I. Introduction

On May 8, 1993, the day when the author of this paper sat down to start drafting its first version, The Washington Post carried a front page story about a division of the U.S. National Institute of Health in Bethesda “in which some female employees were promised promotions and raises if they agreed to have sex with their male managers….” (Jennings, 1993, p. 1). On the same day, an article in The New York Times reported that in China, in the process of privatization of land, “individuals seem to be expropriating the state’s property at bargain prices.” (Wu, 1993, p. 3). The article went on to state that “…today’s real estate boom raises troubling questions about land that is allocated not [so much] by prices but by guanxi.” (ibid). ^1/ Guanxi is the Chinese word for “connections.” The article cited a Chinese businessman to the effect that: “if you have guanxi. it is the time to make big money.” Thus, according to this article, in today’s China, “connections” have great economic value and smart people can capitalize on them to earn large “incomes.” On May 8, 1993 French newspapers were still carrying stories about the suicide of former Prime Minister Pierre Bérégovoy, attributed in part to his despondency as a result of the accusation that he had received an interest-free loan from a friend. And Italian newspapers were providing further details about private entrepreneurs who had paid large bribes to well-placed politicians in order to get lucrative contracts with the Italian public sector. On August 10, 1994, when the final version was completed, front page news in the New York Times and the Washington Post reported on the accusation that gifts given to a high U.S. government official might have resulted in favorable decisions vis-à-vis certain business activities.

What is the unifying thread in all of these stories? Essentially, it is the accusation that the “arm’s length principle” was not applied to economic decisions. This principle requires that personal relationships should play no role in economic decisions involving more than one party. Equality of treatment for all economic agents is essential for a well-working market economy. In the examples cited above, economic decisions--in land markets, in labor markets, in financial markets, in the market for government contracts, and in government decisions concerning private sector’s activities--were allegedly not guided by the arm’s length principle.

II. Government and Markets

Economic theory tells us that if markets were perfect, there would be no need for the government to play an economic role. Furthermore, almost by definition, these markets would be guided by the extreme version of arm’s length relationships among market participants implied by pure or perfect competition. Decisions concerning the selling and buying of goods; the hiring, retaining, and compensating of workers; the lending and borrowing of money; the timing, size, and location of investment projects, and so forth would be determined by economic considerations and profit opportunities alone. Personal relationships among the parties would play no role.

However, markets are not perfect; the existence of public goods, of externalities in production and consumption, of informational deficiencies, and of monopolistic practices, justifies and requires that the state play an economic, corrective role. ^1/ The work of Lindahl, Vicksell, Musgrave, Samuelson and others has been influential in determining the optimal theoretical. or normative, role of the state.

A fundamental but unstated assumption in the theoretical work on the role of the public sector is that public sector officials (both policy makers and civil servants) know what they are doing and are neutral and impersonal in their pursuit of the social welfare. ^2/ When these officials make mistakes, they are honest mistakes. This assumption, though of fundamental importance, is hidden in the pure theory of public expenditure. ^3/

The theoretical work of how the public sector should pursue its corrective role owes a great deal to Max Weber’s ideal or normative type of rational-legal bureaucracy. As Robin Theobald has put it: Weber’s bureaucrats would operate “…according to rational procedures and universalistic principles in which there is no place for personalism, cronyism, and, most of all, the confusion of public with private interests.” ^4/ This bureaucracy “… is run by hierarchically ordered corps of officials who are recruited and promoted according to objective criteria such as educational qualifications and professional experience; who are paid a regular salary which is graded according to rank and qualifications; and who are allocated fixed jurisdictional areas governed by clearly laid down rules and procedures” (see Theobald, p. 56). Weber was aware that “Bureaucracy, thus understood, is fully developed… only in the modern state, and in the private economy only in the most advanced institutions of capitalism.” (see Weber, 1978, p. 56). In other words he was fully aware that he was presenting an ideal type of bureaucracy rather than describing reality. ^1/

The extent to which the real world approximates the Lindahl-Samuelson-Weber ideal is, of course, an open question. The Italian Scienza delle Finanze. which had a lot of influence on James Buchanan’s work and, through him, on the development of public choice, would have been very skeptical on various grounds about the realism, and, indeed, about the usefulness of the normative theory of public expenditure.

In the real world, the economic role of the state is exercised through the use of various instruments such as (a) public spending for government consumption and transfers; (b) taxation and borrowing; (c) various forms of regulation; (d) lending activities; and, (e) occasionally, less orthodox governmental actions such as expropriation, conscription, nationalization, privatization, exhortation, and so forth. Depending on the level of development of a country and on its sophistication, some of these instruments are more used than others. For example, poorer countries are less able to raise large shares of their national incomes in taxes; as a consequence, they tend to abuse the instruments listed under (c), (d), and (e). ^2/

The behavior, of real world bureaucracies (and policymakers) may depart significantly from the Weber ideal for several reasons. Some of these reasons have provided the basis for the public choice literature. ^3/ In this paper the focus is on the economic role of “corruption.” An important conclusion will be that the more real-life bureaucracies diverge from the Weber’s ideal, the less control will the government have-over its policy instruments and the less correction it will be able to bring, through its actions, to the imperfections of the market. In other words the less legitimate and justified will be the corrective role of the government.

The divergence between the optimum of the Samuelson solution and the real outcome of the public sector’s action may be due either to (a) policymakers pursuing policies that are not consistent with the achievement of the social welfare function; or (b) to bureaucrats distorting, in various ways, the signals that they get from the policymakers. The issue of corruption obviously concerns both. There is, thus, administrative and political corruption (see Rose-Ackermann, 1978). This paper deals mostly with administrative rather than political corruption. In other words, it focuses on the activities of individuals who, in their positions as public officials, control various activities or decisions. It does not focus on the corruption associated with particular political systems. The issues discussed transcend the political configuration of the country; they are not limited to whether a country has democratic or authoritarian institutions. Political corruption, however, is likely to be more influenced by a country’s political system.

III. Arm’s Length Relationships and Cultural Factors

On January 25, 1992 The Economist reviewed a paper by Frakash Reddy, an Indian social anthropologist who, reversing the common pattern of Western scholars going to study developing countries’ social behavior, obtained a research grant to study a village in Denmark. He spent a few months in this village--Hvilsager--and registered his impressions of the relations among its inhabitants. ^1/ These impressions formed the basis of Reddy’s paper reviewed by The Economist.

Professor Reddy had been amazed to observe that the villagers hardly knew one another. They rarely exchanged visits and had few other social contacts. They had little information on what other villagers, including their neighbors, were doing and apparently little interest in finding out. Even the relationships between parents and children were not very close. When the children reached adulthood they moved out and, after that, they visited their parents only occasionally.

Professor Reddy contrasted this behavior with that prevailing in a typical Indian village of comparable size. In the latter, daily house visits would be common. Everyone would be interested in, and getting involved in, the business of the others. Family contacts would be very frequent and the members of the extended families would support one another in many ways. Relations with neighbors would also be close.

This story has implications for the concept of arm’s length and, in turn, for the role of corruption. Arm’s length relationships in economic exchanges would be much more likely to prevail in the Danish village than in the Indian village. In the latter, the concept of arm’s length would seem strange and alien. It would even seem immoral. The idea that, economically speaking, one should treat relatives and friends in the same way as strangers would appear bizarre. Relatives and friends would simply expect preferential treatment whether they were dealing with individuals in the private or in the public sector.

If a government were established in each of the two villages, with a bureaucracy charged with carrying out its functions through the instruments described earlier, it would be far easier for the Danish bureaucrats to approximate in their behavior the Weber’s ideal than for the Indian bureaucrats. In the Indian village the attempt to create an impersonal bureaucracy that would operate “…according to…principles in which there is no place for personalism, cronyism, etc.” would conflict with the accepted social norm that family and friends come first. In this society the government employee, just like any other individual, would be expected to help relatives and friends with special treatment or favors even if, occasionally, this behavior might require bending, or even breaking, administrative rules and departing from “universalistic principles.” The person who refused to provide this help would be seen as breaking the prevailing moral code and would be ostracized.

Once civil servants begin to make distinction among the people they deal with according to the degree of family relationship or friendship, they have abandoned the arm’s length principle. It would consequently be a small step to begin to expect some compensation from more distant members of the community for performing for them tasks that should be the duty of the civil servants to perform or for treating them in the same favorable way as others. Without such compensation those who required particular permits or legal documents or other services might have to wait a long time to get them. ^1/ Thus, “speed or oil money” may be required to accelerate the process and payments of various kinds (“bribes”) to get a positive response in cases where the bureaucrat has the power of refusal.

The Indian and Danish villages described above represent polar or extreme cases of how individuals may interact within a community. Whether Professor Reddy’s description of them is or is not accurate, they provide convenient polar cases. Most societies probably fall somewhere between these two extremes, with Nord European and Anglo-Saxon countries probably closer to the Danish village model and many other countries closer to the Indian village model. The Anglo-Saxon concept of “privacy” is probably just a manifestation of arm’s length relationships. ^2/ Many developing countries would probably be closer to the model represented by the Indian village. Some industrial countries, such as Italy and possibly Japan, would also come closer to the latter than the former. Sadly, the very features that make a country a less cold and indifferent place are the same that increase the difficulty of enforcing arm’s length rules so essential for modern efficient markets and governments.

In this connection, the concept of “social capital” introduced by Coleman (1990) is of relevance. ^1/ Coleman criticized the economic theory of perfect competition in a market since it accepts “[the] fiction—that society consists of a set of independent individuals, each of whom acts to achieve goals that are independently arrived at, and that the functioning of the social system consists of the combination of these actions of independent individuals” (p. 300). He maintains that “… individuals do not act independently, goals are not independently arrived at, and interests are not wholly selfish” (p. 301). And that “personal relations and networks of relations” are important to achieve personal goals. He considers these relations and networks as a capital asset for the individual, a kind of social capital. This social capital is not tangible and is not completely fungible but is “… productive, making possible the achievement of certain ends that would not be attainable by the individual in its absence (p. 302).” The concept of social capital can be adapted to our discussion to reach conclusions that are not necessarily those intended by Coleman.

In our interpretation, social capital is an asset to the individual who possess it. It is in essence the summation of all the “I owe yous” that the individual has accumulated vis-à-vis others. Some of these may come from his family background, some from connections developed in school or at work, some from past favors made, and so forth. But, of course, the social capital of an individual does not represent a one-way street in obligations. While the individual can use this capital to ask others to do things for him, others can draw from their social capital to ask him to perform tasks or do things for them. One could, thus, distinguish a gross from a net concept of social capital. The obligations that others have toward the individual might be largely balanced by the obligations that he has toward them. The existence of social capital links individuals in a network of obligations that both increases their opportunities and reduces their individual freedom. It puts strong pressures on individuals to accommodate the needs of friends and relatives and creates a presumption that they will in turn accommodate the individuals’ needs.

In some societies (those closer to the polar case of the Indian village) the gross social capital of individuals, as we have defined it, will be particularly large. The examples of the guanxi in China and the relations in the Indian village all reflect the existence of social capital. The net social capital of individuals is likely to be unevenly distributed. This net social capital will play a large role in determining the distribution of income within countries. The existence of social capital is likely to interfere with arm’s length relationships and, in particular circumstances, to lead to corruption.

IV. On Corruption

Not long ago the word corruption was infrequently used in professional works although the problem reflected by its existence is obviously a very old one. ^1/ That word did not appear often in newspapers and it was rarely mentioned by economists even though sociologists, political scientists, and a few economists did pay some attention to it. Very few economists spent much time assessing its economic significance. ^2/ Recent newspaper articles indicate that this phenomenon is far more widespread and universal than previously thought. Evidence of it is everywhere, in developing countries and, with growing frequency, in industrial countries. A perusal of recent daily newspapers would show that this problem has attracted a lot of attention in Brazil, France, Germany, Greece, India, Italy, Japan, Korea, Russia, Spain and many other places.^3/ Prominent political figures, including presidents of countries and ministers, have been accused of corruption and, as a consequence, some have resigned or have been forced out of office.

Corruption comes in many shapes and forms. It is very difficult to define and at times it is even more difficult to identify. Several definitions have appeared in the literature but none seems to be fully satisfactory. ^4/ Here we shall simply define it as the intentional noncompliance with arm’s length relationship aimed at deriving some advantage, for oneself or for related individuals, from this behavior. In the Danish village described by Professor Reddy, if corruption existed, it would reflect the isolated acts of particular individuals who would try to take advantage of their positions for personal gain. In other words it would reflect a clearly, unwarranted and antisocial behavior not induced by social pressures or prevailing social norms. In this case its identification might be easier since the victims of this corruption would be more likely to report the perpetrators. Its punishment would also be easier since it would conform with generally accepted social norms. However, in societies with close social or interpersonal relationships, its frequency is likely to be greater, its identification much more difficult, and its punishment more problematic. This implies that the models of criminal or illegal behavior of the type pioneered by Becker (1968) and by Allingham and Sandmo (1972) will be of limited applicability since some of the requirements of those models (knowledge of probability of getting caught and knowledge of the penalty) would not be satisfied.

The term corruption comes from the Latin verb to break, rumpere. It, thus, implies that something is broken. This something might be a moral or social code of conduct or, more often, an administrative rule. ^1/ If it is the latter, a requirement must be that the rule that is broken is precise and transparent. Another is that the official who breaks it derives some recognizable benefit for himself, his family, his friends, his tribe or party, or some other relevant group. Additionally, the benefit derived must be seen as a direct quid pro quo from the specific act of “corruption”. This simple description reveals several potential difficulties.

First, there must be evidence that a precise rule is broken. This requires that all the rules must be precisely stated, leaving, thus, no doubts about their meaning and no discretion to the public officials. But what about cases where the rules are not precise or where the bureaucrats are specifically given some discretion? For example, in many countries legislation related to the granting of tax incentives or import licenses has often left a lot of discretion in the hands of the officials who must make the decisions to grant them. They must decide whether an investment or an import is “essential” or “necessary” to the country. These officials are often the sole interpreters of what those terms mean. Thus, in a way, they are in a position of monopoly since they can grant or deny these permits and these permits cannot be obtained from other sources. Or, taking a different example, what about the selection of a tax return for an audit; or, the selection of an enterprise for an in sight tax inspection when there are no precise rules for making these choices? Or the hiring of a new employee when there are no precise objective and honest tests for the selection?

Over the years there has been a lot of controversy among economists on whether economic policy should be guided by precise rules or should have an element of discretion. ^1/ It is evident that the greater is the element of discretion, the greater is the possibility that it might be used to someone’s (rather than the public’s) advantage. Thus the possibility of corruption would seem to create a presumption in favor of precise and rigid rules. ^2/ But, of course, the creation of such rules can itself reflect an attempt on the part of some officials to benefit in a large way from that creation. Some rules may be created just to give some government officials the power to benefit from their applications. Often, it is precisely the excess of rules that creates a fertile ground for corruption. Furthermore, the lack of discretion can make the rules too rigid and create obstacles to the well functioning of the economy or the particular organization. At times workers who have wanted to embarrass an organization have complied rigidly with the existing rules bringing the organization to a stand still. ^3/

Second, when social relations tend to be close and personal, it may be difficult to establish a direct link between an act that could be assumed to reflect corruption and a particular payment for it. An employee who uses his official position to favor an acquaintance--say helps him or her get a valuable license, a government contract, or a government job--may be compensated with an immediate or explicit payment (clearly a bribe). Alternatively, he may be compensated, at a much later time, with a generous gift to his daughter when she gets married; or with a good job offer for his son when he completes his studies. In other words, there may not be any direct, explicit, and immediate compensation for the favor. The payment may be delayed in time and, when made, it may appear completely unconnected with the favor received. ^1/ In many cases the “corrupted” and the “corruptor” may never even have discussed the payment. It would simply be understood that a favor today creates a presumption or even an obligation for a reciprocal favor later. In other words, it contributes to the growth of the giver’s “social capital.” In some societies a shadow market for favors develops with demand and supply and with implicit prices. This market which often does not use money but trades in what could be considered the equivalent of “I owe yous” could lend itself to economic analysis. Implicit prices for favors are established and possibly even discount rates for future favors. A kind of honor system guarantees that favors received today will be repaid tomorrow. In this market, it becomes very difficult to separate genuine favors from favors that are close to being bribes and is thus difficult to clearly identify bribes and punish those who receive or pay them.

This takes us to the third and final difficulty. In societies where family, or other kinds of relationships are very strong and especially where existing moral or social codes require that one helps family and friends, the expectation that the public employee will routinely apply arm’s length principles in his relations with friends and relatives is unrealistic. In these societies the Weberian type of ideal bureaucracy will prove very difficult to install. Century old and widely accepted social norms will often prove more powerful as guides to behavior than new and often imported rules based on arm’s length, impersonal and universalistic principles. When this reality is ignored, disappointment is likely to follow. This explains why some reforms imposed by foreign powers or promoted by foreign advisors, which may implicitly require or assume arm’s length relationship, often do not survive the test of time.

In these societies, the cost of the corrective role of the government in the market is likely to rise. Economic relationships within the private sector will also be affected, thus rendering more difficult the establishment of a well working market economy. ^2/ To argue that the personal relationships that come to be established between public sector employees and individuals who deal with them reflect a “corrupt” society may be correct in a legalistic sense but it misses the point that these relationships simply reflect different social and moral norms.

The instruments that make corruption possible are many. Important examples include: (a) regulations (such as the issuance of licenses of various kinds; zoning and other sorts of regulations which may have great economic value; permits of various kinds); (b) fines for alleged or actual violations of existing legal norms; (c) control over procurement contracts; (d) control over public investment contracts (roads, airports, bridges) which can benefit some areas over others, and some contractors over others; (e) programs related to the provision of tax incentives, subsidized credit, overvalued foreign exchange; (f) controls over hiring and promotions; (g) controls over the assignment of entitlements and other benefits (pensions for disability, scholarships, subsidies); (h) controls over access to underpriced public services (electricity, telephone, water); (i) tax administration decisions (auditing, determination of presumptive income, etc.).

These examples are far from exhaustive. The greater is the use of these instruments by a country, the greater will be the potential for corruption. Control over these instruments can give government employees great power which, given the right social environment, the right incentive systems, and weak and uncertain penalties, may allow them to extract large rents for themselves or for their families and friends. Gary Becker’s analysis of crime, or Allingham and Sandmo’s analysis of tax evasion, can be applied to the analysis of corruption when it occurs in an environment more like that of the Danish village than that of the Indian village. In the latter case, the difficulty of identifying many acts of corruption, of proving that a bribe has actually been paid, and of inducing society to apply significant penalties to these acts is likely to reduce the relevance of that analysis.

V. Economic Consequences of Corruption

When civil servants appropriate, for their own use, the instruments that the government has at its disposal to influence the economy and to correct the shortcomings of the private market, they reduce the power of the state and its ability to play the intended and presumably corrective role. In a way this represents a privatization of the state where its power is not shifted to the market, as privatization normally implies, but to government officials and bureaucrats. This will inevitably damage the functioning of the market.

Assuming that government policies had been or would have been guided by the traditional criteria that justify governmental action, ^1/ corruption distorts the end result in several ways.

VI. Policy Implications

Several factors are likely to determine the extent to which corruption will play a significant role in a country. These are: (a) the role of the state and the range of instruments it uses to pursue that role; (b) the social characteristics of the society, i.e., the extent to which arm’s length relationships prevail in social and economic relations; (c) the nature of the political system; and (d) the penalty system for acts of corruption that are uncovered.

In some countries corruption is prevalently an activity of the political leaders. In others, it is prevalently an activity of the civil servants. In still others, it is an activity of both. In a truly democratic system, with checks and balances exercised through fair elections, through the Parliamentary process, and through a vigorous free press, the extent of corruption by the political leaders will generally be checked or, at least, it will eventually be discovered and hopefully controlled. Corruption will exist but it will rarely reach the extreme level reached in some authoritarian governments (see Harsch, 1993). ^1/ Therefore, if, as it has been argued in a recent bestseller, the inevitable course of history is to transform authoritarian into democratic governments, because of this change, the future should experience less corruption than the past or even the present (see Fukuyama, 1992). But let us consider the other factors that determine corruption focussing mainly on the role of the state and on the social characteristics of the countries’ citizens.

Especially in societies where arm’s length relationships are unlikely to be enforceable (because of the close and continuous contacts among closely-knit groups of citizens who tend to personalize most relations), the larger is the role of the state, the greater is the probability that its instruments will be used by public officials and civil servants to favor particular groups in addition to themselves. When this happens, the cost of government rises and its ability to correct the shortcomings of the market falls. In other words, the effective control that the government has on the economy is reduced. ^2/ In this situation the best policy to reduce corruption will be a reduction of the opportunities to engage in it by scaling down the government’s role in the economy. Both the demand for and the supply of corruptive practices can be contained by a sharp reduction of that role in all its aspects, i.e., spending and taxing activities and, especially, in economic regulations. It is no accident that in centrally planned economies, where regulations of economic activity were most widespread, corruption seemed to be a far more common occurrence than in many other countries and to affect not just the bureaucracy but the whole machinery of government. In his Pulitzer Prize winning account of the last days of the Soviet Empire, David Remnick writes that: “The Communist Party apparatus was the most gigantic mafia the world has ever known” and that “the Party’s corruption under Brezhnev was not a matter of exception …” (p. 183). “But in the Soviet Union no economic transaction was untainted. It was as if the entire Soviet Union were ruled by a gigantic mob family; virtually all economic relations were, in some form, mafia relations.” (p. 185). ^1/

Unfortunately, in countries that had been centrally planned, the process of reducing the role of the state in the economy (by freeing prices, by privatizing state enterprises, etc.) may itself produce enormous opportunities for bureaucratic corruption during the transition when the institutions necessary to limit it have not been developed while the habits developed in the previous period may not have changed. ^2/

When corruption characterizes modern states in the Weberian sense, it can be reduced by increasing penalties on those who engage in it, by increasing the transparency of the rules, regulations, and laws, and by strengthening audits, checks, and other controls on the civil servants. In this environment, the analysis suggested by Becker will be relevant. However, when it characterizes more traditional societies, this option, while still worth pursuing, is not likely to give by itself very positive and, especially, permanent results. History is full of examples of campaigns against corruption (and against tax evasion) that started with great fanfare but that, over the long run, did not accomplish much. By the same token, one should not officially sanction corruption by, for example, reducing the wages of the civil servants on the assumption that they are getting payments under the table. In an African country, for example, three years ago, the government reduced to zero, for six months, the wages of customs officials on the assumption that “they could take care of themselves.” ^3/ Unrealistically low wages always invite corruption and at times lead society to condone acts of corruption. This is the reason why repression of public sector wages, if carried too far, is never a good policy.

Because it is social intimacy that creates the environment that promotes corruption, a policy that has been effective in some cases (for example in tax administration) in reducing corruption is that of forced and periodic geographical mobility for civil servants, in order to remove them from the region where they have their closest social or family relations and to prevent the formation of new relations. Some forms of social relations take time to develop so that for a while, after a government official has moved to a new region or to a new function, they will not play a large role in the contacts between the bureaucrats and the citizens who depend on them. Thus, periodic mobility, especially in a large country, could be an effective policy to reduce bureaucratic corruption. It is no accident that in large countries, such as the United States, corruption is more a problem for local governments where bureaucrats and citizens often know each other and where arm’s length relationships are less likely to prevail. In general, one can speculate that the larger is a country and the more mobile is the population, the less of a problem will corruption be.

VII. Concluding Remarks

Economists have developed elaborate and elegant theories about the workings of markets and the role of public sector’s action in those markets. A normative role has been assigned to the government aimed at correcting for market failure. Public choice economists have, in recent years, stressed that, in addition to market failure, one could also have political failure when political action or the action of civil servants are influenced by objectives other than the need to correct market failures and to promote the public interest.

Over the past decade, economists have begun to pay some attention to phenomena that were largely ignored in earlier periods such as underground economic activities, tax evasion, money laundering, and corruption. These represent the dark side of the economy. All these activities have major implications for the functioning of markets. When these activities are present, markets do not operate efficiently. Acts of corruption by public sector officials often play a role in promoting or sustaining underground economic activities and in facilitating tax evasion and the laundering of dirty money. These acts are facilitated or even stimulated by close interpersonal relations that characterize some societies. They are also facilitated by the many instruments that governments use to promote their often very active role in the economy.

Given the existence of close interpersonal relations in a society, and given that a government is pursuing a very active role in the economy through various policy instruments that lend themselves to be used by some public sector employees for personal gains, corruption is likely to grow with the passing of time. The passage of time will submit public sector employees to growing pressures from friends, acquaintances, relatives or simply those who want to buy favors from them; and the passage of time will teach some government officials (those who are more likely to abuse their power) how to take advantage of the situation. ^1/ This may explain why corruption grew over the years in centrally planned economies and in some of the economies which tried to promote a welfare state.