed.), The essence of Friedman. Stanford: Hoover Institution Press. Hsieh, N. 2005a. Equality, dumpiness, and incomparability. Utilitas, 17: 180-204. Hsieh, N. 2005b. Property rights in crisis: Managers and rescue. In M. Santoro & T. Gorrie (Eds.), Ethics and the pharmaceutical industry in the 21st century. Cambridge: Cambridge University Press. Hsieh, N. 2007. Is incomparability a problem for anyone? Economics and Philosophy, 23: 65-80. Jensen, M. 2001. Value maximization, stakeholder theory, and the corporate objective function. Harvard Business School Working Paper #00-058. Jensen, M. 2002. Value maximizadon, stakeholder theory, and the corporate objective function. Business Ethics Quarterly, 12: 235-56. Machan, T. 1999. Business ethics in a free society. In R. Frederick (Ed.), A companion to business ethics. Maiden, Mass.: Blackwell. Mamie, 1. 2004. Implementing codes of conduct: How businesses manage social performance in global social supply chains. Geneva: International Labor Office. Margolis, J. 2001. Responsibility in organizadonal context. Business Ethics Quarterly, 11:431-54. Margohs, J., & Walsh, J. 2003. Misery loves companies: Rethinking social inidadves by business. Administrative Science Quarterly, 48: 268-305. Mitchell, R., Agle, B., & Wood, D. 1997. Toward a theory of stakeholder identificadon and salience: Defining the principle of who and what really counts. The Academy of Management Review, 20: 65-91. MAXIMIZATION, INCOMPARABILITY, AND MANAGERIAL CHOICE 513 Parfit, D. 1987. Reasons and persons, corrected ed. Oxford: Oxford University Press. Perrini, F. 2005. Italian small and medium-sized multinational enterprises and foreign direct investment in developing countries: The impacts of corporate social responsibility policies and practices. ILO-Muldnadonal Enterprises and Social Policy Working Paper. Phillips, R. 2003. Stakeholder theory and organizational ethics. San Francisco: Berrett- Koehler Publishers. Philips, R., & Margolis, J. 1999. Toward an ethics of organizadons. Business Ethics Quarterly, 9: 619-38. SABAF. 2003a. Charter of values. Available at http://www.sabaf.it/english/Company/ Stock_Exchange/Download[/download.htnil. SABAF. 2003b. Social report. Available at http://www.sabaf.it/english/Company/ Stock_Exchange/Download/download.htnil. SABAF. 2005. Company website. 21 September. http://www.sabaf.it/english/Company/ The_Vision/the_vision.html. Scanlon, T. M. 1998. What we owe to each other. Cambridge, Mass.: Harvard University Press. Sen, A. 1997. Maximization and the act of choice. Econometrica, 65: 745-79. Sen, A. 2000. Consequendal evaluation and pracdcal reason. Journal of Philosophy, 98: 477-502. Sethi, S. P. 2003. Setting global standards: Guidelines for creating codes of conduct in multinational corporations. Hoboken: John Wiley and Sons. Seung, T. K., & Bonevac, D. 1992. Plural values and indeterminate rankings. Ethics, 102: 799-813. Simon, H. 1955. A behavioral model of radonal choice. Quarterly Journal of Economics, 69: 99-nS. Simon, H. 1997. Administrative behavior, 4th ed. New York: The Free Press. Singer, P. 1972. Famine, affluence, and morality. Philosophy and Public Affairs, 1: 229^3. Stemberg, E. 2000. Just business: Business ethics in action, 2nd ed. New York: Oxford University Press. Stocker, M. 1990. Plural and conflicting values. Oxford: Clarendon Press.