Corruption is a complex phenomenon. Its roots lie deep in bureaucratic and political institutions, and its effect on development varies with country conditions. But while costs may vary and systemic corruption may coexist with strong economic performance, experience suggests that corruption is bad for development. It leads governments to intervene where they need not, and it undermines their ability to enact and implement policies in areas in which government intervention is clearly needed-whether environmental regulation, health and safety regulation, social safety nets, macroeconomic stabilization, or contract enforcement. This chapter looks at the complex nature of corruption, its causes, and its effects on development.

How do we define corruption?

The term corruption covers a broad range of human actions. To understand its effect on an economy or a political system, it helps to unbundle the term by identifying specific types of activities or transactions that might fall within it. In considering its strategy the Bank sought a usable definition of corruption and then developed a taxonomy of the different forms corruption could take consistent with that definition. We settled on a straightforward definition-the abuse of public office for private gain.1 Public office is abused for private gain when an official accepts, solicits, or extorts a bribe. It is also abused when private agents actively offer bribes to circumvent public policies and processes for competitive advantage and profit. Public office can also be abused for personal benefit even if no bribery occurs, through patronage and nepotism, the theft of state assets, or the diversion of state revenues. This definition is both simple and sufficiently broad to cover most of the corruption that the Bank encounters, and it is widely used in the literature. Bribery occurs in the private sector, but bribery in the public sector, offered or extracted, should be the Bank's main concern, since the Bank lends primarily to governments and supports government policies, programs, and projects.

Bribery:

Bribes are one of the main tools of corruption. They can be used by private parties to "buy" many things provided by central or local governments, or officials may seek bribes in supplying those things.

Government contracts:

Bribes can influence the government's choice of firms to supply goods, services, and works, as well as the terms of their contracts. Firms may bribe to win a contract or to ensure that contractual breaches are tolerated.

Government benefits:

Bribes can influence the allocation of government benefits, whether monetary benefits (such as subsidies to enterprises or individuals or access to pensions or unemployment insurance) or in-kind benefits (such as access to certain schools, medical care, or stakes in enterprises being privatized).

Lower taxes:

Bribes can be used to reduce the amount of taxes or other fees collected by the government from private parties. Such bribes may be proposed by the tax collector or the taxpayer. In many countries the tax bill is negotiable.

Licenses

Bribes may be demanded or offered for the issuance of a license that conveys an exclusive right, such as a land development concession or the exploitation of a natural resource. Sometimes politicians and bureaucrats deliberately put in place policies that create control rights which they profit from by selling. Time:

Bribes may be offered to speed up the government's granting of permission to carry out legal activities, such as company registration or construction permits. Bribes can also be extorted by the threat of inaction or delay.

Legal outcomes:

Bribes can change the outcome of the legal process as it applies to private parties, by inducing the government either to ignore illegal activities (such as drug dealing or pollution) or to favor one party over another in court cases or other legal proceedings.

The government benefits purchased with bribes vary by type and size. Contracts and other benefits can be enormous (grand or wholesale corruption) or very small (petty or retail corruption), and the impact of misinterpretation of laws can be dramatic or minor. Grand corruption is often associated with international business transactions and usually involves politicians as well as bureaucrats. The bribery transaction may take place entirely outside the country. Petty corruption may be pervasive throughout the public sector if firms and individuals regularly experience it when they seek a license or a service from government. The bribes may be retained by individual recipients or pooled in an elaborate sharing arrangement. The sums involved in grand corruption may make newspaper headlines around the world, but the aggregate costs of petty corruption, in terms of both money and economic distortions, may be as great if not greater.

Theft:

Theft of state assets by officials charged with their stewardship is also corruption. An extreme form is the large-scale "spontaneous" privatization of state assets by enterprise managers and other officials in some transition economies. At the other end of the scale is petty theft of items such as office equipment and stationery, vehicles, and fuel. The perpetrators of petty theft are usually middle- and lower-level officials, compensating, in some cases, for inadequate salaries. Asset control systems are typically weak or nonexistent, as is the institutional capacity to identify and punish wrongdoers.

Theft of government financial resources is another form of corruption. Officials may pocket tax revenues or fees (often with the collusion of the payer, in effect combining theft with bribery), steal cash from treasuries, extend advances to themselves that are never repaid, or draw pay for fictitious "ghost" workers, a pattern well documented in the reports of audit authorities. In such cases financial control systems typically have broken down or are neglected by managers.

Political and bureaucratic corruption:

Corruption within government can take place at both the political and the bureaucratic levels. The first may be independent of the second, or there may be collusion. At one level, controlling political corruption involves election laws, campaign finance regulations, and conflict of interest rules for parliamentarians. These types of laws and regulations lie beyond the mandate and expertise of the Bank but nevertheless are part of what a country needs to control corruption.2 At another level corruption may be intrinsic to the way power is exercised and may be impossible to reduce through lawmaking alone. In the extreme case state institutions may be infiltrated by criminal elements and turned into instruments of individual enrichment.

Isolated and systemic corruption:

Corruption in a society can be rare or widespread. If it is rare, consisting of a few individual acts, it is straightforward (though seldom easy) to detect and punish. In such cases noncorrupt behavior is the norm, and institutions in both the public and private sectors support integrity in public life. Such institutions, both formal and informal, are sufficiently strong to return the system to a noncorrupt equilibrium. In contrast, corruption is systemic (pervasive or entrenched) where bribery, on a large or small scale, is routine in dealings between the public sector and firms or individuals. Where systemic corruption exists, formal and informal rules are at odds with one another; bribery may be illegal but is understood by everyone to be routine in transactions with the government. Another kind of equilibrium prevails, a systemic corruption "trap" in which the incentives are strong for firms, individuals, and officials to comply with and not fight the system. And there may be different degrees of coordination between those taking bribes, ranging from uncontrolled extortion by multiple officials to highly organized bribe collection and distribution systems. Antibribery laws notwithstanding, there are many countries in which bribery characterizes the rules of the game in private-public interactions. Systemic corruption may occur uniformly across the public sector, or it may be confined to certain agencies-such as customs or tax authorities, public works or other ministries, or particular levels of government.3

Corruption in the private sector:

Fraud and bribery can and do take place in the private sector, often with costly results. Unregulated financial systems permeated with fraud can undermine savings and deter foreign investment. They also make a country vulnerable to financial crises and macroeconomic instability. Entire banks or savings and loan institutions may be taken over by criminals for the purpose of wholesale fraud. Popular support for privatization or the deepening of financial markets can be eroded if poor regulation leads to small shareholders or savers withdrawing when confronted by insider dealings and the enrichment of managers. And a strong corporate focus on profitability may not prevent individual employees soliciting bribes from suppliers. Furthermore, when corruption is systemic in the public sector, firms that do business with government agencies can seldom escape participating in bribery.

While noting the existence of fraud and corruption in the private sector and the importance of controlling it, this report is concerned with corruption in the public sector. Public sector corruption is arguably a more serious problem in developing countries, and controlling it may be a prerequisite for controlling private sector corruption.4 Still, Bank activities can also promote the control of bribery and fraud in the private sector by helping countries strengthen the legal framework to support a market economy and by encouraging the growth of professional bodies that set standards in areas like accounting and auditing. In the long run, controlling corruption in the private sector may require improvements in business culture and ethics.

What are the causes of corruption?

The causes of corruption are always contextual, rooted in a country's policies, bureaucratic traditions, political development, and social history. Still, corruption tends to flourish when institutions are weak and government policies generate economic rents. Some characteristics of developing and transition settings make corruption particularly difficult to control. The normal motivation of public sector employees to work productively may be undermined by many factors, including low and declining civil service salaries and promotion unconnected to performance. Dysfunctional government budgets, inadequate supplies and equipment, delays in the release of budget funds (including pay), and a loss of organizational purpose also may demoralize staff. The motivation to remain honest may be further weakened if senior officials and political leaders use public office for private gain or if those who resist corruption lack protection. Or the public service may have long been dominated by patron-client relationships, in which the sharing of bribes and favors has become entrenched. In some countries pay levels may always have been low, with the informal understanding that staff will find their own ways to supplement inadequate pay. Sometimes these conditions are exacerbated by closed political systems dominated by narrow vested interests and by international sources of corruption associated with major projects or equipment purchases.

The dynamics of corruption in the public sector can be depicted in a simple model. The opportunity for corruption is a function of the size of the rents under a public official's control, the discretion that official has in allocating those rents, and the accountability that official faces for his or her decisions.5 Monopoly rents can be large in highly regulated economies and, as noted above, corruption breeds demand for more regulation. In transition economies economic rents can be enormous because of the amount of formerly state-owned property essentially "up for grabs." The discretion of many public officials may also be large in developing and transition economies, exacerbated by poorly defined, ever-changing, and inadequately disseminated rules and regulations. Finally, accountability is typically weak in these settings. The ethical values of a well-performing bureaucracy may have been eroded or never established. Rules on conduct and conflict of interest may be unenforced, financial management systems (which normally record and control the collection of revenues and the expenditure of budgeted resources) may have broken down, and there may be no formal mechanism to hold public officials accountable for results. The watchdog institutions that should scrutinize government performance, such as ombudsmen, external auditors, and the press, may be ineffectual. And special anticorruption bodies may have been turned into partisan instruments whose real purpose is not to detect fraud and corruption but to harass political opponents.

A defining characteristic of the environment in which corruption occurs is a divergence between the formal and the informal rules governing behavior in the public sector. The Bank is unaware of any country that does not have rules against corruption, although not all countries have all the rules that may be necessary. These range from laws making it a criminal offense to bribe a public official to public service regulations dealing with the expected behavior of public officials, conflicts of interest, the acceptance of gifts, and the duty to report fraud. Government agencies-police and army, tax and customs departments, local governments, and public enterprises-may have their own regulations and codes of behavior. Organic laws, often embedded in constitutions, cover budgeting, accounting, and auditing, supported by laws and regulations on public procurement and the safeguarding of public assets. In addition, there are laws on the conduct of elections and the appointment of judges, and codes governing the conduct of legislators. Some of these laws are a colonial inheritance, some have been adapted from countries with a similar legal tradition, and some are additions to existing laws (for example, providing for special anticorruption commissions and other watchdog bodies).

Where corruption is systemic, the formal rules remain in place, but they are superseded by informal rules.6 It may be a crime to bribe a public official, but in practice the law is not enforced or is applied in a partisan way, and informal rules prevail. Government tender boards may continue to operate even though the criteria by which contracts are awarded have changed. Seen in this light, strengthening institutions to control corruption is about shifting the emphasis back to the formal rules. This implies acknowledging that a strong legal framework to control corruption requires more than having the right legal rules in place. It means addressing the sources of informality, first by understanding why the informal rules are at odds with the formal rules and then by tackling the causes of divergence. In some countries the primary reason for divergence may be political, a manifestation of the way power is exercised and retained. This limits what the Bank can do to help outside the framework of its projects. In other countries the reason may be weak public management systems and inappropriate policies, which the Bank can help improve.

What do economics and political science tell us?

While it would be easy to say that all corruption is bad, the Bank must base its approach on evidence and analysis of corruption's effects on development. The political sensitivity surrounding issues of governance underscores the need for such a foundation. In preparing this report, the Bank examined the conclusions drawn by economic researchers working on the topic, the perspectives of disciplines other than economics, and the evidence from the Bank's operational work.

Economic research:

The body of research addressing the economic effects of corruption has grown significantly in recent years. The research is both macroeconomic and microeconomic, theoretical and empirical. Its conclusions depend in part on what the researcher views as the bottom line: short-term economic efficiency in private markets, long-term dynamic efficiency and economic growth, equity and fairness, or political legitimacy.

One strand of literature explores, primarily from a theoretical perspective, the likely economic effects of different forms of corruption.7 Some writings of this group argue that corruption can be efficiency enhancing. First, the argument is made, corruption may not distort the short-run efficiency of an economy if it merely entails a transfer of economic rents from a private party to a government official. Thus a bribe to an official who is allocating, say, foreign exchange or credit in short supply can be seen as a market payment for ensuring that resources go to the party most likely to use them efficiently (the one who can pay the highest bribe). The problem with this line of reasoning is that it fails to take into account any objective other than short-term efficiency. In the long run, expectations of bribery may distort the number and types of contracts put up for bid, the method used to award contracts, and the speed or efficiency with which public officials do their work in the absence of bribes. It may also delay macroeconomic policy reform. In addition, the gains from such bribery may be inequitably distributed (accessible only to certain firms and public officials).

Second, bribes can theoretically increase economic efficiency if they allow firms to avoid overly restrictive regulations or confiscatory tax rates. That is, bribes lower the costs of bad regulations to firms that bribe. There may be some validity to this argument, particularly in the short run. Yet such bribery defuses pressure for broader reform and invites firms to evade good regulations as well as bad. Furthermore, the costs of such a system may fall disproportionately on smaller firms.8 A policy framework based on many legal restrictions and widespread bribery to avoid them is like a highly regressive system of taxes on the private sector, and few would argue for such a system in developing countries. And in some transition economies such restrictions have proliferated in an uncontrolled way with the express purpose of extracting rents. This causes a shift of economic activity to the informal sector.

To summarize, models purporting to show that corruption can have positive economic effects are usually looking only at static effects in the short run. In the long run, opportunities for bribery are likely to lead public officials to change the underlying rules of the game or their own behavior in the absence of bribes, and the results are likely to be costly in terms of economic efficiency, political legitimacy, and basic fairness.

Another strand of literature examines the links between investment, economic growth and the quality of government institutions.9 It finds that weak public institutions, as evidenced by unreliable contract enforcement, unclear property rights, unpredictable policies, inefficient public administration, corruption, and other indicators, significantly reduce private investment and lead to slower growth. While useful in highlighting the broad economic effects of institutional deficiency, much of the literature has been unable to separate the effect of corruption from other dimensions of government quality.10

Finally, there is the uneven performance of countries to contend with. While few would disagree that corruption has undermined development in Africa and has slowed the emergence of well functioning market economies in the former Soviet Union, the coexistence of high growth and systemic corruption in some Asian countries challenges those who believe that corruption is always economically harmful. Several explanations have been suggested. First, perhaps predictability is what matters, and some governments reliably deliver what is "bought" with bribes while other governments do not.11 Second, others view highly concentrated corruption at the top of the political system (cited as more the model in some Asian settings) as less distortionary than uncontrolled corruption at lower levels (as in parts of the former Soviet Union).12 Third, if political systems are well established and the rules of the game are known to all, the transactions costs of rent seeking may be less costly than in less stable, less certain environments. Fourth, corruption may be imposing environmental and social costs that are not captured in national accounts data. We do not know these costs, and country experience differs widely even within Asia. Nobody, however, argues that corruption is good for development, and recent research suggests that corruption may be restraining growth even in Asia.13 What is clear is that the nature and dynamics of corruption vary greatly among countries, making it a diverse and complex phenomenon to address.

Political science:

Political scientists look beyond the visible signs of corruption to the broader setting in which it occurs. They see corruption in relation to the legitimacy of the state, the patterns of political power, and the engagement of civil society. Corruption may be a manifestation of the way political power is contested and exercised. To the leadership the creation and allocation of state rents serves political purposes: rewarding supporters, buying off opponents, ensuring the backing of key groups, managing ethnic diversity, or simply accumulating resources to fight elections. To obtain these resources, leaders may forge alliances with business groups or create and distribute rents through the bureaucratic apparatus. The resulting policies may favor or discourage capital accumulation and economic growth, depending on the nature of the alliances struck.14 Politicians in such countries may be aware of the distortionary consequences of such rents but view them as a necessary tool of political management. If this is the case, the pattern of corruption will change only if the power structure changes, which may result from a popular outcry against corruption.

Political scientists also take a historical perspective. Over time most industrial countries have developed merit-based bureaucratic values, institutionalized competitive politics, established transparent government processes, and fostered an active media and an informed civil society. These mechanisms constrain political and bureaucratic corruption, making it the exception rather than the norm. The transition may be spurred by an enlightened ruler or, more likely, by the growing power of new political groups with an interest in better-performing government. In developing countries, in contrast, government institutions are weaker, civil society is less engaged, and political and bureaucratic processes are less accountable and transparent. An effective state apparatus and capacity for law enforcement may be virtually nonexistent. In such settings, sustained progress in building an honest and effective state apparatus requires addressing the mix of factors in the state and in society that give rise to both corruption and weak social and economic performance. This is an exceedingly complex and long-term effort.15

Public management:

The public management view of corruption is clear-cut. Systemic corruption, in the form of graft and patronage and the inefficiencies that accompanied it, spurred the nineteenth-century reforms in Europe and North America that created the modern bureaucratic state. Corruption opposes the bureaucratic values of equity, efficiency, transparency, and honesty. Thus it weakens the ethical fabric of the civil service and prevents the emergence of well-performing government capable of developing and implementing public policies that promote social welfare.

The machinery of modern government, as it evolved in industrial countries and has been transferred to developing countries, includes systems that protect public organizations from corruption and promote accountability. These systems, including a meritocratic civil service and watchdogs such as supreme audit institutions, ombudsmen, and public service commissions, should not be neglected. Some OECD countries seeking to improve government performance through New Public Management reforms are developing "risk management" perspectives on corruption. But they do so within a framework of strong financial management control systems and a renewed emphasis on the ethical values of public service. While economies may still grow in countries in which corruption is entrenched in the public sector, the public management view is that successive stages of economic and social development will be harder if not impossible to achieve without well-performing government. Ultimately, countries need to create durable institutions to foster and protect integrity in public life if public policy is to achieve the objectives (such as poverty reduction and environmental protection) that are at the core of sustainable economic and social development.

What is the Bank's experience?

Most of the economic and sector work undertaken by the Bank does not directly address corruption. However, a small but growing number of public expenditure reviews have drawn attention to the problem, and survey data gathered in the course of private sector assessments are beginning to illuminate the costs of bribery to entrepreneurs. Based on the economic and sector work that does address the topic, informal country knowledge within the institution, and examples from the Bank's vast store of country reports in which the influence of corruption can be inferred (even if the term is seldom used), the following picture emerges of the many ways in which corruption imposes costs on our borrowers.

Macroeconomic stability may be undermined by loss of government revenue and excessive spending. This can happen through corruption in tax and customs departments, through debt incurred when the scrutiny of finance ministries and central banks is bypassed, through contracts that are awarded to high-cost bidders or without competitive tendering, and through the general erosion of expenditure control. Excessive debt may be incurred through "white elephant" investment projects that owe their origin, in part, to bribes. Macroeconomic stability may also be threatened by debt guarantees and other off-budget contingent liabilities agreed to in corrupt transactions without public scrutiny. It may also be threatened by fraud in financial institutions, leading to loss of confidence by savers, investors, and foreign exchange markets. In transition economies and in many developing countries corruption may reduce revenue collection by driving firms (or their most profitable activities) out of the formal sector and by providing a moral justification for widespread tax evasion.16 The costs of macroeconomic instability are borne by all elements in society but especially by the poor.

Foreign direct investment may still flow to countries in which corruption is systemic but only if bribery is affordable and the results are predictable. Even so, corruption can have a negative effect on foreign investment. Where corruption is large and systemic, investment may be concentrated in extractive industries in which operations can be enclaved, or in light manufacturing or trading operations that can be relocated if corruption costs become unbearable. Or foreign investors may shun the country altogether. For most foreign firms, corruption is a cost of doing business to be recouped from revenues.17 If the costs become too high or unpredictable, foreign firms will disengage unless global marketing or sourcing considerations require them to maintain a presence in that country. High levels of corruption add to the risk of a country being marginalized in the international economy.

Small entrepreneurs may be affected in many developing and transition economies. Evidence from private sector assessments suggests that corruption increases the costs of doing business, that small firms bear a disproportionately large share of these costs, and that bribes can prevent firms from growing.18

The environment may be endangered:

Many countries have enacted laws to protect the environment and have created special agencies to enforce these laws, but there is often a disconnect between policy and its implementation. Complying with environmental regulations imposes on firms costs that can be avoided by bribery. There are huge rents to be earned from activities such as logging in tropical rain forests, where permits can be obtained corruptly or where inspectors can be bribed. The environmental costs of corruption may take the form of ground water and air pollution, soil erosion, or climate change, and can be global and intergenerational in their reach.

The poor suffer:

While poverty assessments have focused more on measuring poverty than explaining it,19 anecdotal and survey evidence reveal the cost of petty corruption to the poor. When access to public goods and services requires a bribe, the poor may be excluded. Given their lack of political influence, the poor may even be asked to pay more than people with higher incomes.20 Furthermore, when corruption results in shoddy public services, the poor lack the resources to pursue "exit" options such as private schooling, health care, or power generation.

A survey of 3,600 firms in 69 countries carried out for the 1997 World Development Report provides further evidence of the widespread existence and negative effects of corruption. As noted in the report:

The survey confirmed that corruption was an important-and widespread -problem for investors. Overall, more than 40 percent of entrepreneurs reported having to pay bribes to get things done as a matter of course. . . . Further, more than half the respondents worldwide thought that paying a bribe was not a guarantee that the service would actually be delivered as agreed, and many lived in fear that they would simply be asked for more by another official. . . . The consequences of corruption often do not end with paying off officials and getting on with business. Government arbitrariness entangles firms in a web of time-consuming and economically unproductive relations. . .21

The survey also confirms the negative correlation between the level of corruption (as perceived by businesspeople) and the level of investment in an economy.