reveal the value of bribes transacted. Collusion makes the separation of powers ineffective in providing checks and balances. Public monitoring by the press and civil society then remains the most effective way to deter corruption. Corruption is not so much due to the culture of greed and the dispassionate pursuit of self interest, or due to wrong policies such as low wages in the public sector, but stems from the inherent design of non-market institutions set up to promote the social good.

Paper written for the project entitled "State of Democracy in South Asia" at Lokniti (Institute of Comparative Democracy), Centre for the Study of Developing Societies, New Delhi.


Introduction

There are different accounts that are presented with regard to corruption. Some prefer to follow up on the individual misdeeds of politicians and bureaucrats and make comments about greed, probity in public life, and the need for individuals with high morals and vigilance by civil society. Others lay the blame on the types of policies followed by the state, such as low levels of wages in the bureaucracy, high government intervention in the economy, and the lack of competition from efficient foreign players that generate incentives to create rents which are skimmed away through corruption. In sharp contrast, this paper takes the position that it is not so much the type of person and his propensity to be corrupt, or policy variables that create avenues for corruption, but rather the nature of public institutions that we should focus on when thinking about corruption. Taking a cue from some of the recent literature (Banerjee, 1997; Acemoglu and Verdier, 2000) this paper argues that the case for governments to intervene is when markets cannot be relied on to allocate resources efficiently or equitably. This requires the use of institutions – bureaucracies – to implement policies and make decisions about allocation. Bureaucrats have discretion and power because there is an information asymmetry as to which agents are to be allocated resources through this non-market institution. Red tape is used as a screening device to sort out those individuals who deserve the resource. Red tape is therefore useful but is subject to opportunism as some individuals may masquerade as belonging to the category of those deserving an allocation of the resource when they are not and can even resort to speed money. In cases where the red tape separates those deserving of resource transfers from the undeserving, collusion and side-payments (bribes) between the bureaucrat and citizen by concealing information also promotes corruption. We also distinguish between external corruption and internal corruption – collusion between bureaucrats in an organization, typically between a supervisor and his subordinate, that allows subordinates to transfer proceeds from bribes to higher levels in the hierarchy in return for protection. Finally, the bureaucracy is supposed to be accountable to politicians. However, politicians and bureaucrats can also collude and politicians in India exercised control over an otherwise independent bureaucracy through using their power over transfers and postings. Transfers reveal the value of the bribes transacted in a post as bureaucrats bid for lucrative posts. Politicians share in the proceeds and replenish party funds which are then used to influence the electorate.

When different branches of the government collude with each other separation of powers is ineffective in deterring corruption. In such situations public monitoring by the press and other elements of civil society is the important means to deterring corruption. Given the high levels of collusion between bureaucrats and those wrongly favoured with non-market public allocations, between different levels of the bureaucratic hierarchy, and between bureaucrats and politicians, increasing transparency and information in the public domain about the social costs of such activities makes it more difficult for them to thrive. When there is corruption government activity which is promoted to allocating resources conditional on recipients of those resources possessing certain attributes (such as being poor for instance), ends up transferring resources unconditionally and this is socially inefficient. Government allocation is preferable to market allocation but is costlier because corruption may misallocate resources even more than the normal process of bureaucratic allocation without corruption would. This does not mean asking for a complete withdrawal of the state because bureaucratic allocation is still socially useful. Only when the market failure is not pronounced enough to justify government intervention i.e., the cost of monitoring and allocation by bureaucracies is higher than the misallocation by markets, is the case against government intervention really strong. Otherwise, it requires us to design mechanisms that make corruption costly for bureaucrats and politicians. After looking at some empirical evidence on corruption we expand on this theme in the rest of this essay.


Empirical Pointers

Corruption is difficult to study empirically but in the recent past a literature on what makes corruption more widespread in some countries than others has begun to emerge. Daniel Treisman, a political scientist from the University of California Los Angeles, has done the most extensive cross national study on corruption (Treisman, 2000). Treisman develops a series of hypotheses¹ about corruption:

Amongst the interesting results from this study of 100 countries is that whether a country is democratic today or the quality of its democracy as measured by Freedom house’s rating of political rights makes no difference to how corrupt it is. What seems to matter is whether or not it has been democratic for decades. Countries with a long duration of democracy had some impact on reducing corruption. The impact of democracy on corruption is small – those countries with at least 40 years of consecutive democracy behind them enjoyed a small corruption dividend,⁵ -40 consecutive years of democracy reduces corruption by close to one point on a 10-point scale.

Openness to trade reduces corruption but to make a noticeable difference it would require a radical opening rather than a marginal shift – Turkey would require to boost the share of imports in GDP from about 23% in 1994 to about 80% to reduce its corruption rating by about one point.

Long-lived aspects of countries’ cultural or institutional traditions affect corruption more significantly than current state policies. – (a) the percentage of Protestants in the population is a robust predictor of lower corruption. Protestant churches often developed in counterpoint with the state and may have stimulated a more autonomous civil society. The impact of religion is not so much cultural as institutional.⁶ (b) Countries that were British colonies have significantly lower perceived corruption.⁷ This British heritage effect is linked to the inheritance of a common law tradition. In common law systems, law is made by judges on the basis of precedent, rather than on the basis of codes drawn up by scholars and promulgated by governments. British colonizers left behind an accumulation of precedents and case law and a legal culture that emphasized procedural justice over substantive issues far more than in countries colonized by other powers.

There was strong evidence that the process of economic development reduces corruption – richer countries are less corrupt than poor ones. This occurs presumably through the rationalization of public and private roles and the spread of education which makes abuses harder to conceal.

A study by Lederman, Loayza and Reis Soares (2001) found contrary to Treisman’s results that openness of an economy has no effect on corruption whereas British legal tradition which has a negative effect in Treisman’s paper has a positive and significant effect on corruption. This difference is not due to differences in data sets. Lederman et al show that if they remove the political variables from the regression specification then British legal tradition has a negative effect on corruption. British legal tradition is strongly associated with democracy, stability, and freedom of the press and once these political variables are taken into account the culture associated with the British legal tradition – the informality of British law, where practices are based on unwritten rules – seems to be more subject to corruption than civil law traditions. After political institutions are accounted for Lederman et al find that variables that are usually associated with corruption – openness, wages in the public sector, legal tradition, etc – lose all their relevance as explanatory variables for corruption.

What emerges is that policy decisions have little impact on corruption. Higher civil service wages may not reduce corruption and the extent of state intervention is highly correlated with corruption – a correlation that vanishes when we control for predetermined factors such as the level of economic development that may influence both policy choices and corruption levels. If the extent of state intervention or wage differentials between the public and private sectors have little significant impact on corruption then we need to focus less on policy and more on institutions and history. Specifically we need an understanding of what it is in state intervention that produces undesirable outcomes such as corruption. To do this we proceed by constructing an account of the design of state intervention and then use that to understand how the design is undermined through corruption.


State Intervention and Design Subversion

Let us begin with a proposition that few would disagree with – the choice between markets and governments is not binary. The choice is often between different combinations of the two and different degrees of one or another mode of allocating resources. Even if the predominant choice to allocate resources is in favour of the market, a significant role for the government still remains. This role centres around the production of pure public goods such as define and national security, to establishing and maintaining the legal and other environmental conditions that are essential for the functioning of markets, and to the provision of redistributive services and programmes that reflect the standards of support and distributive justice which the society implements. In short, markets cannot be called on to allocate resources in many situations where there are externalities – the private and social values of goods are different. Many government activities revolve around this setting – the issuing and renewing of licences, passports, product quality certificates, assigning procurement contracts, etc. Even where the state provides private goods in cases such as welfare benefits and food rations, the social value is higher than the private value as there is a social preference for equity. Because social value differs from private value consumers cannot be allowed to allocate the benefits through markets. For example, we can’t expect a suspect to arrest himself if given a choice between that and setting himself free (Prendergast, 2003). As consumers cannot be trusted to allocate such goods efficiently an alternative mechanism is used which is a bureaucracy. When consumers cannot be provided for by markets because they exclude in an inefficient manner bureaucracies are required.

Government intervention thus requires bureaucrats to collect information, make decisions, and implement policies. The pattern of government intervention is usually explained via a ‘public interest’ theory or an ‘interest group’ or capture theory (Laffont and Tirole, 1991). The public interest theory emphasizes the government’s role in correcting market imperfections and agencies of the government are viewed as benevolent maximizers of social welfare. The interest group theory emphasizes the role of interest groups in public policy formulation. This work has ranged from emphasizing how big business has influenced public policy to how small business groups can capture regulatory processes because they find it easier to overcome collective action problems.⁸ These attempts to understand government intervention overlook an important issue. They do not take into account information asymmetries, the absence of which makes it difficult for interest groups and politicians to initiate policies that favour some agents over others. It is the lack of perfect information that results in the possibility of discretion and is the basis of stakes in the system and the exercise of power. The information asymmetry arises either because there is private information with the interest group – it has information about an activity which results in a divergence between private and social costs for instance such as the use of a technology which damages the environment – or the bureaucrat has private information because he has control over a good or service that is to be allocated.

The reason bureaucracies exist is because there is a need for an agency that collects information ad implements policies in situations where markets cannot allocate resources efficiently. Allocating resources such as procurement contracts and licenses is to be done on the basis of the fulfillment of certain criteria on the part of the persons who are to be provided the resource. In the case of an anti poverty programme for instance the person should be poor. Screening by the bureaucrat establishes whether a person meets the allocative criterion that has been devised to minimize the distortion created by the diversion between social and private benefits and costs if the good were to be priced in markets. Screening is done through the use of red tape – a neutral term for the device which may be a questionnaire or a test that allows the bureaucrat to distinguish between types of persons and select those who should be allocated the resource. Red tape helps sort out those who deserve from those who do not deserve the resource and administering red tape requires specific skills such as those possessed by a bureaucracy. Red tape is therefore informative but also costly (Guriev, 2004).

Red tape is useful but gives rise to two problems that are associated with asymmetric information – adverse selection and moral hazard. Adverse selection arises because the type of individual to whom the resource is to be allocated is known only to that individual and can be verified only through red tape. Some individuals may masquerade as a type when they actually belong to another group type (the nonpoor for instance posing as the poor so as to receive the benefits of an anti­poverty programme). Those who are not allocated the resource correctly due to this deficiency in the screening mechanism could complain and since information about an incorrect allocation of the resource is revealed the upper rung in the hierarchy of the bureaucracy could intervene and affect the prospects of the bureaucrat who is in charge of implementing the allocation. Bureaucrats are not monitored via pay for performance as could be the case in the private sector (it would be ridiculous to pay police officers on the basis of the number of arrests as that would give wrong incentives) and the standard way of monitoring them is through investigations (via commissions, Tribunals, etc) when a mistake or a malfeasance is reported. Bureaucracies respond to this imperfection in the screening mechanism and the resulting possibility of a customer complaint and investigations that can affect career paths by increasing the amount of red tape. This caused delays but it increases the chance that the resource gets in the hands of the beneficiary for whom it was intended. I do not want to be apologetic or supportive of most bureaucratic inefficiencies but it does seem to be the case that due to the problems of screening in a non-market allocation of a resource excessive red tape emerges as a way of addressing the chance of adverse selection and customer complaints.

With excessive red tape the informativeness of the screening mechanism is more readily available to the bureaucrat implementing the allocation than to others in the bureaucratic hierarchy. Those who apply for the resource now face the problem that because of asymmetrical information about whether they are really the type of beneficiary the resource was intended to be transferred to, there has emerged a longer application process which lays down several criteria they must possess. Waiting costs are now higher and the bureaucracy takes longer to make decisions. Their valuation of the resource that the bureaucracy allocates may not be high compared to the waiting cost and they may drop out of the queue. For a number of them, however, the valuation of the resource they are applying for is high and they are not prepared to wait and go through the procedures laid down in the red tape. So they collude with the bureaucrat by providing “grease money” that makes the application process shorter and reduces their waiting costs. In this case speed money is being paid to reduce the amount of red tape before the applicant is screened for being worthy of an allocation or not. The bribe reduces the waiting costs but could be negotiated by a person who would be provided the resource eventually as society had valued giving it to him over other applicants (the deserving) or offered by a person who would not under normal screening and evaluation processes have been considered for the resource transfer from the government (the undeserving).

Of course, realizing the unintended consequence of red tape corrupt officials may delay the queuing process and put in place tedious reporting requirements and form filling so as to induce bribe giving (Myrdal, 1968). And yet again we could have what should strictly be labeled not bribery, but extortion – the act of threatening an applicant with some cost (such as exclusion from an allocation) by a bureaucrat unless he is benefited with a money transfer or something of value to which he has no prior right. In contrast to extortion, a bribe is a payment of money or something of value to a bureaucrat in exchange for his giving special consideration that is incompatible with the duties of his office. Though this distinction is useful we do not pursue here whether a citizen offered a benefit or a bureaucrat demanded it. Some analysts (Leff, 1964; Lui, 1985) consider grease money as a lubricant to the squeaky wheels of a bureaucracy. Whilst conceding that in the case of a deserving applicant it may increase growth through a reduction in waiting time we do not subscribe to their view that growth occurs through overcoming the obstacle of a rigid and often inefficient public administration. For us public administration is necessarily slow and imposes large costs on customers but this emerges from the good or resource not being efficiently allocated via markets. Bureaucratic allocation is inefficient because markets are not efficient either.

Once red tape has done the job of screening those in whose control the resource would provide the social benefits that are being sought to be achieved by bureaucratic allocation, then, yet again there is the likelihood of collusion between a bureaucrat and a citizen. In this case the bureaucrat and the citizen collude to conceal what the information emanating from red tape has revealed. This is different from the speed money case in that it occurs after information about the suitability of the applicant for the resource transfer is available. The incentive to increase red tape here is even greater for a bureaucracy that is intent on receiving bribes. On the basis of the information received it is revealed to the bureaucrat who are the genuine deserving and who not. The genuine deserving customer can be labeled a legal applicant whose application should result in the transfer of the resource by the bureaucrat. The bureaucrat could, however, decide not to process this application and try to extort something of value from the applicant (think of the policeman who demands money to send a valid police clearance certificate to the passport office). The information from red tape also reveals who is an undeserving applicant and instead of rejecting this person the bureaucrat may collude with this person and obtain a reward for hiding the information as to his status if the true information was available.

So far we have not brought in the fact that bureaucratic organizations are not unitary but rather are hierarchical. Once we admit this into the analysis we need to separate out external corruption from internal corruption. External corruption occurs in the unofficial transaction between the applicant citizen and the bureaucrat implementing the transfer of the resource. Internal corruption is a collusion between bureaucrats within the organization which results in a systematized sharing of the proceeds from corruption. In a hierarchical set up the bureaucrat involved in external corruption is subject to monitoring and hierarchical review by a supervisor. The supervisor faces a cost of spending effort on monitoring and sometimes only, given the secrecy with which bribes are given and received, is the supervisor able to detect the bribe. In that case disciplinary measures are initiated against the erring bureaucrat and the supervisor may be rewarded with a bonus or some other measure of appreciation. When detected by the supervisor, the bureaucrat may in turn bribe the supervisor so as to avoid the disciplinary punishment which could range from a fine to a transfer, temporary suspension, dismissal, or even prison sentence.⁹ When the bureaucrat and his supervisor collude to hide information about the transaction of a bribe there occurs internal corruption. Knowing that monitoring is part of the institution’s functional set up the bureaucrat and his supervisor may enter into an agreement to share the proceeds from bribery ex ante at the application stage itself rather than ex post after detection in the monitoring stage.

It could also happen that a bureaucrat’s actions are put under the spotlight as a result of complaints made by those subject to extortion when they were legal applicants deserving of the resource transfer, or by those denied a transfer because the undeserving applicant received it by bribe giving. The complaint results in a review procedure that is distinct from the monitoring procedure. Complaints are usually registered higher up in the bureaucracy – the immediate supervisor, or the supervisor’s supervisor. Reviewers have to take into account that since some sort of rationing mechanism is used for allocation any person getting the allocation enjoys rents which a person denied the allocation (rightly or wrongly) misses thereby giving incentives for complaints. Just like monitoring, review may not be successful in identifying whether an infraction has occurred or not on the part of the bureaucrat. If the review does establish that bribes have been transferred disciplinary action and a penalty would be imposed on the bureaucrat and correspondingly the reviewer, which may be a committee, may receive a reward.

Unlike the monitoring stage, however, the reviewer has to have a greater incentive to collude with the bureaucrat because information has to be hidden not only within the organization about deceitful actions but it has also to be justified in some way outside the organization to the complainant. The ability to do so depends on the extent to which alternative reliable sources of information are available such as whistleblowers and the media. Supervision shifts the power of revelation and information disclosure higher up in the hierarchy of the organization and commensurate with that larger shares of the proceeds of bribery go into the hands of supervisors. Such internal corruption and collusion in the bureaucracy where subordinates transfer the proceeds form bribes to upper levels in the hierarchy in return for protection will result in growing corruption as each person in the hierarchy will act to maximize the proceeds of corruption that can be collected in the hierarchy. The lower the reward to a supervisor for truthful reporting and the larger the bribe and the penalty on the deviant bureaucrat, the higher we can expect the extent of collusion. Raising the penalty on corruption is often advocated to effectively impose a deterrent effect. At the same time as it increases the expected opportunity cost of corruption, it also simultaneously requires that the potential bribe be sufficiently large to offset this deterrent effect and this can induce even more indulgence in bribery by the bureaucracy because of the protection that collusion offers. Higher penalties when there is collusion paradoxically increases the willingness to pay for safely taking bribes. Harsh penalties for deterring corruption can therefore be counterproductive and may even intensify the extent of corruption when collusion is extensive.

We must be careful to acknowledge that bureaucrats are not homogeneous – there will be some amongst them who are predisposed towards honesty and some will possess the trait of dishonesty. Each person who is interested in the longer term information-conveying effects of his behaviour will also consider how behaviour affects reputation and future prospects. However, individual behaviour is imperfectly observed and it is also affected by the track record of the organization where a person serves. If the generally perceived notion of an organization is one of being a party to corrupt practices some of this behavioural trait is attributed even to honest individuals because behaviour cannot be perfectly observed. Beyond some threshold level of public perception of indulgence in corruption an individual’s reputation for honesty is always suspect (the organization’s reputation is attributed to the individual) and reduces the incentive for him to remain honest. The costs of being corrupt for an individual is not only the institutional penalty but also the cost of his reputation being damaged which in turn depends on the proportion of people in his organization who are corrupt. Every organization has a code of behaviour or a norm that is shared commonly and which informs the members of the organization whether they should be straight and stick to their duties. For those who do not adhere to the norm formal sanction by laid down procedures, peer pressure, and even ostracism may be experienced. The individual may even feel embarrassment, guilt, anxiety, and shame, which are standard means of sustaining norms. However, in an organization where corruption is prevalent these sanctioning processes and other costs are not as intensively experienced and reduce the incentive for honesty. If corruption and especially internal corruption is widespread in the organization large penalties have no impact as the chance of convicting a person is low given that recording this requires the active cooperation of the members of the organization. We could thus expect that a bureaucratic organization and its history of fair dealing will be an important determinant of the propensity for new entrants to engage in corruption activity. When reports from the Policy Group (1985) estimate that 76% of government tax auditors in India took bribes we should expect internal corruption to be large and new honest officers to have diminished incentives to remain so.


Executive-Legislative Collusion

In India the bureaucracy has elements of what a Weberian bureaucracy is deemed to have – meritocratic recruitment through competitive examinations, procedures for hiring and firing rather than political appointments and dismissals, and filling of higher levels of the hierarchy through internal promotion. The bureaucracy was designed to be apolitical and the institution through which governance of the nation was administered. As bureaucrats implement policies they have access to and control over information and can distort what they reveal to the political leadership to whom they are accountable. Politicians realizing this deficiency often attempt to compensate for this by soliciting views from independent experts and even from interest groups. In the case of corruption by the bureaucracy affected parties such as legal, deserving applicants, and those illegal applicants who consider the bribe demand to be higher than justifiable, as well as the fourth estate would often increase the information available to the politician about the misdeeds of the bureaucrat.

But just as bureaucrats collude and make a mockery of the monitoring and review process, politicians and bureaucrats also collude. And just as collusion at the review stage results in a large share of the bribe passing up the hierarchy, collusion between politicians and bureaucrats is usually a high margin and much less a volume business, and so less visible than the daily collections from a large number of people that the lower level functionaries engage in. Politicians in India who otherwise had little control over an independent bureaucracy were able to elicit patronage through exercising their power over postings. Since different points of contact with citizens – departments allocating licenses, development authorities for land, subsidized civil supplies, etc – command different illegal transfers that are known about within the bureaucracy because of internal collusion, there is also a latent demand within the bureaucracy for access to posts which enable the extraction of higher value bribes. Bureaucrats assess the illegal surplus that can be skimmed off in a post differently because of imperfect information or differential administrative capabilities and make offers to politicians as to the amount of graft they could transfer to them if appointed to that post. The various bids for a post made by members of the bureaucracy reveals the value of the post to the politician who then offers the post not necessarily to the highest bidder but to the one who would generate high transfers from graft along with secretiveness and tact, and at the same time meet some notional standard of service delivery. As there is a variability to the assessment of how much bribe money can be extracted and as the politician is not in a position to judge correctly whether he is receiving a fair share of the bribe, he has an incentive to make transfers frequent so as to keep receiving information from new bids by bureaucrats as to the value of the post.¹⁰ This in turn causes bureaucrats to be predative and they seek to extract as much bribes that they can from a post so as to acquire the resources to bid for, other lucrative posts or even to retain their present one as well as to garner within a short period of time what they would otherwise have done over more extensive time periods. A bureaucrat who owes his posting to a politician is also more pliable in favouring politicians with help in securing contracts, or getting admissions to hospitals or schools for his electorate, which enlarges the electoral support base of the politician. Bribe money can also be used to increase party funds that are then spent on activities that influence the electorate such as providing free medicine, clothes, etc. Less attention gets paid to allocating goods and services according to social value for which purpose the institution was designed, and more emphasis goes to how to raise the yield on private benefits associated with extracting bribes. The fact that promotions are decided on seniority and not job performance also gives little incentive to pay attention to the efficient provision of services from the organization.

It should be emphasized yet again that not all bureaucrats and politicians are corrupt Some are honest to the t and some are tempted once in a way to engage in corruption when the benefit from this outweighs the possible loss of reputation for honesty that gives them self-esteem. Those who are honest and possess technical skills that are required for the running of a department as well as demonstrate administrative capability do rise to higher levels in the system. These often occur in posts where either public contact is low or where systemic corruption has economy-wide ramifications such as is the case with regulatory bodies in the capital markets. But the point we are making is at the systemic level. When Transparency International ranks India as amongst the most corrupt countries with a rank of 71 amongst 102 countries covered in 2002, and the recently released Human Development Report 2004 of the UNDP lists Mumbai and New Delhi as amongst the top cities in the world where people are affected by corruption and bribery, our analysis should not be driven by exceptions and outliers. Empirical evidence has documented widespread corruption. It has been acknowledged as prevalent. Kohli (1990) quotes Rajiv Gandhi as stating on the 100th anniversary of the Congress Party that “Corruption is not only tolerated, but regarded as a hallmark of leadership”.


Mitigation Mechanisms

Our analysis points to the possibilities that arise for considering corruption mitigation. It is obvious that mechanisms that reduce the informational problem associated with government activities will tend to reduce corruption. Corruption arises in government as a consequence of rents – decisions by bureaucrats generate rents ex post to those benefited by the decision – and monitoring failures. The corruption levels are directly an outcome of the incentives faced by bureaucrats and citizens due to the slackness of monitoring and the existence of rents. The idea of preventing the abuse of power that arises with the checks and balances mechanism as a result of separation of powers, with different government bodies disciplining each other, so that citizens are better off has limited effectiveness in addressing the problem. When different branches of the government – executive and legislative – collude with each other, separation of powers is not effective in unearthing and deterring corruption. Even an independent judiciary would have its task cut out when the information it requires to decide on a case of corruption is not independently verifiable because of collusion within the bureaucracy and between the bureaucracy and politicians. Separation of powers is a cornerstone of a liberal society as it makes regulatory capture more difficult. Article 16 of the Rights of Man of 1789 states in fact that a society in which the guarantee of rights is not assured, nor the separation of powers provided for, has no constitution. Without this separation of powers institutions of governance are easier to subvert. In such a situation transparency and freedom of the press and expression become imperative so that misdeeds on the part of the government can be publicized and through the impact on electoral behaviour reduce corruption.

This raises the problem of balance at a different level as freedom of the press can result in the deliberate publication of defamatory falsehoods. Two important values are involved here – freedom of speech and freedom from defamation. Newspapers can monitor corruption and be auditors of politicians and bureaucrats. However, they could be deterred by defamation damages. Being harsh on defamation, however, leads to a lack of incentive to expose corruption which is a bad outcome for society. On the other hand, low compensation for defamation increases the incentive for the newspaper to print lies about honest public servants. May be using two tools could bring about the balance – the penalty on public servants for correctly revealed corruption should be set as high as possible and could even involve confiscation of assets disproportionate to known sources of income in addition to imprisonment. The compensation for defamation on the other hand should be lower than the damage to reputation of the politician and should be limited to forfeiting the increased revenues to the newspaper from increased circulation and covering the litigation costs of the politician. When information about corruption is hidden by collusion a society should think about increasing the freedom of the press even if as a result some politicians and bureaucrats are accused wrongly of dishonesty.

Apart from a free press, empowering the public through instruments such as the Freedom of Information promotes transparency and can reduce corruption. Other tools that are employed such as E-governance and voice mechanisms such as the report card system of the Public Affairs Centre in Bangalore have their pluses in enabling public scrutiny of public authorities. These mechanisms are also fraught with difficulties. Citizens cannot be trusted either to be fair when it comes to the public allocation of resources for transferring such resources provides rents to those who benefit from them and distorts their incentives to report truthfully about such transfers. As we argued earlier, bureaucratic allocation is the result of market failure and in a non-market setting consumer preferences are not aligned with social benefits. We should ask for administrative reforms to simplify laws and procedures,¹¹ citizens' charters¹² for handling grievances, and codes of conduct prescribing standards of integrity for the public services.¹³ We should also ask for an independent judiciary that takes on cases dealing with malfeasance by public officials.¹⁴ Higher levels of government intervention, a more closed economy, and lower public sector wages relative to private sector wages in the industrial sector are also often attributed as reasons for higher corruption in an economy and demands made for withdrawal of the state, more openness to trade, and a more highly paid and smaller bureaucracy. Whilst important we would argue that these are proximate causes of corruption. I would rather place democratic institutions as being the important determining factor behind corruption rather than policy variables.

The extent of openness of an economy or the wages of its bureaucracy would more be determined by the goals of a society as expressed through its sovereign democratic institutions. Those institutions function according to social goals if there is no collusion and effective separation of powers and checks and balances exist. When there is collusion we do ask for administrative reforms, and an independent judiciary but we should also strengthen the ability of civil society to monitor the public institutions through ensuring freedom of the press. Accountability is more easily fostered by the separation of powers. If individuals involved in governance are benevolent then they would always use their discretion and the information at their disposal to further the social good. It is when individuals are non-benevolent and pursue personal agendas by colluding that separation of powers is a means of limiting the scope for indulgence in socially wasteful activities. The mechanism of separation of powers does not work when agents are strategic and can manipulate evidence and affect unwarranted benefits to third parties as it is in their interests to cooperate at the expense of the institution. The idea of separation of powers is based on the principle that it is more likely that with checks there is mutual monitoring and policing. In India, however, this power to check misreporting and collusive behaviour is forfeited in favour of the use of the power, which is bestowed by the privileged status of being in the executive or legislative branch, for the pursuit of personal benefits that makes the system of accountability defunct. This requires a greater role for non-constitutional agents – the press, business associations, whistleblowers, academics, NGOs – who keep reporting deceit and fraud and put pressure to increase accountability in the system.

Errol D’Souza Professor, Economics Area Indian Institute of Management Ahmedabad 380 015.
Tel.: +91-79-2630-8357/58 (Board)
              -2632-4866 (Direct)
              -2632-4870 (Secretary)
Fax: +91-79-2630-6896
Email: errol@iimahd.ernet.in



BIBLIOGRAPHY

Acemoglu, D. & T. Verdier (2000) – “The Choice Between Market Failures and Corruption”, American Economic Review, 90(1), 194-211.
Alvarez, M., Cheibub, J.A., Limongi, F., Przeworski, A. (1996) – “Classifying political regimes”, Studies in Comparative International Development, 31(2), 3-36.
Banerjee, A. (1997) – “A Theory of Misgovernance”, Quarterly Journal of Economics, 112(4), November, 128-1332.
Basu, K., S. Bhattacharya, and A. Mishra (1992) -"Notes on Bribery and the Control of Corruption", Journal of Public Economics, 48, 349­ 59.
Goswami, O., I. N. Gang, and A. Sanyal (1991) -"Bureaucratic Corruption, Income Tax Enforcement, and Collection", mimeo.
Gupta, A. (1995) – “Blurred boundaries: the discourse of corruption, the culture of politics, and the imagined state”, American Ethnologist, 22(2), 375-402.
Guriev, S. (2004) – “Red tape and corruption”, Journal of Development Economics, 73, 489-504.
Kohli, A. (1990) – “Democracy and Discontent: India’s Growing Crisis of Governability”, Cambridge University Press.
Kumar, A. (1999) – “The Black Economy in India”, Penguin Books India, New Delhi.
Laffont, J.J &D. Martimort (1999) – “Separation of regulators against collusive behaviour”, RAND Journal of Economics, 30(2), 232-262.
Laffont, J – J and Tirole, J. (1991) – “The Politics of Government Decision-Making: A Theory of Regulatory Capture”, Quarterly Journal of Economics, November, 1089-1127.
Lederman, D., N. Loayza and R. Reis Soares (2001) – “Accountability and Corruption – Political Institutions Matter”, The World Bank, November, mimeo.
Lui, F.T. (1985) – “An equilibrium queuing model of bribery”, Journal of Political Economy, 93, 760-781.
Myrdal, G. (1968) – “Asian Drama, Vol. II”, Random House, New York.
Policy Group (1985) – “Estimates of tax collection and evasion in respect to the 1985-86 budget”, New Delhi, mimeo.
Potter, David C. (1986) – “India’s Political Administrators, 1919-1983”, Oxford: Clarendon Press.
Prendergast, C. (2003) – “The Limits of Bureaucratic Efficiency”, Journal of Political Economy, 111(5), 929-958.
Tirole, J. (1996) – “A Theory of Collective Reputations (with applications to the persistence of corruption and firm quality)”, Review of Economic Studies, 63, 1-22.
Treisman, D. (2000) – “The causes of corruption: a cross-national study”, Journal of Public Economics, 76, 399-457.
Verma, A. (1999) – “Cultural roots of police corruption in India”, Policing: An International Journal of Police Strategies & Management, 22(3), 264-279.
Wade, Robert (1985) – “The Market for Public Office: Why the Indian State is Not Better at Development”, World Development, April, 13:4, 467-97.

¹ Two hypotheses we do not report here are about the effect of ethnically divided societies and federally structured states on corruption.
² Common law systems found mainly in Britain and its former colonies developed from precedents established by judges usually allied with the property owning aristocracy against the Crown. Civil law systems found mainly in continental Europe and its former colonies developed from codes drawn up by jurists at the sovereign’s bidding for purposes of state building and controlling economic life.
³ A country is considered to be democratic if (a) the chief executive is elected, (b) the legislature (at least its lower house) is elected, (c) more than one party contests elections, and (d) during the last three elections of a chief executive there has been at least one turnover of power between parties. See Alvarez et al (1996).
⁴ Political instability was measured as the average number of leaders the country’s government had per year.
⁵ The highest estimate of the impact of 45 years of uninterrupted democracy was that it would reduce the corruption score by about 1.5 points on a 10-point scale.
⁶ Protestant cultures are more likely to discover and punish official abuses. Protestants believe that individuals are personally responsible for avoiding sin whereas the Catholic Church places more emphasis on the inherent weakness of human beings, their inability to escape sin and error, and the need for the church to be forgiving and protecting. Protestant countries are more individualistic and less familistic and a focus on the family rather than individuals can slip into nepotism.
⁷ The corruption scores were significantly lower for former British colonies once one controls for income. The four most corrupt countries on the list in 1996 – Bangladesh, Kenya, Pakistan and Nigeria – were all British colonies – and they were rated less corrupt than would have been expected given their poverty.
⁸ This work (which emanated from Stigler) is usually seen as originating from the Chicago School (Stigler, Peltzman, Becker) or the Virginia School (Tollison and Tullock).
⁹ Again there are possibilities here such as the supervisor extorting a transfer rather than the supervise offering a bribe which for simplicity we do not list. Supervisors may also report bribes wrongly so as to affect the prospects for making illegal transfers.
¹⁰ 80% of IAS officers held posts for less than two years, with a majority staying for less than one year according to Potter (1986).
¹¹ See the report of the Commission on Review of Administrative Laws set up in 1998.
¹² See the 1997-98 Annual Report of Ministry of Personnel, Public Grievances and Pensions, Government of India.
¹³ This has been endorsed by the Chief Ministers’ Conference and is being given legal shape.
¹⁴ In India unlike many European countries such as UK and France the officers and staff of the courts are under the control of the Chief Justices of the court and there is no interference from the executive or the legislature. However, the number of judges per million population varies from 90 to 100 in Europe whereas in India it is just 11. Many posts are vacant and judicial finances are a shortcoming.