See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/4730866

Market Socialism: A Case for Rejuvenation

Article  in  Journal of Economic Perspectives · August 1992

DOI: 10.1257/jep.6.3.101 · Source: RePEc

CITATIONS READS

97 177

2 authors:

John Roemer Pranab Bardhan

Yale University University of California, Berkeley
283 PUBLICATIONS   7,391 CITATIONS    261 PUBLICATIONS   11,044 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

How we do and could cooperate: A Kantian explanation View project

State and Development View project

All content following this page was uploaded by John Roemer on 12 January 2016.

The user has requested enhancement of the downloaded file.

Journal of Economic Perspectives—Volume 6, Number 3—Summer 1992—Pages 101–116

Market Socialism:
A Case for Rejuvenation

Pranab Bardhan and John E. Roemer

I
t is unfortunate that the momentous events in socialist countries since
1989 have persuaded many that socialism as a political, economic, and
intellectual movement is now to be dismissed as bankrupt and practically
moribund. The economic experiment that has clearly failed was characterized
by three features: (1) public or state ownership of the means of production;
(2) non-competitive, non-democratic politics; and (3) command/administrative
allocation of resources and commodities. In this paper, we will outline a feasible
economic mechanism of "competitive socialism" in which (2) and (3) are
negated—there would be competitive politics and competitive allocation of
most commodities and resources—but in a major part of the economy we do
not replace state or public ownership of the principal means of production with
traditional private ownership.
Public ownership in the narrow sense of state control of firms is not
necessary to achieve one of socialism's goals, a relatively egalitarian distribution
of the economy's surplus. We take public ownership, in a wider sense, to mean
that the distribution of the profits of firms is decided by the political democratic
process—yet the control of firms might well be in the hands of agents who do
not represent the state. What the Eastern European experience has shown is
that a system of pervasive state control of firms, plus the absence of markets,
does not work. Our claim is that competitive markets are necessary to achieve
an efficient and vigorous economy, but that full-scale private ownership is not
necessary for the successful operation of competition and markets. Contrary to
popular impression, this claim has not yet been disproved by either history or

• Pranab Bardhan is Professor of Economics, University of California, Berkeley,

California. John E. Roemer is Professor of Economics, University of California, Davis,
California.
102 Journal of Economic Perspectives

economic theory. It is the failure of both the political right and the left to
disentangle the concepts of private ownership and the competitive market that
has led to the premature obituaries for socialism.
To be sure, one standard response in some East European countries to
proposals of market socialism is that some variants of it have been tried (for
example, in Hungary since 1968) and they have failed. Kornai (1990) states it
flatly: "[T]he time has come to look this fact in the face and abandon the
principle of market socialism... . It is time to let go of this vain hope once and
for all." Yet when one studies the Hungarian socialist reform process over the
last two decades in some detail, as in the account of Brus and Laski (1989) or of
Kornai himself, it is clear that whatever has been tried was at best piecemeal;
market socialist reform in some integrated pattern, with institutional restructur-
ing to address the various incentive and agency problems, has never been tried,
and certainly not with any measure of political democracy or market competi-
tion. Quite often the crucial decisions on the entry and exit of firms and the
selection, promotion, and dismissal of managers have remained in the hands of
the all-powerful party nomenklatura. Sachs (1991) claims, on the basis of his
recent experience in Poland, that market socialism involving "liberalization
without privatization" is particularly pernicious, because it gives the managers
and workers of a public firm autonomy without responsibility, often leading to
their joint cannibalization of the firm's assets. But this only means that the key
incentive and agency problems in the management of a public firm have to be
addressed; we claim that privatization is not the only or even a better way of
handling those problems.
We take the main problem with public ownership, as traditionally prac-
ticed, to be the inefficiencies associated with the failure to separate political
from economic criteria in decision-making concerning the firm. (The "soft
budget constraint" is an instance of this failure.) Our proposals involve remov-
ing firms from the state's orbit, but not granting individual citizens unbridled
private property rights in them. Citizens will have rights to profits of firms,
and, in one variant, even to trade such rights to some extent, but they will not
be able to capitalize these profit streams. Allowing citizens or institutions to
trade these rights can provide some of the signals that a capitalist stock market
provides concerning the effectiveness of firms' managements. The final disci-
pline in monitoring managements, however, will be administered by banks
responsible for financing the firms' operations and, in some cases, also by other
firms in the same financial group.
A natural question to ask of any proposal of market socialism like ours, as
Martin Weitzman posed to one of us, is why bother with these complicated
institutional arrangements which mimic capitalism, when you can have the real
thing? East European economists sometimes put it in a slightly different form:
we are tired of experiments, let us not waste any more time, but go for the only
time-tested system that works, however imperfect it may be—capitalism.
 Pranab Bardhan and John E. Roemer 103

First, we are skeptical that the option of the "real thing," Western-style
capitalism, is available to some of the East European countries, China, or
Vietnam, however much some people in these countries may crave it. The
institutions of Western capitalism, including its legal, political, and economic
infrastructure, evolved over a long period. Some of them are not easily
replicable. In fact the bank-centric organization that we shall describe is a way
of mitigating an historical handicap in capital market institutions. It is impor-
tant to realize that it was the underdevelopment of capital markets in late
19th-century Germany that gave rise to its present system of heavy bank
involvement in the financing and management of industrial companies. Even in
the case of Japan, as Horiuchi (1989) points out, the main bank system
originated in the highly imperfect financial markets and economic uncertainties
of the immediate postwar period.
Secondly, at the risk of belaboring the obvious, the whole purpose of
designing blueprints for market socialism that can achieve production efficiency
roughly similar to that of capitalism, instead of rushing for the "real thing," is
our conviction that such a system would have a more egalitarian income
distribution, and would be more sensitive to social needs, such as education,
health care, and environmental protection. Under market socialism the social
dividend, the surplus after payment of wages, interest and taxes in large firms,
can be redistributed in the form of private consumption for workers or social
consumption and investment. We do not believe that entrepreneurial functions
require, for their elicitation, the large drain on the social surplus that corporate
capitalists usually exact, nor that inherited wealth serves a useful social
purpose.
Further, the more egalitarian distribution of profits may also lead to fewer
public "bads," with which we associate many of capitalism's ills. To illustrate
this point, imagine the economy as consisting of a firm that hires all citizens and
that produces the single good that all consume. The firm also produces, as a
joint product, a public bad. Equivalently, the public bad may be viewed as a
production input, whose level determines the production possibilities of the
firm. Suppose that the equilibrium output of the good is an increasing function
of the level of the public bad, as is the profit of the firm and the equilibrium
wage. One may think of several examples of such a bad, the most obvious of
which is pollution. Another one is the speed of the assembly line (as a proxy for
working conditions more generally): this is a public bad for workers in the firm,
but is positively correlated (at least up to a point) with output and profits. A
third example is noxious advertising by the firm (as in the "Joe Camel" ads by
R. J. Reynolds). (Here, the public bad is associated with increasing profits of the
firm, but not increased output.) Assume that citizens have identical preferences,
increasing in the good and decreasing in the public bad, but that they earn
different wages and own different shares of the firm. Assume that both a
person's wage level and his or her share of ownership of the firm are increasing
104 Journal of Economic Perspectives

functions of "skill." (Citizens with higher wages have invested more in the firm
in previous periods.)
The political economy works as follows. Citizens regulate the level of the
public bad in an election—they decide, for example, upon the level of pollution
that the firm may emit. For any given level of the public bad, the production
function of the firm is determined and so is the competitive equilibrium in the
economy. Each citizen has an indirect utility function over levels of the public
bad induced by his utility in the equilibria associated with different levels of the
public bad. The simplest theory of politics is that the level of the public bad
chosen will be that which is preferred by the median voter—where the median
is with respect to the distribution of "skill" in the population.
Roemer (1991) shows that, for a large class of preferences, the level of the
public bad optimal for a citizen is an increasing function of that citizen's share
of the firm, which is to say, of the agent's "skill." If the initial distribution of
profit shares is highly inegalitarian, as in every capitalist country today, then
the median voter owns less than the population per capita share of the firm. If
that distribution approaches an egalitarian one, as it would under our propos-
als in this paper, then the median voter's share of the firm must eventually
increase, as it approaches a population per capita share. This would suggest
that the level of the public bad decided by a democratic election would increase.
There is, however, another, more plausible, "interest group" theory of the
political process in which the opposite would occur. With an egalitarian distri-
bution of profits, no coalition of shareholders desire as high a level of the public
bad as is desired by the coalition of large shareholders in a (capitalist) economy
with an inegalitarian distribution of shares. Large shareholders are powerful:
they fund political parties and, more generally, are influential in the political
process. It would therefore not be surprising if the level of public bad which is
the outcome of the political process is higher under an inegalitarian profit
distribution than under the egalitarian alternative.
The general point is that negative externalities are associated with the high
degree of concentration of ownership of firms in capitalist economies, if the
political process is also strongly influenced by the wealthy. Equalizing the
distribution of profits can reduce these externalities.
Why is social democracy not an adequate cure for the inegalitarianism of
capitalism? Perhaps, when it works, it is—or rather, perhaps it produces as
egalitarian an outcome as our market socialism would. However, we believe
that social democracy requires rather special political circumstances that are
absent in many countries for which our market socialism proposal may be
feasible. Since it permits a powerful capitalist class to exist (90 percent of
productive assets are privately owned in Sweden), only a strong and unified
labor movement can win the redistribution through taxes that is characteristic
of social democracy. It is idealistic to believe that tax concessions of this
magnitude can be effected simply through electoral democracy without an
organized labor movement, when capitalists organize or finance influential
 Market Socialism: A Case for Rejuvenation 105

political parties. Even in the Scandinavian countries, strong apex labor organi-
zations have been difficult to sustain and social democracy is somewhat on the
decline now. We believe that some variant of the market-socialist proposal we
describe below is politically feasible in the existing or former socialist countries
and in some of the developing countries, but that once a capitalist class becomes
entrenched, social democracy may be politically difficult to achieve given the
non-existence of strong and unified labor organizations. Besides, public pro-
duction—as opposed to capitalist production with redistributive fiscal transfers
—may be necessary sometimes in meeting complex social objectives, like ensur-
ing a certain quality of social life or certain standards in education or arts and
culture.
The argument in the rest of the paper proceeds as follows. The following
section looks at the question of the "soft budget constraint" as an agency
problem under market socialism. We then propose two variants of a bank-centric
system of insider monitoring as a viable solution to the agency problem. The
next section discusses the essential problem of political accountability and the
difficulty of credible pre-commitment in avoiding the soft budget constraint
problem, and suggests ways of minimizing this problem in our proposed
system. We then conclude by addressing some of the other standard objections
to a proposal for market socialism.

The Soft Budget Constraint as an Agency Problem

The classic Lange model of market socialism, like the standard textbook
model of a capitalist market economy, concentrates on the primacy of the role
of prices in resource allocation. In a world of imperfect information, non-price
mechanisms (like contracts and reputation) often play a more important role,
and models of market socialism, like the models of market economies they
mimic, must take serious account of them. Stiglitz (1990) offers a discussion of
these issues in the context of market socialism.
Here, we must address the key question of any model of market socialism:
how to motivate the managers of public firms to maximize profits. Under
private ownership, the entrepreneur has a stake in the firm and gains or loses
money depending on the performance of the firm. The salaried manager of a
public firm usually has much less at stake, and therefore may not have the full
drive or incentive to pursue the Langean rules of the game. In particular, that
manager may operate under the built-in expectation of what Kornai calls "the
soft budget constraint." Various political considerations interfere with the harsh
exit mechanism of the market and the state remains the ultimate rescuer of
losing concerns. Political accountability prevails over financial accountability.
Kornai (1986) spells out the mechanisms of softening the budget constraint in
terms of soft subsidies, open-ended and negotiable; soft taxation—easily
arranged tax-relief; soft credit—easy renegotiation of debt, often forced upon
106 Journal of Economic Perspectives

suppliers and other creditor firms; and soft administered prices, often involving
cost-plus pricing.
The essential problem of soft budget constraints presents at least two
conceptually separable elements: one is an information or agency problem, the
other is a political problem largely involving the problem of credible pre-
commitment on the part of the state. Let us take the agency problem first. The
state, as the principal, even when it has the "political will" to demand efficiency
of management, may not have the full information to decide the extent to
which a firm's bad performance is due to factors beyond the manager's control.
This agency problem is clearly absent in owner-managed firms under private
ownership.
But if one goes beyond 19th-century owner-entrepreneurial capitalism and
looks to sectors outside the small-scale sector of trade, crafts, services, and
agriculture, large-scale enterprises under corporate capitalism also face qualita-
tively similar agency problems in management.1 With the separation between
ownership and management in such a capitalist firm, managers may not
maximize the share value of the firm and may instead feather their own nests,
or simply take wasteful or foolhardy decisions. The large body of shareholders,
the collective principal in this case, may have a difficult monitoring problem at
hand, as the individual investor has neither the ability nor the full incentive to
monitor. Just as in a socialist firm which although owned by everybody is really
owned by nobody (because nobody takes responsibility), so a capitalist firm
owned by thousands or even millions of investors may have difficulty in
ensuring responsibility. Only a small part of the agency costs under corporate
capitalism can be gauged from the astronomical salary raises that chief execu-
tive officers in American and British companies regularly give themselves—this
is clearly a case of the soft budget syndrome, with respect to the shareholders'
money rather than the taxpayers'.
Finance theorists concerned with the agency problem in corporate
capitalism—for example, Alchian and Demsetz (1972), Jensen and Meckling
(1976), and Fama (1980)—claim that the primary disciplining of managers
comes through the capital market and the managerial labor market (both
within and outside the firm). In principle, market socialism can reproduce the
managerial labor market, if a manager's reputation and future wages crucially
depend on the performance of the currently managed firm. It requires time
and considerable depoliticized institution-building, but not necessarily a capital-
ist property system, to nurture a corporate culture of competitive bidding in
the market for professional managers. But reproducing the capital market
without private ownership is much more difficult. Socialism, as usually
1
As a matter of fact, Lange (1938) insisted that "public officials must be compared with corporation
officials under capitalism and not with private small-scale entrepreneurs." With respect to small-scale
industry and farming, Lange was of the opinion that "private property of the means of production
and private enterprise may well continue to have a useful social function by being more efficient
than a socialized industry might be."
 Pranab Bardhan and John E. Roemer 107

conceived, lacks an institution like the stock market that is supposed to provide
a mechanism of continuous assessment of managerial performance. The threat
of corporate takeover (along with stock options in managerial compensation) is
supposed to keep managers honest and the firm efficient, and thus to resolve
the conflict of interest between those who bear risk and those who manage risk.
But the financial discipline of corporate takeovers is usually a delayed and
wasteful process. Jensen (1989) notes that in the United States, takeover and
leveraged buyout premiums average 50 percent above market price, which
illustrates how much value corporate managers can destroy before they face a
serious threat of disturbance. Even the takeover process contains a basic
asymmetry of information: managers are more informed about the source of a
firm's problems than outside buyers. As Stiglitz (1985) suggests, takeovers are
like buying "used firms," and are subject, as such, to Akerlofs "lemons
principle."
We also should not forget that corporate raids, a peculiarly Anglo-Ameri-
can game, have not been necessary for strong firm performance in some
countries of continental Europe (like France or Germany), nor in Japan. The
predominant practice in postwar Japan (at least until the middle 1970s) of
mutual stock-holding of private companies within the keiretsu, a corporate
financial grouping often with a "main bank" as the nucleus, provides an
alternative model of monitoring by involved parties. We have drawn upon
some of the features of the Japanese system in our proposed alternative
financial system of monitoring under market socialism, presented in the next
section.2 Even in the United States in recent years, as Jensen (1989) points out,
new organizational forms, such as the leveraged buyout association, are evolv-
ing, in which the key organizational principle is the active involvement by
investors, who hold large equity or debt positions, in the long-term strategic
direction of the firm. In other words, in the trade-off between risk diversifica-
tion (facilitated by the diffuse stock ownership system) and control (which is
diluted by that system), the balance is shifting in favor of more control by large
investors.

A Proposed System of Insider Monitoring

We propose two designs for corporate control in a market socialist econ-

omy, which are intended to solve the managerial incentive problem, while
maintaining a roughly egalitarian distribution of total profits in the economy.
2
As M. Aoki has pointed out to us, in Japan there are two types of keiretsu, overlapping but
conceptually distinguishable. One is a financial corporate grouping across industries, bound by
mutual stock-holding and a main bank as the nucleus; the other is a hierarchical grouping of firms
connected by inter-industrial input-output relations, with a major manufacturing firm at its apex.
Although our proposed system emphasizes the former, we have also borrowed one or two
institutional features from the latter system.
108 Journal of Economic Perspectives

Both designs rely on banks as the main monitors of firms. In the first scheme,
as elaborated in Bardhan (1991), the state would not own a public firm directly.
The firm will be a joint stock company with some of its shares owned by its
workers, but also a major part of its shares owned by other public firms
(including their workers) in the same financial group, together with the main
investment bank and its subsidiaries.3 The share-owning workers in one firm
will have the motivation and some leverage in prodding other firms in the
group to maximize profits. Some shares will be owned by companies outside
the group: other financial institutions, pension funds, local governments, and
so on. The firm will also borrow from the main bank (where the state owns a
majority interest) which may sometimes organize loan consortia for the firm. As
Horiuchi (1989) suggests for the Japanese system, the primary role of the main
bank may be that of what Diamond (1984) has called a "delegated monitor;"
through its commitment to the affiliate firm, the main bank communicates to
other investors and lenders about the firm's credibility. All this, however,
applies to large firms.
Our scheme does not preclude small entrepreneurs running private firms,
getting rich "gloriously" (to use Deng's famous maxim in China). The issue of
socialist takeover (with compensation) arises only after the owner-entrepreneur
firm goes public and the shareholder-manager agency problem becomes seri-
ous. Our proposed scheme does not interfere with the process of innovations
that sometimes take place in small entrepreneurial firms. Most small firms,
under capitalism, either die after some time or are bought by large firms, and
the same trajectory could be expected under market socialism, where large
corporations would purchase successful private firms.
The shares of a large firm can be sold to the main bank. At the first signs of
significant attempts by other firms at unloading the shares of a particular firm,
and usually much earlier, the main bank will take measures to prod and
discipline the management, renegotiate the debt contract if necessary, orches-
trate financial rescue strategies, help the firm with an interest moratorium and
emergency loans, and arrange for technological assistance from affiliated firms
and for the (temporary) sale of the firm's stock assets (in other firms) to cover
its operating losses. With the bank's substantial share holdings, it will even have
the power to take over the management of the ailing firm temporarily, if
necessary. In cases where bankruptcy cannot be prevented, the assets of the
firm will be disposed of by the bank among a number of other enterprises. Aoki
(1988) gives the example in Japan of Sumitomo Bank in the mid-1970s taking
over the management of the distressed Toyo Kogyo Company, the maker of
Mazda cars, until it was salvaged and nursed back to health. The main bank is
motivated to arrange the rescue operation, a disproportionate share of which
cost is borne by the bank, by its desire to retain its reputation of credibility as a

3
When lenders are also important equity holders, credit-rationing and other onerous terms of
lending may be largely avoided, and more risk-taking encouraged.
 Market Socialism: A Case for Rejuvenation 109

delegated monitor, in a system of reciprocal delegated monitoring with a small

number of other main banks. It also does not want to lose the intangible asset it
has accumulated specific to its relationship with the affiliate firm.4 As Berglof
(1989) has found in his comparative study of alternative financial systems,
creditor reorganization of ailing firms is relatively common in bank-oriented
financial systems. Such reorganization is more informal and less costly than
involvement with outsiders like courts or corporate raiders. In the United
States, venture capital often plays a similar role, getting involved in active
management of a company in times of trouble.
The maximum size of a corporate group should not be large, and would
depend on the monitoring ability and technical and financial expertise of the
main bank. On the other hand, it should not be too small, at least for purposes
of risk diversification. It will be desirable for members of a corporate group to
be technologically somewhat interrelated, either at the vertical upstream-down-
stream level or at the horizontal contracting level, for three reasons:
(a) technological interrelatedness makes it easier for firms to be somewhat
knowledgeable about each other's production and market conditions, so that
sharing information, closer monitoring and early detection of trouble become
feasible; (b) there may be spillovers in the results of research and development,
in which case the externalities in the generation and diffusion of technology can
be internalized within the mutual stock-holding corporate group; and (c) the
main bank can specialize in some relatively narrow and well-defined technologi-
cal area, for the purpose of monitoring and scrutinizing its loans and equity
involvements in the associated companies. On the other hand, if the technologi-
cally interrelated firms are prone to suffer covariate risks, the main bank needs
to have a sufficiently diversified portfolio of loans and equities in firms outside
the corporate group to reduce the danger of bank failure.
The proposed bank-centric financial system thus largely mitigates the
planner-manager principal-agent problem, and does so in a way potentially
superior to that of the stock market-centric system. The main bank and the
group partners have a larger stake in, and more "inside" information about, a
company than the ordinary shareholders in a stock market-centric system, are
likely to be capable of detecting and acting on early signs of trouble more easily
than a diffuse body of stockholders, and are prone to take a longer view in the
matter of risk-taking and innovations. Under the stock market system, even
fully rational investors, in a situation of highly imperfect information about the
activities of the firm, may be too much concerned with short-run profitability.
In a model with asymmetric information, von Thadden (1991) has shown the
likelihood of myopic firm behavior caused by the fear of withdrawal of outside
finance in case of failure in an early stage of a project, even when firms and

4
Inthe Japanese case, long-term workers also have an incentive to work harder to avoid liquidation
of the firm, which involves a significant loss of firm-specific benefits and seniority.
110 Journal of Economic Perspectives

investors can conclude binding long-term contracts. A firm's planning horizon

can be lengthened if the lending bank, as a delegated monitor, can provide
informational insurance to the firm, thus mitigating its incentives to produce
quick profits, and the cost of this insurance is lowered by the scale economies of
monitoring by the bank. Empirically, Berglof (1989) confirms that a feature
that distinguishes the bank-oriented systems from their stock market-oriented
counterparts is the longer-term shareholdings in the former.
This proposal is designed to work particularly in economies where a fully
developed stock market and its concomitant financial institutions are absent or
do not provide the major role in financial discipline (as in Japan up to the
middle '70s). In economies where functioning stock markets have played a
more central role, we suggest a second possible alternative. The first step is to
distribute vouchers to all adults, giving each a per capita claim on the profits of
each large, or formerly state-owned firm in the country. More realistically,
mutual funds might be set up, each of which would initially hold the same
portfolio of all large firms in the country, and the vouchers distributed to adult
citizens would entitle each to a per capita share of the income of each mutual
fund. If a stock market were then opened, prices would eventually equilibrate
for shares of these mutual funds and/or firms. Many citizens would be likely to
sell their shares to the rich and leave the market. This might well be myopic
behavior; moreover it could lead to the negative externalities associated with a
high concentration of ownership.
Therefore, we propose that the stock market be limited as follows: while
citizens would be free to trade their stock in mutual funds for stock in other
mutual funds, they could not liquidate their portfolios, that is, turn them into
cash. We might call this a clamshell economy: all citizens initially would receive
an equal endowment of vouchers denominated in "clamshells," which could be
used to purchase shares of mutual funds. Shares could only be purchased with
clamshells, not money, and money could not purchase clamshells. Thus, our
proposal differs from the voucher schemes discussed recently in Poland and
Czechoslovakia.
The first virtue of the clamshell stock market is that it should provide the
same signals that a capitalist stock market does, apart from providing some
risk-bearing by citizens. If a firm's management is performing poorly, the
mutual funds holding its stock will sell and its stock price (in clamshells) will
fall. This would be a signal to the banks that the firm was in trouble, and the
firm would be hard-pressed to finance its operations. That financing, in this
model, would come from bank loans: the clamshell stock market would not be a
source of equity, but only a market for property rights in the profits of firms.
The board of directors of a firm would be comprised of representatives of those
mutual funds which own it, and of the main bank which largely arranges for
financing it.
A second virtue of the clamshell economy is that it would prevent the
concentration of ownership of firms in the hands of a small class. The mecha-
nism of requiring citizens to hold shares of mutual funds, and not firms
 Pranab Bardhan and John E. Roemer 111

directly, would help prevent the unworldly from losing their stock assets by
making poor investments. There would be some state regulation of the mutual
funds.
We can think of a number of criticisms of the clamshell proposal. The main
one is that by not allowing citizens to capitalize their rights to profit streams
from firms, we prevent them from shifting consumption from the future to the
present, and hence create Pareto inefficiency. Our response is that, apart from
private behavior being sometimes too myopic from the social point of view, a
negative externality is associated with the concentration of shares that might
follow if citizens could liquidate their shares, the creation of a small class of
large shareholders with concomitant political power, as discussed earlier. (For
that matter, not allowing citizens to sell their votes in elections, which most
democracies insist upon, also creates inefficiency by prohibiting the develop-
ment of a market.)
It can also be objected that financial markets will find ways of de facto
allowing people to capitalize their shares. For example, some firms might pay
out their capital stock as dividends. This problem could be addressed by setting
a maximum fraction of profits that could be paid out to shareholders. As a way
of mitigating some of these problems, we suggest that citizens be allowed to use
their clamshell assets as collateral with banks for (large) loans. For the period of
the loan, the bank would manage the borrower's portfolio, deducting the loan
payments from the income stream. Another problem would exist if foreign
investment were permitted. Foreigners would not have clamshells; their invest-
ments in domestic firms must necessarily take the form of real assets. Thus,
citizens could invest money in domestic firms through foreign agents. Again,
this possibility would require some regulation of foreign investment.

The Soft Budget Constraint as a Political Problem

The major question one must raise, when depending on the main bank as
the primary monitor of the public firms in a corporate group, is who monitors
the monitor? If the main bank depends substantially on the state for finance,
the political aspect of the soft budget constraint again looms large, exposing the
soft underbelly of socialist economies.
Whenever the benefits of state polices of leniency in underwriting losses, of
refinancing, or of providing relief or subsidies, are concentrated and highly
visible, while the costs of such policies are diffuse, inevitable political pressure
on the state builds to follow such policies, whether in a capitalist or a socialist
country. But such pressure is less resistible in a socialist country. In capitalist
countries, while large bail-outs by the state are not uncommon, the prevailing
hegemonic ideology makes lay-offs and bankruptcies politically more tolerable.
All systems make costly mistakes from time to time; under socialist monitoring
(including under our proposed system) what are called Type 2 errors (bad
projects are allowed to continue too long) are likely to be more common than
112 Journal of Economic Perspectives

Type 1 errors (projects are abandoned too soon) that seem to characterize the
harsh, if occasionally myopic, exit mechanisms of capitalist market economies.
Different societies have different degrees of tolerance for these two types of
error. Societies that value stability and security more than mobility and change
seem to have a larger degree of tolerance for Type 2 errors.
While it is difficult to escape completely the politics of the soft budget
constraint, the problem may be less virulent under our proposed insider
monitoring system with proper safeguards. We spell out several reasons:
(a) In our system, between the public firm, which is an independent joint
stock company, and the state treasury, there is a hard layer formed by
equity-holding, technologically interdependent affiliate firms (or mutual funds)
and a main bank which orchestrates reciprocal monitoring. This layer provides
some financial discipline on public firms and acts as a buffer against direct
political accountability. This, of course, is not enough to prevent the whole
affiliate group from acting as a lobby with government for a troubled member.
(b) The reputational concerns of the main bank's managers may act as an
antidote to susceptibility to political pressures. In Japan, even though the banks
have been closely regulated by the Ministry of Finance, bank managers exhibit
some keenness to preserve their reputations as good monitors, and banks
compete in seeking the position of main bank for well-run firms. In our
proposed system, it would not be difficult to keep track of the reputational
record of bank managers, since the number of main banks would be relatively
small. The managerial labor market may not "forget" if a bank manager
forgives bad loans or non-performing firms too often.
(c) It is important to introduce incentive features in the payment struc-
tures of main bank (as well as mutual fund) managers linked to their perfor-
mance in monitoring firms. While the social loss from a bad project may be
many times the personal loss to the bank manager's linked income, that
personal loss may be a significant enough fraction of that manager's income to
make negligence rather costly.
(d) It is important to keep open the doors of international competition, as
a check on the institutional monitors' laxity. The international market signals
can also provide valuable guidelines and comparative reference points in the
main banks' monitoring process and raise cost and quality consciousness all
around. There are genuine cases for infant-industry protection, but to prevent
the overly common degeneration of infant industries into inefficient geriatric
protection lobbies, a clearly specified fixed duration should be announced for
such protection, after which the firm must sink or swim in international
competition. To make such pre-commitments credible, some binding interna-
tional trade agreements may be tried.
(e) It is often claimed that under the soft budget constraint, the state
remains as the risk-absorber of last resort, and so little incentive remains on the
part of managers to avoid overly risky projects. Yet in actual cases of public
sector management, one often finds too few, rather than too many, risks taken
by the managers. This is due largely to too much political accountability: the
 Market Socialism: A Case for Rejuvenation 113

managers are constantly wary of taking bold decisions that might be seen by
their political bosses as rocking the boat of the pre-existing patronage distribu-
tion system. Even in our proposed system, it may be difficult for the state to
make a credible pre-commitment not to intervene too often in the main bank's
decision-making process. Some difficult-to-change constitutional guarantees of
the infrequency of state intervention in the short- to medium-run operations of
the banks may be necessary.
(f) There should also be, as Sah and Weitzman (1991) have suggested,
well-publicized pre-commitments before public sector projects are launched that
they will be liquidated should their cumulative performance at pre-specified
dates in the future not exceed certain threshold levels. The rescue strategies by
the main bank of a corporate group (which we have indicated above) will
be subject, by prior legislative enactment, to this kind of liquidation pre-
commitment. Of course, the major constituency opposed to liquidation or
scaling down of unprofitable enterprises is the workforce. Sah and Weitzman
have pointed out the advantages of profit-sharing payment schemes in this
context. If pre-commitment to profit-sharing is part of a public sector project
from the beginning, and if workers must sign on to this provision when they
take a job, then in chronically unprofitable concerns the attraction of clinging
to the job is obviously much less and to that extent the resistance of the
workforce may be weaker.
(g) Although in our system the state is to own directly a majority of the
shares of a main bank, some significant fraction of the shares will be owned by
pension funds, insurance companies, and other banks, to allow for some
diversification of interest in and professional control of the main bank's lending
operations (even though they may not be opposed to bailing out by the state).

Other Problems with the Proposed System

An additional problem with our proposed bank-centric corporate groups is

the possibility of collusion and industrial concentration facilitated by interlock-
ing shareholding and the exchange of inside information. It is therefore
important to preserve the discipline of product market competition, along with
some antitrust regulation, in this system. Major competitors must be placed in
separate corporate groups around different main banks. We are not ruling out
cases of a firm leaving one corporate group and joining another (although in
the Japanese case the relationship between a main bank and its customers is
usually quite stable), but applications of new entrants to a group should be
subject to strict scrutiny by an independent antitrust authority for collusion
possibilities.
In some situations, particularly when the market size is small, economies of
scale may make it inefficient to have many competing firms in the same
industry. In these situations, a corporate group with mutual stock-holding
among companies linked in input–output interdependence might be helpful in
114 Journal of Economic Perspectives

providing some mutual accountability. For example, a steel firm having a stake
in a coal company belonging to the same group may, through its own levers of
control and those of the main bank, pull up the coal company if it indulges in
monopoly-induced sloth and high cost operations.
Of course, partial vertical integration through mutual stock-holding may
increase market power and make entry difficult. Here, international competi-
tion can provide a crucial safeguard. There are lessons to learn from the cases
of South Korea and Taiwan, where the state has often energetically used the
carrot of easy loans and other benefits and the stick of international competi-
tion to prod the firms (many of them in the public sector) on to the technologi-
cal frontier. The East Asian cases where the dominant state has worked closely
with the market and has used world market signals to keep domestic firms on
their toes provide important counterexamples to Kornai's (1990) conviction
that "the systemic tendency of self-reproduction of the bureaucracy" will snuff
out reform attempts at market coordination, a conviction that feeds his pes-
simism about "third ways" like market socialism.
Even though our proposed system is likely to be more egalitarian than
capitalism, there will be several departures from egalitarian distribution to the
extent there are various incentive payment schemes for managers and workers.
We believe this is a price worth paying for efficiency. What about the bureau-
crats' cut from the social dividend under market socialism? In our scheme,
markets allocate most resources and the state is not involved in production.
Top-level bureaucrats will not be any more powerful than they are in mixed
economies like those of France or the Nordic countries. There will be an
elaborate monitoring machinery in the public banks which is absent in the stock
market-centric capitalist financial system. But as we have argued earlier, the
agency costs of stock market capitalism and the wastefulness of the corporate
takeover process can be viewed as the opportunity costs of not adopting a
Japanese-style banking bureaucracy.
There are alternative models of market socialism, oriented more than ours
to the labor market and the important issues of worker participation and
motivation, as opposed to the emphasis on financial systems and managerial
motivation in our model of market socialism (although in our proposed system
as well, shareowning workers play an important role in the whole incentive
scheme). Those alternative models are aimed at some other goals of socialism
which we have not discussed, like shop-floor democracy, less alienating work
organization, worker solidarity and autonomy, reduction in the inequality of
wage income, less of unemployment as a worker disciplining device and so on.
There is a large and significant literature on market socialism in the form of
worker-owned or labor-managed firms. We note in passing that our proposed
insider-monitoring system also provides a solution for some of the adverse
incentive and agency problems that many critics of labor-managed firms have
pointed out.
Jensen and Meckling (1979), for example, have identified the "horizon
problem"—workers will not value cash flow beyond their term of employment,
 Pranab Bardhan and John E. Roemer 115

leading to suboptimal choices of investment and maintenance of capital—and

the "common property problem"—only projects maximizing profits of the firm
per worker will be chosen, leading to suboptimal employment and rejection of
many worthwhile projects. These problems can be solved in the labor-managed
firm if one introduces, as Barzelay and Thomas (1986) have suggested, the
floating of non-voting shares for non-employees to raise outside equity. If, in
the spirit of our proposed financial system, we have the main bank provide
insider monitoring, one can see a way out for a problem which Barzelay and
Thomas have not quite solved for the labor-managed firm: the main bank will
monitor and discipline the firm against the possible tendency toward excessive
wage payments or myopic capital consumption, which the non-voting shares of
outsiders will not quite stop otherwise.
This paper is about blueprints and not so much about their implementa-
tion. We do not harbor illusions about the formidable problems, political and
economic, of the possible transition to our proposed system of market socialism.
That transition would surely require the development of new institutions, but
possibly not many more than, or organizationally more difficult than, those
required for the transition to capitalism. In the immediate future both types of
transition will involve a set of common and difficult problems: for example,
breaking state monopolies, ending large-scale public subsidies, introducing
markets and competition along with their inevitable painful readjustments and
dislocations, organizing joint stock companies and a viable commercial banking
system, overhauling the legal system and so on. However, the bank-centric
monitoring system may be less difficult to introduce in some developing
countries with a pre-existing set of public investment banks and financial
institutions.
We have claimed that introducing competition, markets, and political
democracy are the crucial parts of the reform program, not large-scale privati-
zation. Some of the popular horror stories about inefficient public firms, or of
parastatals in developing or socialist countries which have become white ele-
phants draining the public treasury, may have more to do with their being
monopolies than with their being publicly owned. There are many examples of
efficient public firms in competitive environments around the world. Empirical
evidence of significant efficiency differentials between public and private firms
after adjusting for market structure (and regulatory policy) is quite scanty.5 We
have argued that, with appropriate institutional restructuring, incentive and
efficiency issues can be handled without privatization. One hopes that this point
of view will not be summarily dismissed simply because, in the current populist
discourse in some of the East European countries, the word "socialism" brings
bad memories of something imposed on them in its name, or because, in the
simple-minded ideology of the free marketeers in those countries and their

5
As Vickers and Yarrow (1991) note in their survey in this journal of the evidence on ownership
and efficiency, in competitive industries even in cases where private ownership seems to have the
edge, competition rather than ownership per se is the key to efficiency.
116 Journal of Economic Perspectives
Western patrons, the market mechanism can function only with full-scale
capitalist property rights.
• We are grateful to Joseph Stiglitz, Carl Shapiro and Timothy Taylor for their very
helpful editorial comments. This work is related to a project on East-South System
Transformations, supported by a grant from John D. and Catherine T. Mac Arthur
Foundation.
References
Alchian, A. and H. Demsetz, "Production,
Information Costs, and Economic Organiza-
tion," American Economic Review, December
1972, 62:5, 777–95.
Aoki, M., Information, Incentives, and Bar-
gaining in the Japanese Economy. New York:
Cambridge University Press, 1988.
Bardhan, P., "Risktaking, Capital Markets,
and Market Socialism," University of Califor-
nia at Berkeley, Department of Economics,
Working Paper No. 91-154, January 1991.
Barzelay, M. and L. R. Thomas, III, "Is
Capitalism Necessary? A Critique of the Neo-
classical Economics of Organization," Journal
of Economic Behavior and Organization, Septem-
ber 1986, 7:3, 225–33.
Berglof, E., "Capital Structure as a Mecha-
nism of Control: A Comparison of Financial
Systems." In Aoki, M., B. Gustafsson and
O. E. Williamson, eds., The Firm as a Nexus of
Treaties. London: Sage Publications, 1989,
237–62.
Brus, W. and K. Laski, From Marx to Market:
Socialism in Search of an Economic System. Ox-
ford: Clarendon Press, 1989.
Diamond, D., "Financial Intermediation and
Delegated Monitoring," Review of Economic
Studies, July 1984, 51:3, 393–414.
Fama, E., "Agency Problems and the The-
ory of the Firm," Journal of Political Economy,
April 1980, 88:2, 288–307.
Horiuchi, A., "Informational Properties of
the Japanese Financial System," Japan and the
World Economy, July 1989 1:3, 255–78.
Jensen, M., "Eclipse of the Public Corpora-
tion," Harvard Business Review, September–
October 1989, 89:5, 61–74.
Jensen, M. and W. Meckling, "Theory of
the Firm: Managerial Behavior, Agency Costs
and Ownership Structure," Journal of Finan-
cial Economics, October 1976, 3:4, 305–60.
Jensen, M. and W. Meckling, "Rights and
Production Functions: An Application to La-
bor-Managed Firms and Codetermination,"
Journal of Business, October 1979, 52:4,
469–506.
Kornai, J., "The Soft Budget Constraint,"
Kyklos, 1986, 39:1, 3–30.
Kornai, J., The Road to a Free Economy. New
York: W. W. Norton, 1990.
Lange, O., On the Economic Theory of Social-
ism. Minneapolis: University of Minnesota
Press, 1938.
Roemer, J., "Would Economic Democracy
Decrease the Amount of Public Bads," Univer-
sity of California at Davis, Department of Eco-
nomics, Working Paper, 1991.
Sachs, J., "Comparing Economic Reform in
Latin America and Eastern Europe," Hicks
Lecture, Oxford University, March 1991.
Sah, R. and M. Weitzman, "A Proposal for
Using Incentive Pre-commitments in Public
Enterprise Funding," World Development, June
1991, 19:6, 595–605.
Stiglitz, J., "Credit Markets and the Control
of Capital," Journal of Money, Credit, and Bank-
ing, May 1985, 17:2, 133–52.
Stiglitz, J., "Whither Socialism? Perspec-
tives from the Economics of Information,"
Wicksell Lectures, Stockholm, May 1990.
Vickers, J. and G. Yarrow, "Economic Per-
spectives on Privatization," Journal of Economic
Perspectives, Spring 1991, 5:2, 111–32.
von Thadden, E. L., "Bank Finance and
Long-Term Investment," WWZ Discussion Pa-
per, University of Basel, July 1991.
This article has been cited by:

1. Andrei Shleifer,, Robert W. Vishny. 1994. The Politics of Market Socialism. Journal of Economic
Perspectives 8:2, 165-176. [Abstract] [View PDF article] [PDF with links]
2. Pranab Bardhan,, John E. Roemer. 1994. On the Workability of Market Socialism. Journal of Economic
Perspectives 8:2, 177-181. [Abstract] [View PDF article] [PDF with links]
3. Samuel Bowles,, Herbert Gintis. 1993. The Revenge of Homo Economicus: Contested Exchange and
the Revival of Political Economy. Journal of Economic Perspectives 7:1, 83-102. [Abstract] [View PDF
article] [PDF with links]

View publication stats