Global Policy Forum

Democratic Development as the Fruits of Labor


Keynote Address
Industrial Relations Research Association

Joseph Stiglitz
January 25, 2000

Table of Contents

Development Strategies and the Labor Movement

  • Objectives of Development
  • More instruments
  • The objectives and organization of this lecture

    Neoclassical Perspectives on Labor To Markets and Beyond

    Market Failures: Search Costs and Macro-economic rigidities

  • Search costs and asymmetric bargaining power
  • Macro-economic equilibrium

    Labor as a Stakeholder in Corporate Governance

    On the Importance of Residual Control Rights: Lessons from the East Asia Crisis

    Principal-Agent Problems and Worker Involvement

    Two Industrial Relations Systems: the Low Road and the High Road

    Development Strategy for Labor: From the Low Road to the High Road

  • The Need for - and Limitations of - Collective Action
  • Is there a role for collective action within the public sector?
  • Systemic problems

    Development as Democratic Transformation

    Toward Economic Democracy

    Concluding Remarks



    Development Strategies and the Labor Movement

    Objectives of Development

    Today, there is growing recognition that the objectives of development go beyond simply an increase in GDP: we are concerned with promoting democratic, equitable, sustainable development.(1) If that is our objective, then it is natural that we should pay particular attention to the issue of how the plight of workers changes in the course of development; and we should look not only at their incomes, but broader measures -- at their health and safety, and even at their democratic participation, both at the workplace, and within the broader political arena. Workers' rights should be a central focus of a development institution such as the World Bank.

    I am just completing serving three years as Chief Economist of the World Bank. During that time, labor market issues did arise, but all too frequently, mainly from a narrow economics focus, and even then, looked at even more narrowly through the lens of neo-classical economics. Wage rigidities—often the fruits of hard fought bargaining—were thought to be part of the problem facing many countries, contributing to their high unemployment; a standard message was to increase labor market flexibility—the not so subtle subtext was to lower wages and lay off unneeded workers.

    Even when labor market problems are not the core of the problem facing the country, all too often workers are asked to bear the brunt of the costs of adjustment. In East Asia, it was reckless lending by international banks and other financial institutions combined with reckless borrowing by domestic financial institutions—combined with fickle investor expectations—which may have precipitated the crises; but the costs—in terms of soaring unemployment and plummeting wages—were borne by workers. Workers were asked to listen to sermons about "bearing pain" just a short while after hearing, from the same preachers, sermons about how globalization and opening up capital markets would bring them unprecedented growth. And nowhere, in all of these discussions, did issues of workers' rights, including the right to participate in the decisions which would affect their lives in so many ways, get raised?

    It was finance ministers and central bank governors—and outsiders who often seemed to be representing international financial interests—that had the seats at the table, not labor unions or labor ministers. Indeed, even as debate on reforming the international economic architecture proceeded, these people, who would inevitably face much of the costs of the mistaken policy, were not even invited to sit in on the discussions; and I often felt myself to be the lone voice in these discussions suggesting that basic democratic principles recommended that not only should their voice be heard, but they should actually have a seat at the table. To be sure, increasing attention did get focused on safety nets, but: was it simply an attempt to assuage feelings of guilt, providing too little, too late, or even worse, an attempt to moderate public criticism of "globalization without a human face?" The suspicion of the international institutions evidenced in Seattle was perhaps the not unsurprising outcome of the attitudes and policies of recent decades.

    As Chief Economist, I faced several problems. I simply could not ignore the standard arguments about the adverse effects of inflexible labor markets—and while I agreed with some of the arguments, there were others that left me unconvinced. I had to tackle those issues on terms that the economists themselves—viewing the world from their particular perspective—could understand. But there was a more positive agenda: Improving labor relations, including promoting core labor standards.

    More instruments

    These concerns quickly led to the older and broader questions of labor and development. There is a strange "studied inattention" to the possible role of the labor movement in economic development. Organized labor has played a well-known role in the already developed countries of North America, Europe, and East Asia/Pacific (e.g., Japan, Korea, and Australia) to:

  • stabilize industrial relations;
  • contribute to preserving firm-specific knowledge and organizational capital; and
  • mitigate the income inequalities that might be aggravated by the unchecked power of employers.(2)

    In addition, the labor movement in many countries has played highly constructive social and political roles; broadly construed, it has promoted

  • the adult education movement;

  • the mutualism movement in the form of credit unions, mutual banking, mutual insurance, cooperatives (consumer, marketing, worker), friendly societies, and other self-help associations;

  • the democratic movements to extend civil rights and the franchise to all adults;

  • Health and safety standards and improved working conditions; and finally,

  • Child labor standards.

    These movements all made important contributions to the economic and social development of the broader population in the now-developed world.(3)

    These perspectives bring me back to two themes that I have been stressing over the past three years. The first is that not only was the Washington consensus too narrow in its objectives—in its focus on GDP—but also in what it saw as the instruments of development, the improvement of resource allocation, through trade liberalization, privatization, and stabilization. The second, related theme, is that development needed to be seen as a transformation of society, a change in mindsets. If that is the case, then workers have to be at the center of the development transformation, and workers' organizations can be a key institution in the development process.

    The objectives and organization of this lecture

    This lecture thus has three objectives. First, I want to review the standard economic arguments; while there was much in that analysis with which I agreed, there were hidden assumptions that limited the domain of applicability, even given the limited objectives of economic efficiency on which the economics literature focused. Indeed, in recent years, there has developed a large literature that has stressed the efficiency benefits of worker involvement in the work place. Today, the World Bank sees one of its roles as disseminating knowledge. If it were clear that increased worker involvement led to increased efficiency, then one could let the matter rest at that: simply informing firms of this new finding would lead them to change their behaviors. There would be no need for government intervention—the profit motive would drive firms towards a "high worker involvement" workplace. My second objective, then, is to ask why that may not be the case, why government action may be required.

    Over the past three years, I have been particularly concerned about developing countries, and this brings me to the third objective: are there particular reasons, rationales, for government intervention in developing countries? Development entails not only creating market institutions, but also political institutions—and the two are intimately intertwined. I began this lecture by emphasizing both a broader conception of development—the transformation of society—and a broader set of objectives—democratic development. That, in turn, entails participation and involvement both within the workplace and at higher political levels.

    Neoclassical Perspectives on Labor

    If one didn't know better, it might seem as if the fundamental propositions of neoclassical economics were designed to undermine the rights and position of labor:

  • In the standard formulations of general equilibrium, e.g., Arrow and Debreu [1954] and Debreu [1959], labor is just another factor, denoted by its own subscript, xL, just like capital and land—or any intermediate good. There is nothing special about labor, nothing to suggest that labor should be treated differently from any other factor.

  • Indeed, from John Bates Clark on, there has been the view that it does not matter whether labor hires capital, or capital hires labor.

  • Coase [1937] went one step further: not only did these institutional arrangements not matter, but the distribution of wealth did not matter; so long as property rights were well defined, outcomes would be (Pareto) efficient, and indeed, in some circumstances at least certain aspects of the patterns of resource allocation would not even depend on distribution.

  • Thus, the central tenet of the Fundamental Theorems of Welfare Economics was that issues of distribution could be separated from issues of efficiency; again, so long as property rights were well defined—and so long as none of a limited number of market failures, such as externalities, arose—then the economy would be efficient. Issues of distribution could be left to a separate "branch" of government (Musgrave [1959]), and need not concern most policymakers.

  • And Coase went one step further: even when there were externalities, so long as there were well defined property rights and transactions costs could be ignored, efficient outcomes would emerge as a result of bargaining.

  • Even issues of workers' working conditions could be embraced within the standard neoclassical formulation. If workers value working conditions, then the optimal "contract" will entail firms spending on improved working conditions an amount such that the marginal value of improving working conditions still further is just equal to the marginal cost. More generally, competition among firms forces them to be "good" employers, paying full attention to all efficiency aspects of the workplace, from working conditions to organization design (e.g. the extent of involvement of workers in decision making); and issues of distribution should be handled not through labor market legislation, but through general legislation directed at redistribution.

  • Keynes (at least in Hicks's interpretation [1936, 1937]) traced the problem of unemployment back to rigid wages.(4)He thus provided a ready prescription for the doctors of modern capitalism approaching the developing countries confronted with chronic unemployment: what is needed is more flexible labor markets, which reads: abolish minimum wages, lower wages, eliminate job protection, and privatize social security. When possible, the doctors of the international financial community force the painful medicine on the country in its times of need, when it comes to the international financial institutions for help.

    If establishing these propositions that served to eviscerate the rights and positions of workers can be viewed as one of the great achievements of economics during the century from 1850 to 1950, one of the great achievements of economics during the last half century has been to show the fragility of each of the propositions. Indeed, one might argue that the real achievement of neoclassical theorists was to find the singular set of assumptions—involving perfect markets, perfect information, etc.—under which the propositions were valid, and then to dress up this highly restrictive set of assumptions in the fancy clothes of mathematical generality.

    The fundamental weaknesses—the assumptions concerning economics, e.g. information and markets—were not even listed by Debreu as assumptions; what were listed as assumptions were mathematical properties of the relevant functions. General equilibrium theorists spent much energy during the subsequent decades showing that those mathematical assumptions could be weakened, paying scarce attention to the underlying economics. It was only with the development of information economics, and the broader focus on transactions costs (Williamson [1975, 1979, 1981]) that the lack of generality of general equilibrium became apparent.

  • Labor is not like other factors. Workers have to be motivated to perform. While under some circumstances, it may be difficult to coach a machine to behave in the way desired (e.g., trying to get a computer not to crash), what is entailed in eliciting the desired behavior out of a person and out of a machine are, I would argue, fundamentally different.

  • Indeed, some might go further and argue that the central (or at least one of the central) issues of labor economics is the design of the appropriate mix of incentives and monitoring arrangements, and that some of the most important (physical and social) innovations have been those that have altered monitoring costs and devised new incentive arrangements.· Indeed, some might go further and argue that the central (or at least one of the central) issues of labor economics is the design of the appropriate mix of incentives and monitoring arrangements, and that some of the most important (physical and social) innovations have been those that have altered monitoring costs and devised new incentive arrangements.

  • But whenever information is imperfect—that is essentially always—the Fundamental Theorems of Welfare Economics do not hold. The economy is in general not (constrained(5)) Pareto efficient, and issues of efficiency cannot be separated from issues of distribution [Shapiro and Stiglitz 1984]. Whether or not the economy is Pareto efficient can depend on the distribution of income. An economy in which workers own their own land may be efficient; there are no agency costs. But an economy in which wealth is concentrated in a few hands may not only be less productive—as agency costs lead to an undermining of productivity—but there is actual scope for government intervention to make both workers and capitalists better off. There are pecuniary externalities that arise that have real consequences. The distortions associated with static resource allocation may be increased over time, through distorted incentives to innovate (the savings in private costs for labor saving innovations need not coincide with social benefits).(6) More generally, firm incentives to provide improved working conditions may depart significantly from what would be required by efficiency.

  • The fact that redistributions are not costless (there are no lump sum transfers) implies that issues of distribution cannot simply be left to the "distribution" branch of government. Distribution issues should—and do—come into every aspect of public policy, from the design of expenditure programs to labor legislation. The redistribution of bargaining power that results from collective bargaining may lead to redistributions that would not occur through the tax/welfare system, or which would be far more costly to effect in that way. Indeed, by redressing the asymmetries in bargaining associated with costly search, it is even conceivable that the overall efficiency of the economy will be enhanced.

  • These information costs undermine Coase's attempt to resolve market failures (See, e.g. Farrell [1987])

  • With incomplete contracting and imperfect risk markets, it does matter whether labor rents capital, or capital hires labor: It determines not only who bears the residual risk, but who has residual control rights—the rights to take actions not specified in the contract.

  • And recent changes in macro-economic analyses have suggested that increased labor market flexibility could actually exacerbate economic fluctuations.(7)

    Before turning to an elaboration of the unique features of labor markets that make it more likely that such markets not conform to the Arrow-Debreu idealization—and that such markets are more likely to be characterized by "market failure"—there is one attribute of the labor market to which I would like to call attention, that highlights the differences between labor and other goods.

    To Markets and Beyond

    There is an old saying that a kid with a hammer sees everything as a nail. An economist with a neoclassical toolkit sees every social problem as a "market" waiting to be developed and perfected. With a little hammering here and there, the market will solve the social problem. While the neoclassical perspective sees a role for government action in addressing market failures, but there is a deeper question of what should or should not be "on the market" in the first place. (8) Some "markets" should fail—and government has a role to see that they stay "failed." We do not believe that individuals should buy or sell votes—regardless of whether the resulting exchanges might or might not be Pareto improvements.

    The point is even more important within the labor market, and understanding this helps us understand the limitations in the idea of a competitive labor market as we see it in general equilibrium theory, e.g., the Arrow-Debreu model.

    For capital goods or land, there is the buy or rent decision. One can rent the durable entity for a period of time (e.g., buy the services of several car-days from a car-rental agency), or one can buy the entity itself (i.e., all the services the durable can provide plus the remaining salvage value). But this free market choice of rent or buy is not available for people! Alfred Marshall [1920, Chapter IV and V of Bk. VI] noted this as the first in a number of peculiarities of labor. Paul Samuelson also recognized the first peculiarity.

    Since slavery was abolished, human earning power is forbidden by law to be capitalized. A man is not even free to sell himself; he must rent himself at a wage. [1976, 52 (emphasis in the original)]

    The voluntary self-enslavement contract would be a contract to sell all of one's present and future labor services. Although now outlawed, the idea of such a contract is not of just antiquarian interest. The corresponding contract for the citizens of a country would be the Hobbesian contract of subjugation wherein people alienate their rights of self-determination to an absolute sovereign. The contemporary Harvard philosopher Robert Nozick(9) would allow the Hobbesian contract between individuals and a "dominant protective association." Nozick goes further:

    The comparable question about an individual is whether a free system will allow him to sell himself into slavery. I believe that it would. [Nozick 1974, 331]

    This ‘sophisticated' form of madness does not stop at the Harvard Philosophy Department; it is hidden within the assumptions of the competitive general equilibrium model. As Gerald Debreu puts it: a consumer/worker "is to choose (and carry out) a consumption plan made now for the whole future, i.e., a specification of the quantities of all his inputs and all his outputs." [Debreu 1959, 50] Thus a worker could sell all of her future labor at once. If sold to one buyer, it would essentially be the slavery contract. If complete future markets in labor were not allowed, then there could not be "capitalist acts between consenting adults" (Nozick's phrase) that would be a Pareto improvement, so one could not have the fundamental theorem that a competitive equilibrium is Pareto optimal. Therefore the fundamental efficiency theorem requires revising constitutional law to allow voluntary slavery contracts! Needless to say, this labor market peculiarity is not emphasized in the standard texts but it has been occasionally pointed out. The John Hopkins economist, Carl Christ, made the point in no less a forum than Congressional testimony.

    Now it is time to state the conditions under which private property and free contract will lead to an optimal allocation of resources.... The institution of private property and free contract as we know it is modified to permit individuals to sell or mortgage their persons in return for present and/or future benefits. [Christ 1975, 334; quoted in Philmore 1982, 52].

    Thus the fantasy world of the "idealized" competitive equilibrium model is not only unrealistic—a point I have emphasized throughout my career; it has even been illegal since the abolition of slavery (involuntary and voluntary). Thus those of us who were trained as neoclassical economists should not feel too guilty as we try to devise institutional solutions that do not fit well into the idealized competitive model of the textbooks.(10)

    These results stand in stark contrast to the previously prevailing wisdom. In the limited time available, I want to elaborate on questions of market failures, labor's role in corporate governance and industrial relations systems.

    Market Failures: Search Costs and Macro-economic rigidities

    Search costs and asymmetric bargaining power

    In the standard model, no firm and no worker has any bargaining power. There are an infinite number of firms offering identical jobs, and any firm attempting to lower its wage below the market wage would immediately lose all of its workers; and any worker attempting to increase his or her wage above the market wage would immediately find no job opportunity. But in the real world, there are costs to search, to finding out what wages other firms are offering, and even greater costs associated with moving from one employer to another. Even small costs can have large effects.(11)

    For instance, with e search costs, the equilibrium wage is the monopsony wage; even when there are many employers, each firm has considerable bargaining power—indeed the same that it would have had had it been the only employer. Assume that all firms paid a wage above the monopsony wage. Then if any single firm were to lower its wage by less than e, it would lose no workers. It thus pays each to lower its wage. The process continues until the wage is lowered to the monopsony wage.(12)

    Thus, whenever there are search and moving costs, there may be, in effect, a mini-bargaining problem; outcomes of such bilateral bargaining problems—even when the scope of bargaining is limited—may depart markedly from those associated with the perfect competition model, especially when there are sunk firm-specific costs and incomplete contracting. Whenever one firm raises its wage to attract more workers, it may induce a worker elsewhere to quit (indeed, that was the intent), imposing additional training costs on other firms. (13)

    Macro-economic equilibrium

    Traditional Keynesian analysis focused around a single market failure, wage rigidity, and ignored key aspects of the dynamics of adjustment. If wages and prices start to fall more rapidly, in an economy facing downward rigidities in interest rates (e.g. nominal interest rates cannot fall below zero—a problem confronting Japan), then greater wage and price flexibility leads to higher real interest rates and decreasing investment and overall aggregate demand, potentially exacerbating the economic downturn.

    Moreover, with incomplete contracting, there are large redistributions associated with unanticipated decreases in wages (and prices). Bruce Greenwald and I have argued that such redistributions may have large aggregate demand effects, because of compounding imperfections in capital markets. The unanticipated fall in wages and prices implies that firms have to pay back in real terms more than they had anticipated—a redistribution from debtors to creditors. Debtors (firms) contract investment and consumption expenditures more than creditors expand in their response to the redistribution. The net effect is a large contraction in aggregate demand, and possibly even in aggregate supply, as firms' real working capital also deteriorates, and cannot be compensated for (especially in a recession) by additional borrowing or equity issues. [Greenwald and Stiglitz 1993]

    On a priori grounds, it cannot be ascertained whether the standard "wage" rigidity effect predominates over the real interest rate/redistribution effects just discussed. It is certainly theoretically possible that greater wage and price flexibility actually leads to a lowering not just of welfare (as in the discussion above of efficiency wages) but of output. Recent cross section estimates of output variability and the likelihood of recessions suggests that greater wage flexibility is either unassociated with greater output stability or may actually contribute to an enhanced likelihood of a recession. (See Easterly, Islam, and Stiglitz [1999]).

    Labor as a Stakeholder in Corporate Governance

    Much has been written over the past seventy years, and especially over the past fifteen years, about the consequences of the separation of ownership and control;(14) in modern parlance, we say there is a principal-agent problem. Shareholders have limited ability and incentives individually to monitor fully managers; and managers have an incentive to make it more difficult for themselves to be monitored. [Edlin and Stiglitz 1995]

    A central theme of the literature on corporate governance is that there are differences in interests among the various stakeholders in the firm. One strand of literature has argued that there is a variety of mechanisms by which a greater congruence can be obtained, e.g., by making workers partial owners, as under ESOPs, or by making banks also equity holders. Unfortunately, these same practices often lead to conflicts of interest; a bank which is also an equity holder may have an incentive to make an excessively risky loan, partially at the public expense (as a result of the government guarantee to depositors).

    There are several advantages to bringing workers within the fold of corporate governance beyond enhancing this congruence of interests. First, the sharing of information may lead to less conflict; under some theories, strikes are a result of imperfections of information—strikes are a costly way of conveying information between the parties. If firms have to disclose the same information to workers as they do to other board members, then the credibility of that information is enhanced; workers are more willing to accept a firm's claim that it cannot pay higher wages without threatening the viability of the firm.

    There are also arguments that worker participation in decision making, even if only through representatives, may increase the sense of "fairness" of any decisions made, and fairness in turn can affect worker morale and productivity. [See Akerlof and Yellen 1988]

    Second, workers are often in a better position to monitor the firm than are creditors, since they are continuously on the spot. They can verify—or challenge—management claims about what is actually happening within the firm. It is for this very reason that management may resist having worker participation. It may limit the power that management exerts (and its rents) by reducing the asymmetries in information.

    On the Importance of Residual Control Rights: Lessons from the East Asia Crisis

    The issues with which I have been concerned here are not just theoretical niceties, but can have take on first order importance, not only in affecting the prospects of particular individuals in particular firms, but in determining the nature of the market equilibrium. This was brought home forcefully in the context of the recent East Asia crisis. Standard contracts in place gave all (or at least a predominant proportion) of residual control rights to shareholders. Consider first the situation prior to the crisis. When workers came to work at a firm, they typically did not know the full risk profile of the firm, in particular, its uncovered foreign exchange exposure. Though this risk represented a potential real cost to the worker, he would benefit little if any from any returns. Had workers had a significant voice in management, they would have strongly argued against the firm taking on such a position, unless the firm provided adequate job security (severance benefits) to the workers. As it was, the workers had no say.

    Next, consider how firms responded to the unanticipated macro-economic events (the soaring interest rates, the falling exchange rates, and declining demand). Workers' interests would probably have been maximized if the firms had gone into a speedy chapter 11 reorganization; but this would hurt creditors, and it would have hurt equity owners (the presumed decision makers) even more. In this case, there was a natural alliance between international creditors and equity owners (supported by one of the international financial institutions, whose officials, in trying to ward off such a default, spoke repeatedly of the sanctity of contracts, paying little heed to bankruptcy being a core part of capitalism, and the at least implicit contracts between workers and their firms that were being torn up, all in the name of protecting creditor rights). The point is that because workers were not represented—either at the firm decision making counsels, or within the international institutions that were attempting to address the crisis—the outcomes were clearly not in workers' interests, and were probably not even efficient. The costs to workers were not adequately weighed against the benefits to creditors or equity holders.

    Principal-Agent Problems and Worker Involvement

    The problem of corporate governance is, of course, nothing more than a manifestation of the general principal agent problem. I want to spend a few moments focusing on agency problems in industrial relations systems. The competitive model has some rough plausibility in a market populated by small firms. In our Jeffersonian tradition, we celebrate the family farm, the owner-operated shop, and the small independent craftsmen. In terms of the principal-agent relation, all these cases are characterized by the unity of principal and agent, i.e., the people are self-employed (no agency costs). Here the assumption that the principal and agent have the same knowledge and incentives is obviously true, but how should that assumption be generalized to larger firms? For larger firms, the assumption that the principals (e.g., far-flung shareholders) are able to enforce their incentives on the agents (managers and workers in the firm) and monitor the behavior of the agents becomes a rather heroic assumption. Yet without that assumption, the basic theorem fails. [Greenwald and Stiglitz 1986] How in the actual world can the virtues of the limit case of self-employed individuals be best generalized to multi-person firms?

    Let us consider some of the informational and incentive problems inherent in the employment relation and in other agency relationships. The shareholders are principals relative to the managers as their agents, and the managers can be seen as principals relative to the workers as agents. In such pyramided principal-agent relationships, there may be many gaps in monitoring and incentives between the shareholders and workers: managers have imperfect incentives to ensure that the two are well aligned.(15)

    One approach to the effort question was developed in efficiency-wage theory as an explanation for involuntary unemployment in otherwise flexible labor markets.(16) Employers are assumed to fire workers who are caught shirking (working at chronically low effort levels) but if labor markets clear at the prevailing wage, then workers will be able to readily find new employment at the prevailing wage. Assuming effort is costly and the probability of being detected shirking is low, workers would tend to operate at low effort levels. Employers may pay more than the market-clearing wage, i.e., an efficiency wage, in order to give workers something extra to lose if they are fired for working at low effort levels. At that efficiency wage, there would be more job-seekers than vacancies resulting in involuntary unemployment. A higher level of unemployment would make dismissal more costly. Since fired workers would join the unemployment queues, there would be an incentive for work at a high effort level so as not to risk being detected shirking and fired.

    For purposes of this paper, three observations concerning efficiency wage theories are relevant:

    A. The extent of the agency problem depends on the distribution of wealth. Agency problems are less important in societies in which wealth is more equally distributed.

    B. When there are agency problems, the market is, in general, not even constrained Pareto efficient. Costly supervision implies that workers are imperfectly monitored; when bad performance is discovered, a worker is fired, but if firing is to have a cost, there must be surplus associated with holding the job; there may be unemployment in equilibrium. In this case, firms have no incentive to provide job benefits (severance pay, grievance settlement mechanisms) which reduce the cost of leaving—even if such benefits are welfare enhancing. Indeed, the firm has an incentive to look for ways that increase the costs of separation, because in doing so it can lower wages.

    C. This use of an efficiency wage to elicit high effort is best suited to what is typically called the low involvement or "low road" workplace(17) since it assumes a credible threat to fire shirking workers and it assumes no other motivation for high effort levels.

    A high involvement or "high road" workplace (as in what we might call the "Japanese-style" workplace) uses quite different methods to elicit high effort from the workers. A high involvement labor participation program deals not just with economic variables such as pay (perhaps performance-related) and benefits but with a range of other factors such as worker-team involvement in decision-making and control on the shop floor or in the office. At the risk of oversimplifying the psychology, the idea is to increase each worker's involvement in and identification with the firm so that there will be some unification of agent and principal and a resulting tendency for higher effort. The high involvement workplace, by approaching principal-agent unification (however in quite a partial manner), provides a different method of eliciting high effort than the "risk of unemployment" motivation emphasized in the "low road" approach.

    In 1990, Laura Tyson and David Levine [1990] surveyed 43 empirical studies on the connection between participation and productivity. They found that the effect of worker participation on productivity was usually positive though sometimes small or statistically insignificant—but almost never negative. The effect improves the more the participation was close to the shop floor or office.

    Several dozen new studies have been conducted since then, several of which have particularly strong research designs and data quality. Their conclusions reinforce the earlier findings: A small-scale employee involvement plan, just as a small amount of training or a modest change in pay systems, may have some beneficial effects, particularly in the short run. Furthermore, a system of high involvement, strong rewards, and high levels of skill and information, integrated with a corporate strategy that relies on front-line employees' ideas and creativity, is capable of impressive improvements in organizational performance. [Levine 1995, 81]

    The literature on ESOPs and other forms of employee-ownership has generally, but not always, found a positive relationship between ownership and performance. But when the ownership is coupled with genuine participation, the positive relationship is quite clear.(18)

    Two Industrial Relations Systems: the Low Road and the High Road

    The contrasts between the low and high involvement workplaces are part of a larger story about the interlocking attributes of different types of systems (see Aoki [1994]). Indeed, one way to look at the East Asian crisis is as the turmoil that occurs at the interface between two systems just as an earthquake is produced by the collision and rubbing of tectonic plates. In a system of information-rich and stable but highly leveraged relationships between firms and financiers, distress is handled with understanding and lenience on the part of the lenders. The high trust in the firm-financier relationship pairs together with the high leverage as part of a workable system.

    But when the same firms start to become indebted with arm's-length short-term borrowing, there will be little slack in the face of distress and the high leverage may lead to crisis. Low trust and arm's-length finance relationships need to be paired with lower debt-equity ratios to provide more flexibility under distress. The point is not that one system or the other is "better" but that an unwise mixture of the two systems may be quite prone to crisis.

    Yet the world changes; new circumstances arise. Each system must find ways to adjust their set of interlocking attributes to address the new realities and yet avoid unstable mixtures prone to crisis. Just to be even-handed, let me mention the opposite sort of problem when a low trust system adopts a sub-system from a high trust system. Many firms in the US and Europe have wanted to cut inventory costs and to foster the problem-solving induced by the just-in-time inventory system. But the JIT inventory system interlocks with rather cooperative labor relationships both in the firm and in the suppliers and truckers who supply the parts "just in time." In a more confrontational labor environment, the JIT system is rather unworkable.

    I will proceed by outlining the two systems introduced as two ways of addressing the effort problem in the agency relationship and the more general problem of opportunistic behavior in contractual relationships. The two stylized industrial relations systems will be characterized with generic characteristics (e.g., low trust and low involvement of the low road versus high trust and high involvement of the high road) rather than geography—even though one system is commonly thought to be Anglo-American and the other is identified with Japanese and German systems.

    In terms of Hirschman's exit-voice distinction [1970], dissatisfaction in a relationship leads in a low trust/involvement system to exit and the search for a better partner. In a high trust/involvement system, contractual relationships are more stable and long-term. Each partner is expected to have higher commitment to the relationship and more trust that the partner will not act opportunistically.(19) In the high road workplace, dissatisfaction would be addressed through various voice mechanisms (e.g., collective bargaining, grievance procedures, and labor-management committees) rather than termination and exit. To paraphrase an old advertising slogan, the partners would "rather fight [to resolve the differences] than switch.

    High trust is developed between workers and managers by managers exercising the self-restraint to not use their power to enrich themselves and to take advantage of the workers. On their side, the workers choose to be cooperative without feeling that they are exposing themselves to being opportunistically exploited by self-aggrandizing managers. That mutual cooperativeness in the high trust management-labor relationship is the basis for high "X-efficiency."(20) In a high trust and involvement environment, the genuine participation of the workers leads to their increased buy-in to the goals of the immediate work group, if not to some goals of the broader enterprise. As a result of this socialization into the enterprise, the worker tends to identify with and to affect the goals of the whole effort. Instead of better threats and monitoring to reduce opportunistic behavior in the agency relation, the high trust/involvement system strives towards identification of principals and agents.(21) In a 1991 symposium on "Organizations and Economics," Herbert Simon emphasized the importance of identification.

    Although economic rewards play an important part in securing adherence to organizational goals and management authority, they are limited in their effectiveness. Organizations would be far less effective systems than they actually are if such rewards were the only means, or even the principal means, of motivation available. In fact, observation of behavior in organizations reveals other powerful motivations that induce employees to accept organizational goals and authority as bases for their actions. [The] most important of these mechanisms …[is] organizational identification. [Simon 1991: 34]

    Moreover the greater congruence between the goals of the agents and the goals of the firm can be achieved by adjusting both instead of only the former.(22)

    The body of employees is, together with the body of shareholders, explicitly or implicitly recognized as a constituent of the firm, and its interests are considered in the formation of managerial policy. [Aoki 1987, pp. 283-4]

    We have thus seen how the system tries to generalize to larger enterprises the virtues of the family farmer, small producer, or shopkeeper who is self-employed. In doing so, we have seen several different levels of analysis:

    a) Implicit contracts—reputational relationships, with incomplete contracts, with adjustments in response to changing circumstances based on voice and trust—may be more effective than explicit contracts with, say, one side having all "residual" rights to control and all residual income.

    b) Adaptations in workers' preferences, identification, may be more effective in eliciting desired behavior of workers than incentive based contracts; how to achieve such identification is one of the major challenges facing management. Profit sharing, which in terms of standard incentive theories, may be fairly ineffective, may still be effective because of its effects in facilitating identification. Identification can also be facilitated by

    c) Firms convincingly changing their stated objectives as going beyond simply profit maximizing, to include the welfare of their workers, not only as means to end, but as ends themselves.

    The following table tries to concisely give the flavor of the two systems and how their internal interlocks might be played out in different markets.(23)

    Low RoadHigh Road
    Inducement to high effortHigh unemployment and efficiency wageHigh involvement induces effort even
    with low unemployment
    CompensationContractual wagesWages plus profit sharing
    Wage differentialsHigh differentials as incentive for
    individual advancement
    Low differentials for increased group
    solidarity and cohesiveness
    Employment securityLow: dismissal is credible threat for
    High security to promote identification
    with enterprise
    Training costsPaid by individual to increase
    Paid by firm as long-term human capital
    Macro-environmentCan adjust to and contribute to larger
    recessions with layoffs
    Works better with and contributes to
    fewer and smaller recessions by avoiding

    RelationshipArm's-length, market-oriented, and
    Long-term relation based on
    commitment, trust, and loyalty
    ProductStandardized (to foster competition)Customized to buyer or seller
    Curb to opportunismExit and competitionVoice, commitment, and trust

    RelationshipArm's-length and market-oriented
    Long-term relational finance
    Time perspectiveShort-term since hard-to-monitor; human
    capital investments downplayed
    Long-term and patient to reap return to
    human capital investments
    Debt/equity ratiosNeed low D/E ratio to provide flexibility
    in face of unforgiving market
    Can have higher D/E ratios with
    involved, more flexible workers
    Low costs of equityLow costs since no sharing of income or
    control rights with workers
    Lower costs for internal equity since
    workers already share some income &
    control rights

    Development Strategy for Labor: From the Low Road to the High Road

    In the developing world, the picture of a market economy is often based on images of the "dark satanic mills" of 18th and 19th century England. Market-based development was conceptualized using what we have called the "low road" system of industrial relations. The modern experience in Japan and other East Asian countries shows that there is another model that more closely approximates the high road system. Until the East Asia crisis struck, there was, at least among some circles, the view that the "high road" had distinct advantages over the low road: macro-stability would be greater, productivity growth higher, worker morale stronger. I worry that one of the more adverse consequences of the East Asian crisis may be the abandonment of the "high road," as firms are being encouraged to break long standing implicit contracts with workers, to "downsize" in response to the new economic realities—even if downsizing implies forcing long term workers into unemployment. Such long standing relationships are viewed as contributing to market rigidities, impeding the quick adjustments needed in the nimble world of modern globalization.

    To be sure, excessive labor market rigidities (almost tautologically) can have adverse effects. But long term social contracts between firms and their workers may make them more accepting of—and more promoting of—change and progress. Indeed, the breaking of the social contract and the undermining of social capital is increasingly being given "credit" for the huge decreases in productivity in the former Soviet Union.(24) But given imperfections of information (e.g. between workers and firms) arm's length market based relationships will lead to an underinvestment in firm specific human capital (relative to the first best optimum) and higher labor turnover. (See Arnott and Stiglitz [1985].) (25)

    The Need for–and Limitations of—Collective Action

    I have stressed these market failures, often implicit in the discussions of labor relations, for an important reason. In the absence of the kinds of imperfections noted earlier, firms would have an incentive to have the "optimal" amount of worker participation in decision making—there would be no need for government intervention in governance. If the evidence that the "high road" is as compelling as many seem to believe, firms will move in that direction.

    But the market failures depicted earlier explain why they may not move as much or as fast as is socially desirable, and provide a clear rationale for collective action. There is at least the possibility that government interventions in the labor market—through regulations affecting working conditions, collective bargaining, and more broadly workers' rights—will bring about redistributions that might not otherwise be achieved. Such interventions may, under certain circumstances, actually be Pareto improvements.

    But I hasten to add that there is a delicate balance: excessively strong unions can through collective action "hold up" the rest of the economy, reduce product market competition, and interfere in other ways with the efficiency of the economy. This is particularly problematic in areas in which there is a natural monopoly, or a government created monopoly or near monopoly. Wage increases can be passed on to consumers, and workers in these industries have in country after country been able to use their market power to extract wages far in excess of their opportunity cost. When the service is publicly provided—such as education—market discipline may too be limited. Though eventually voters may raise concerns about public employees being paid wages considerably in excess of market wages, the process is a slow one, and before the political process responds, considerable rents may be extracted from the public. Of particular concern are those instances in which, in order to maintain their rents, unions attempt to suppress competition, as many would argue has been happening in the United States, with unions' vehement opposition to vouchers.

    Is there a role for collective action within the public sector?

    I was going to ask, more broadly, what is the justification for unions in the public sector? The arguments presented earlier suggested that profit maximizing private firms would act to exploit their monopsony power over workers, or otherwise engage in wage setting policies that ignored the externalities that those policies exerted on others. Clearly, the government need not do so; with good economic advisers, they would be tamed against abusing any monopsony power and act in a way that is acutely sensitive to externalities. Indeed, correcting externalities is one of the major rationales for collective action. This reasoning would suggest that while there is a need for government-enforced collective bargaining rights in the private sector, such rights should not extend to the public.

    But there is a strong qualification to this argument. Principal agent problems arise in the public sector just as they do in the private. The manager of a public school may not act in a way that fully reflects the public interest. He is likely to see his job as producing high quality education at a minimum cost—just as the manager of a private school would—and as such he would have every incentive (absent collective bargaining) to keep wages as low as possible. More broadly, political control mechanism are far from perfect—far less effective than market control mechanisms. In such cases, the exercise of voice—through unions—may be particularly important. (See Hirschman [1970].)

    Systemic problems

    The case for redressing the imbalance in bargaining power seems clear, but it provides only part of the rationale for the role for government in regulating corporate governance, especially given the magnitude of managerial discretion and the role that management plays in deciding on corporate governance. The fact that there exist quite different systems of corporate governance (those with and without a large role for workers) suggests(26) that there may be multiple equilibria, in which case government can play a role in moving the economy from one equilibrium to another.(27)

    Norms are established that govern "appropriate" behavior—not only the extent of consultation, but appropriate compensation differentials between management and workers. (Does one really believe that the ratio of managerial productivity to worker productivity differs so markedly across countries?) In a society in which workers participate on a regular basis in decision making, any firm that took away that right would be castigated. But in a society in which workers do not regularly participate in decision making, granting that right might not serve to make the firm that much more attractive; indeed, a self-selection process may be set in motion, whereby workers who feel most strongly about participation—and who in other ways are the most militant—may be attracted to the firm. Similarly, in societies that value long term relations, the signal sent by a worker leaving his firm is markedly different from that sent in an economy like the United States.(28)

    Development as Democratic Transformation

    Finally, I would like to view this developmental strategy for labor within a broader framework for development. In my Prebisch Lecture [1998], I emphasized the concept of development as transformation.

    Development represents a transformation of society, a movement from traditional relations, traditional ways of thinking, traditional ways of dealing with health and education, traditional methods of production, to more "modern" ways. For instance, a characteristic of traditional societies is the acceptance of the world as it is; the modern perspective recognizes change, it recognizes that we, as individuals and societies, can take actions that, for instance, reduce infant mortality, increase lifespans, and increase productivity.

    If a change in mindset is at the center of development, then it is clear that attention needs to be shifted to how to effect such changes in mindset.(29) Such changes cannot be "ordered" or forced from the outside , however well-intentioned the outsiders may be. Change has to come from within. The kinds of open and extensive discussions that are central to democratic processes are, I suspect, the most effective way of ensuring that the change in mindset occurs not only within a small élite, but reaches deep down in society.

    Change is also often threatening–and sufficiently risk-averse individuals are willing to pass up opportunities for expected gain to avoid the downside risks. Democratic and participatory processes involving labor unions and other social organizations ensure that these concerns are not only heard, but addressed; as a result, these processes dissipate much of the resistance to change. Consider an example that is particularly relevant in a time of globalization. As one who supports lowering trade barriers, I am nonetheless dismayed to note that all too often ardent free-trade advocates cavalierly dismiss the opponents, including those who stand to lose by free trade, and refer to them as "special interests" trying to protect their existing "rents". But among those hurt by trade reforms will be many who will lose their jobs. And if the society lacks an adequate safetynet, the unemployed worker therefore risks true impoverishment, with disastrous effects on the lives of all family members. What is of concern is not just the loss of "rents", but the loss of a family's livelihood. Inclusive democratic processes involving unions and other popular organizations make it more likely that these legitimate concerns will be addressed.

    We should be clear: workers in much of the world have grounds for suspicion. Capital market liberalization in East Asia did not bring the benefits that were promised, except to a few wealthy individuals. It did impoverish many—both through lower wages and increased unemployment. Worse still, workers have seen decisions that affect their lives and livelihoods being seemingly forced upon their countries, with hardly a nod towards the concerns of the workers, apart from sermons about the virtues of bearing pain. I believe, for instance, that there is some chance that some of the disastrous economic decisions that were made in responding to the East Asian economic crisis would not have occurred had workers had a voice (let alone a voice commensurate with their stake in the outcome) in the decision making.(30) And even if similar decisions had been made, at least workers would have felt that they had had their say.

    Thus, I would argue that economic democracy is essential to effect the systemic change in mindset associated with the democratic transformation, and to engender policies that make change—which is at the center of development—more acceptable. And because labor and other affected social groups have had a voice in shaping the changes, in making them more acceptable, change is likely to be accepted or even embraced, rather than reversed at the first opportunity.

    The economic benefits of workplace democracy are, however, as I have already suggested, more pervasive than just the acceptance of change: there is a growing literature arguing that participation in decision making increases efficiency.(31) Changes in technology would be expected to be associated with changes in the degree and nature of worker participation, as the efficiency benefits and economic costs of participation change. New information technologies and production modes hold open the promise of greater worker participation—just as they hold open the promise of more effective citizen participation in public governance.(32)

    Toward Economic Democracy

    So far, I have largely cast the analysis in traditional economic terms. I have argued that, given traditional objectives (Pareto efficiency) and traditional modeling of individual behavior, simply taking into account imperfections and costs of information, and the imperfections of mobility and asymmetries in bargaining to which they give rise, there is a rationale for government intervention in labor relations.But I want to put forward a stronger hypothesis. We care about the kind of society we live in. We believe in democracy, regardless of whether it increases economic efficiency or not.(33)

    I would argue that democracy entails far more than majority voting.(34) Indeed, there is a whole tradition that identifies "government by discussion" as key to democracy. (35) Democratic processes must entail open dialogue and broadly active civic engagement, and it requires that individuals have a voice in the decisions that affect them, including economic decisions. Thus, we can speak of industrial or economic democracy in the workplace—where unions play a key role—and local democracy at the community level, as well as democracy at the national level.

    A society in which there is a widespread view of class conflict between workers and capitalists is fundamentally different—and functions fundamentally differently—from a society in which there is perceived to be a wider congruence of interests. A society in which large portions of the labor force are marginalized, are treated as if their voice does not count, who see their opportunities for advancement highly restricted, will function in ways that are fundamentally different from a society in which there is perceived to be greater respect for every individual and greater attempts at comprehensive inclusion.

    Economic democracy is thus an essential part of a democratic society. The limits and bounds of economic democracy are evolving, just as democracy itself is changing. Though democracy has a long tradition—in the West, it dates back at least to the Greek city-states—even political democracy has been slow to evolve. It was only in the century just ended that universal political suffrage became the norm. Many countries have been slow to grant those basic rights—of a free press, free speech, the right to organize to pursue common objectives (both in general, and for workers in particular)—that are so necessary for an effective democracy. Many governments continue not to recognize the people's fundamental "right to know," pursuing secrecy well beyond the domain where it is needed for national security.

    There have been comparably great strides in economic democracy. Today, management is more willing to listen to the concerns of workers—they do not view this as an intrusion into managerial prerogatives. Even language is changing, as one speaks of partnerships, teams, community. One need not be pollyannaish—believing that there is complete congruence of interests—to believe that such a change in language represents a fundamental shift in mindset, a move towards greater openness, to delineating more clearly the sources of conflict, clarifying the asymmetries of bargaining power that arise from costs of labor mobility, limited worker resources, and asymmetries of information.

    Democracy is also fragile. Repeatedly, we have seen high levels of social disorder lead to calls for strong (read "anti-democratic") government to restore the basic foundations of law and order without which individuals cannot live and work together. We have seen how economic policies, and the manner in which they are adopted, can either contribute to social cohesion or to social disorder. The world has experienced economic crises of increasing frequency and severity. There is a growing consensus on the causes of the crises, and on the policies that must be adopted to reduce their frequency and severity and to mitigate the consequences, e.g., by developing stronger safety nets. But there is no safety net that can fully replace the security provided by an economy running at full employment. No welfare system will ever restore the dignity that comes from work. It is imperative that countries not only work to put into place policies that prevent crises and minimize their magnitude and adverse consequences, but respond to these crises in ways that maintain as high a level of employment as possible. Too often, in advising countries on policies that they should pursue, the focus has been too narrow. While potential efficiency benefits were stressed, the downside risks were given short shrift; worse still, little attention was put on sequencing—ensuring that the country had in place the institutions that would enable the country (and especially the most vulnerable workers within it)—to bear the risks. (36) And in exposing the country, and its workers, to these risks, we not only put at risk the lives and livelihoods of the workers, but more fundamentally the systems of economic and political democracy.

    Concluding Remarks

    As we end the millennium and begin another, it is time to view the issues of labor relations through new lenses—and begin a shift in the prevailing paradigm. Few writing a history of capitalism in the United States would venture that organized labor did not play an important role not only in restructuring the relationships between workers and firms, partially redressing an imbalance of power, but also in improving living standards. Critics who say that these changes would have come on their own, simply as a result of higher GNP, are simply not credible.

    But the world today is markedly different from the world 75 or 50 years ago. The statistics suggest that unions are playing a far less important role within the private sector than they did in the years immediately following world War II. Yet that does not mean that issues of labor relations have disappeared. Rather, the grounds have shifted—for instance, to issues concerning the role of workers in ownership and governance. There may be a need for government to facilitate this shift in economic organization, just as it did earlier in the century, in facilitating the growth of unions. Many of the developing countries—some of which are just emerging from a history of feudal relations—face more traditional problems of redressing fundamental imbalances of power. Those of us in the business of dispensing development advice must be aware of the social, political, economic and historical context in which that advice is given: advising countries to have more flexible labor markets may be tantamount to telling them to give up hard wonadvances in labor standards. And even the welfare gains may be problematic, once the social costs of the risks imposed and the adverse macro-economic effects described earlier are taken into account. The streets of Seattle bear testimony to the sense of frustration that many within the developing world feel about how the international community has addressed their concerns.

    But even more fundamental than the issues of economic efficiency are those concerning economic democracy: the kind of society we are attempting to create. There is more that we can do than just following the dictum of "do no harm"—though some might argue that that would, by itself, be going a long way. While globalization provides new challenges for sustainable democratic development, it also offers new opportunities to loosen the fetters of the past and to promote the democratic processes essential for long-run success. By becoming advocates of stronger workers' rights and representation, at every level—from the workplace, to the local, regional, and national level, to the international level—I believe that we can achieve much more than improvements in efficiency. Labor unions and other genuine forms of popular self-organization are key to democratic economic development. That is why today, the World Bank supports the Labor Standards of the ILO, including the rights to organize and collectively bargain.


    (1) See, e.g. Stiglitz [1998].

    (2) See Freeman and Medoff [1984] and Kochan and Osterman [1994].

    (3) "Industrialism has, wherever it has appeared, produced some form of trade unionism and some kind of exclusive organization of the working class. In all industrialized countries in Europe, trade unions and the working class movement connnected with them have formed the most lively defenders of democracy. Even in Eastern countries like Japan and India industrialism has produced the need fo rtrade unions and trade unions are the beginning of democracy." [Lindsay 1943, 192]

    <4> But operating not just through the labor market, as in classical economics, but through the product market as well.

    (5) That is, taking into account the costs of obtaining information or creating markets. See Greenwald and Stiglitz [1986].

    (6) See Braverman and Stiglitz [1986].

    (7) See Easterly, Islam, and Stiglitz [1999].

    (8) See Kuttner [1999] for more discussion.

    (9) See Philmore [1982] for elaboration.

    (10) Robert Solow in his own way seeks to provide succor to the neoclassical faithful: "I want to emphasize to economists that it is not a betrayal of the structure of economic theory generally to admit the likelihood that labor is a peculiar sort of commodity and the labor market correspondingly a peculiar sort of market." [1990, p. 30]

    (11) One of the important insights to emerge from the economics of information is that the economy is rife with convexities, so that small costs (or more generally small perturbations) can have large effects. See, e.g. Stiglitz [2000].

    (12) See Diamond [1971]. Under some circumstances, the wage may be somewhat higher than the monopsony wage, or the equilibrium may be characterized by a wage distribution. See Stiglitz [1985a, 1987].

    (13) For an early discussion of the inefficiencies associated with labor turnover, see Stiglitz [1972]. The problem is that imperfections of information and incomplete contracting make it impossible to distinguish between efficient quits, e.g. when the worker is badly matched with the firm, as a result of basic differences in preferences, and inefficient quits. See Arnott and Stiglitz [1985].

    (14) See Berle and Means [1932] and Stiglitz [1985b].

    (15) These agency relationships are a major topic in information economics, since many aspects of the agents' behavior are unobservable. Many of the unobservable factors that affect labor productivity are grouped under the label "effort." Managers cannot directly observe the workers' effort level, and yet effort can have a considerable effect on productivity and profits.

    (16) See Shapiro and Stiglitz [1984] and the other papers in Akerlof and Yellen [1986].

    (17) In Shapiro and Stiglitz [1984], this was described as the "lower-paid, lower-skilled, blue-collar occupations."

    (18) See Blinder [1990], Gates [1998], and the updated reports on posted by the National Center for Employee Ownership in Oakland, California.

    (19) See Goldberg [1980] for a treatment of the contrast between relational contracting and arm's-length contracting.

    (20) Where we might take X = effort. See Leibenstein [1984, 1987].

    (21) These considerations raise a host of important questions that cannot be pursued here. See McGregor's classic treatment of "Theory Y" [1960] and the more recent literature on intrinsic motivation such as Lane [1991], Frey [1997], and Kreps [1997].

    (22) This formulation does not make clear whether the firm is supposed to abandon the goal of profit maximizing, or only that it has to recognize that to achieve that goal, it must act as if it incorporated the concerns of workers within its own objectives. See Akerlof and Kranton [1999].

    (23) See Levine and Tyson [1990]; Levine [1995] for many of the same points and see Clark [1979] or Dore [1987] for similar tables comparing Anglo-American-type and Japanese-type firms.

    (24) See Stiglitz [1999a, 1999b] and Hussain, Stern, and Stiglitz [1999].

    (25) Similar arguments have been put forward to suggest arm's length financial relationships associated with securitization may be less efficient than "relationships" associated with banking. See, e.g. Stiglitz [1992] and the references cited there.

    (26) It is only suggestive, since the legal environment differs in other respects as well.

    (27) The role of government in moving an economy from one equilibrium (that is Pareto inferior) to another is illustrated well by Basu's discussion of child labor. He argues that there may be a "low level equilibrium trap" in which wages are low; because wages are low, families send out their children to work; and because they send their children out to work, the equilibrium wage is low. There is another high wage equilibrium, in which at the high wage, families chose not to send out their children to work. Prohibiting child labor ensures that the economy avoids the low level equilibrium trip. (See Basu [1998].) Similarly, Basu, Genicot, and Stiglitz [1999] show that there may be a high unemployment equilibria, with families sending many members into the labor market because of the risk that some will not be hired. A guaranteed family income may eliminate this "bad" equilibrium.

    (28) A similar argument has been used to show that while it may be more efficient for firms to have more extensive risk sharing arrangements with their workers (Weitzman [1984, 1985, 1995]), it will not pay any firm to introduce such an arrangement. The firms that are most likely to do so are those whose profit expectations are most bleak. In effect, profit sharing arrangements are analogous to the firm issuing shares to its workers, and so the adverse selection arguments for why equity markets work poorly apply here. (See Greenwald, Stiglitz, and Weiss [1984]) By forcing all firms to have risk sharing arrangements with their workers, one dilutes, if one does not eliminate, the adverse signaling effect.

    (29) "All [vicious development] circles result from the two-way dependence between development and some other factor, be it capital or entrepreneurship, education, public administration, etc. But the circle to which our analysis has led us may perhaps lay claim to a privileged place in the hierarchy of these circles inasmuch as it alone places the difficulties of development back where all difficulties of human action begin and belong: in the mind." [Hirschman 1958, 11]

    (30) See Stiglitz [1998c].

    (31) See Blinder [1990] and Levine [1995].

    (32) See Stiglitz [1999c].

    (33) Within the development literature, there is a large and controversial literature addressing the issue of the two-way relationship between growth and democracy. See, e.g., Knack and Keefer [1997], World Bank [1997], Stiglitz [1999d].

    (34) "In theory, the democratic method is persuasion through public discussion carried on not only in legislative halls but in the press, private conversations and public assemblies. The substitution of ballots for bullets, of the right to vote for the lash, is an expression of the will to substitute the method of discussion for the method of coercion." [Dewey 1939, p. 128]

    (35) See, for example, John Stuart Mill [1972 (1859)], Walter Bagehot [1948 (1869)], James Bryce [1959 (1888)], John Dewey [1927, 1939], Ernest Barker [1967 (1942)], Frank Knight [1947], and Charles Lindblom [1990].

    (36) Thus, a large literature now bears testimony to capital and financial market liberalization, whatever the efficiency benefits that might be derived from them (and some recent literature has even questioned that—see Stiglitz [1999e] and Rodrik [1998]), unambiguously contributing to economic volatility and an increased probability of financial and currency crises and recessions. See Demirgüí§-Kunt and Detragiache [1997] and Easterly, Islam and Stiglitz [1999].


    Akerlof, George 1984. An Economic Theorist's Book of Tales. New York: Cambridge University Press.

    Akerlof, George and Rachel Kranton. 1999. Economics and Identity. Draft, Presented at World Bank Summer Research Workshop, July.

    Akerlof, George and Janet Yellen, Eds. 1986. Efficiency wage models of the labor market. Cambridge: Cambridge University Press.

    ________.1988. Fairness and Unemployment. American Economic Review, 78(2), May: 44-49.

    Aoki, Masahiko 1987. The Japanese Firm in Transition. In The Political Economy of Japan. Kozo Yamamura and Yasukichi Yasuba Eds. Stanford: Stanford University Press.

    __________ 1994. The Japanese Firm as a System of Attributes: A Survey and Research Agenda. In The Japanese Firm: Sources of Competitive Strength. Masahiko Aoki and Ronald Dore Eds. Oxford: Oxford University Press: 11-40.

    Arnott, R. and J.E. Stiglitz. 1985. Labor Turnover, Wage Structure & Moral Hazard: The Inefficiency of Competitive Markets. Journal of Labor Economics, 3(4), October: 434-462.

    Arrow, K.J. and G. Debreu 1954. Existence of an Equilibrium for a Competitive Economy. Econometrica. 22: 265-290.

    Arrow, K.J. and F.H. Hahn 1971. General Competitive Analysis. San Francisco: Holden-Day.

    Bagehot, Walter. 1948 (1869). Physics and Politics. New York, Knopf.

    Barker, Ernest. 1967 (1942). Reflections on Government. London: Oxford University Press.

    Basu, Kaushik. 1998. Child Labor: Cause, Consequence, and Cure, with Remarks on International Labor Standards Policy Research Working Paper No. 2027 Washington, DC: World Bank.

    __________, Garance Genicot and Joseph E. Stiglitz. 1999. Household Labor Supply, Unemployment, and Minimum Wage Legislation Policy Research Working Paper No. 2049 Washington, DC: World Bank.

    Berle, A. and G. Means 1932. The Modern Corporation and Private Property. New York: MacMillan Company.

    Blinder, Alan S., Ed. 1990. Paying for Productivity. Washington DC: The Brookings Institution.

    Bowles, Samuel 1985. The Production Process in a Competitive Economy: Walrasian, Neo-Hobbesian, and Marxian Models. American Economic Review. 75 (March): 16-36.

    Braverman, A. and J. Stiglitz 1986. Landlords, Tenants and Technological Innovations. Journal of Development Economics. 23(2 October) 313-32.

    Bryce, James. 1959 (1888). The American Commonwealth. New York: G. P. Putnam's Sons.

    Christ, Carl F. 1975. The Competitive Market and Optimal Allocative Efficiency. In Competing Philosophies in American Political Economics. John Elliott and John Cownie Ed. Pacific Palisades, CA: Goodyear: 332-338.

    Clark, R. 1979. The Japanese Company. New Haven: Yale University Press.

    Coase, R. 1937. The Nature of the Firm. Economica, 4, 386-405.

    Debreu, G. 1959. Theory of Value. New York: John Wiley & Sons.

    Demirgüí§-Kunt, Asli and Enrica Detragiache. 1997. The Determinants of Banking Crises: Evidence from Industrial and Developing Countries Policy Research Working Paper No. 1828. Washington, DC: World Bank and International Monetary Fund.

    Dewey, John. 1927. The Public and Its Problems. Chicago: Swallow Press.

    __________ 1939. Freedom and Culture. New York: Capricorn.

    Diamond, P. 1971. A Model of Price Adjustment. Journal of Economic Theory, 3(2), June: 156-168.

    Dore, R. 1987. Taking Japan Seriously. Stanford: Stanford University Press.

    Easterly, W., R. Islam, and J.E. Stiglitz. 1999. Shaken and Stirred: Volatility and Macroeconomic Paradigms for Rich and Poor Countries. Michael Bruno Memorial Lecture, given at the XII World Congress of the IEA, Buenos Aires, August 27.

    Edlin, A. and J.E. Stiglitz. 1995. Discouraging Rivals: Managerial Rent-Seeking and Economic Inefficiencies. American Economic Review. 85(5): 1301–12.

    Farrell, J. 1987. Information and the Coase Theorem. Journal of Economic Perspectives, 1(2), Fall: 113-129.

    Freeman, Richard and James Medoff 1984. What Do Unions Do? New York: Basic Books.

    Frey, Bruno 1997. Not just for the money: An economic theory of personal motivation. Cheltenham: Elgar.

    Gates, Jeff 1998. The Ownership Solution. Reading: Addison-Wesley.

    Goldberg, V. 1980. Relational exchange: economics and complex contracts. American Behavioral Scientist 23 (3, January/February): 337-352.

    Greenwald, Bruce and Joseph Stiglitz 1986. Externalities in Economics with Imperfect Information and Incomplete Markets. Quarterly Journal of Economics. 101(2), May: 229-264.

    __________ 1988. Pareto Inefficiency of Market Economies: Search and Efficiency Wage Models. American Economic Review. 78 (2 May): 351-5.

    __________ 1993. Financial Market Imperfections and Business Cycles. Quarterly Journal of Economics, 108(1), Feburary: 77–114.

    Banks versus Markets as Mechanisms for Allocating and Coordinating Investment. In The Economics of Cooperation: East Asian Development and the Case for Pro-Market Intervention, J.A. Roumasset and S. Barr (eds.), Westview Press, Boulder, 1992, pp. 15-38. (Presented at Investment Coordination in the Pacific Century: Lessons from Theory and Practice Conference, given at the University of Hawaii, January 1990.)

    Greenwald, Bruce, Joseph E. Stiglitz and Andrew Weiss. 1984. Informational Imperfections in the Capital Markets and Macroeconomic Fluctuations. American Economic Review, 74(2), May: 194-199.

    Hicks, J. 1936. Mr. Keynes' Theory of Employment. Economic Journal, 46(182), June: 238-253.

    __________ 1937. Mr. Keynes and the ‘Classics': A Suggested Interpretation. Econometrica, 5(2), April: 147-159.

    Hirschman, Albert O. 1958. The Strategy of Economic Development. New Haven: Yale University Press.

    __________ 1970. Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States. Cambridge: Harvard University Press.

    Hussain, A., N. Stern and J.E. Stiglitz. 1999. Chinese Reforms from a Comparative Perspective. Unpublished paper, the World Bank.

    Knack, S., and P. Keefer. 1997. Does Social Capital Have an Economic Payoff? A Cross-Country Investigation. Quarterly Journal of Economics 112(4): 1251-1288.

    Knight, Frank. 1947. Freedom and Reform. New York: Harper & Row.

    Kochan, Thomas and Paul Osterman 1994. The Mutual Gains Enterprise: Forging a Winning Partnership among Labor, Management, and Government. Boston: Harvard Business School Press.

    Kreps, David 1997. Intrinsic Motivation and Extrinsic Incentives. American Economic Review. 87: 359-65.

    Kuttner, Robert 1997. Everything for Sale: The Virtues and Limits of Market. New York: Knopf.

    Lane, Robert E. 1991. The Market Experience. New York: Cambridge University Press.

    Leibenstein, H. 1984. The Japanese Management System: An X-Efficiency Game Theory Analysis. In The Economic Analysis of the Japanese Firm. M. Aoki Ed. Amsterdam, Elsevier: 331-357.

    __________ 1987. Inside the Firm: The Inefficiencies of Hierarchy. Cambridge: Harvard University Press.

    Levine, David I. 1995. Reinventing the Workplace. Washington: The Brookings Institution.

    __________ and Laura D'Andrea Tyson 1990. Participation, Productivity, and the Firm's Environment. In Paying for Productivity: A Look at the Evidence. Alan Blinder Ed. Washington: Brookings Institution: 183-237.

    Lindblom, Charles. 1990. Inquiry and Change. New Haven: Yale University Press.

    Lindsay, A.D. 1943. The Modern Democratic State. London: Oxford University Press.

    Marshall, Alfred 1920. Principles of Economics. New York: Macmillan.

    McGregor, Douglas 1960. The Human Side of Enterprise. New York: McGraw-Hill.

    Mill, John Stuart. 1972 (1859). On Liberty. In J.S. Mill: Utilitarianism, On Liberty and Considerations on Representative Government. H. B. Acton Ed. London, J.M. Dent & Sons.

    Musgrave, R.A. 1959. The Theory of Public Finance. New York: McGraw-Hill.

    Nozick, Robert 1974. Anarchy, State, and Utopia. New York: Basic Books.

    O'Toole, James 1995. Leading Change: Overcoming the Ideology of Comfort and the Tyranny of Custom. San Francisco: Jossey-Bass.

    Philmore, J. 1982. The Libertarian Case for Slavery: A Note on Nozick. Philosophical Forum. XIV (Fall 1982): 43-58.

    Samuelson, P. 1976. Economics. New York: McGraw-Hill.

    Shapiro, Carl and Joseph Stiglitz 1984. Equilibrium Unemployment as a Worker Discipline Device. American Economic Review. 74 (June): 433-44.

    Simon, Herbert 1991. Organizations and Markets. Journal of Economic Perspectives. 5 (Spring): 25-44.

    Solow, Robert 1990. The Labor Market as a Social Institution. Cambridge MA: Basil Blackwell.

    Standing, Guy 1999. Global Labour Flexibility: Seeking Distributive Justice. London: Macmillan.

    Stiglitz, Joseph E. 1985a. Equilibrium Wage Distribution. Economic Journal, 95, September: 595-618.

    ______. 1985b. Credit Markets and the Control of Capital. Journal of Money, Banking, and Credit, 17(2), May 1985, pp. 133-52.

    ______. 1987. Competition and the Number of Firms in a Market: Are Duopolies More Competitive Than Atomistic Markets? Journal of Political Economy, 95(5): 1041-1061.

    ______. 1992. Banks versus Markets as Mechanisms for Allocating and Coordinating Investment. In The Economics of Cooperation: East Asian Development and the Case for Pro-Market Intervention, J.A. Roumasset and S. Barr (eds.), Boulder: Westview Press. 15-38.

    ______. 1998a. More Instruments and Broader Goals: Moving toward the Post-Washington Consensus (Helsinki, Finland: World Institute for Development Economics Research, United Nations University).

    ______. 1998b. Towards a New Paradigm for Development: Strategies, Policies, and Processes. Given as Raul Prebisch Lecture at United Nations Conference on Trade and Development (UNCTAD). Geneva. October 19. For speeches:

    ______. 1998c. Responding to Economic Crises: Policy Alternatives for Equitable Recovery and Development. Paper presented to North-South Institute Seminar, Ottawa, Canada, September 29.

    ______. 1999a. Whither Reform? Ten Years of the Transition. Keynote address to the 1999 Annual Bank Conference on Development Economics, Washington, DC, April 28-30.

    ______.1999b. Quis Custodiet Ipsos Custodes? Challenge, 42(6), November-December, pp. 26-67.

    ______. 1999c. Public Policy for a Knowledge Economy. Paper presented to the Department for Trade and Industry and Center for Economic Policy Research, London, UK, January 27.

    ______. 1999d. Participation and Development: Perspectives from the Comprehensive Development Paradigm. Paper presented in Seoul, Korea, February 27.

    ______. 1999e. Capital Market Liberalization, Economic Growth, and Instability. World Development. (forthcoming)

    ______. 2000. Knowledge about Knowledge: Reflections on Economics at the Turn of the Century. Quarterly Journal of Economics, forthcoming.

    Weitzman, Martin 1984. The Share Economy: Conquering Stagflation. Cambridge: Harvard University Press.

    ______. 1985. The Simple Macroeconomics of Profit Sharing. American Economic Review(December): 937-953.

    ______. 1995. Incentive Effects of Profit Sharing. In Horst Siebert, ed. Trends in Business Organization: Do Participation and Cooperation Increase Competitiveness? International workshop (Tubingen: Mohr (Siebeck)), pp. 51-78.

    Williamson, Oliver E. 1975. Markets and Hierarchies: Analysis and Antitrust Implications. New York: The Free Press.

    ______. 1979. Transaction-Cost Economics: The Governance of Contractual Relations, Journal of Law and Economics, 22(2): 233-261.

    ______. 1981. The Modern Corporation: Origins, Evolution, Attributes. Journal of Economic Literature, 19(4), December: 1537-1568.

    World Bank. 1997. Crime and Violence as Development Issues in Latin America and the Caribbean. Mimeographed. Office of the Chief Economist, Latin America and the Caribbean. Washington, DC: World Bank. Banks versus Markets as Mechanisms for Allocating and Coordinating Investment. In The Economics of Cooperation: East Asian Development and the Case for Pro-Market Intervention, J.A. Roumasset and S. Barr (eds.), Westview Press, Boulder, 1992, pp. 15-38. (Presented at Investment Coordination in the Pacific Century: Lessons from Theory and Practice Conference, given at the University of Hawaii, January 1990.)

    World Bank Economist Felt He Had to Silence Criticism or Quit (December 2, 1999)
    More Information on the World Bank

    FAIR USE NOTICE: This page contains copyrighted material the use of which has not been specifically authorized by the copyright owner. Global Policy Forum distributes this material without profit to those who have expressed a prior interest in receiving the included information for research and educational purposes. We believe this constitutes a fair use of any such copyrighted material as provided for in 17 U.S.C íŸ 107. If you wish to use copyrighted material from this site for purposes of your own that go beyond fair use, you must obtain permission from the copyright owner.


    FAIR USE NOTICE: This page contains copyrighted material the use of which has not been specifically authorized by the copyright owner. Global Policy Forum distributes this material without profit to those who have expressed a prior interest in receiving the included information for research and educational purposes. We believe this constitutes a fair use of any such copyrighted material as provided for in 17 U.S.C § 107. If you wish to use copyrighted material from this site for purposes of your own that go beyond fair use, you must obtain permission from the copyright owner.