Arvind K. Jain uses the concept of agency theory to analyse the effects of corruption of the decision makers on the level of foreign debt.
There is hardly an international forum or publication that has not discussed some aspect of the Third World debt crisis since 1982. Much of the discussion is devoted to understanding what caused the crisis and what can be done about it. There is shortage neither of diagnoses nor of prescriptions. Most of the suggested solutions, however, treat the crisis as one of liquidity and not as one of solvency. The main thrust of the prescriptions tends to be towards ‘managing’ the crisis in the hope that over time, the countries will regain their liquidity and be able to service their debts. Although crisis management of this kind is essential, it is also important that fundamental problems, if any, be addressed so that the crisis does not repeat itself.
The nature of the political structure within the borrowing countries may be one such fundamental problem. Borrowing decisions in a country are made by a group of individuals, that is, the political leaders and the bureaucrats who are supposed to act in the best interests of the public at large. It is, however, idealistic to assume that politicians and the public officials who make these decisions are pathologically honest and will completely set aside their personal interests when making resource allocation decisions. Access to power creates a temptation to misuse the power and politicians can be expected to take decisions that benefit either themselves or a close group of their supporters. Mature political systems, like those in some of the industrialised countries, have mechanisms to control such self-serving actions of the politicians. In many of the Third World countries, however, the population often has few mechanisms for controlling the leaders even when they have some degree of choice as to who will lead them. Lack of such control can allow politicians, especially if they are prone to corruption, to make decisions that are undesirable from the perspective of the population.
In this chapter, the concept of agency theory is used to analyse the effects of corruption of the decision makers on the level of foreign debt. The scope of agency relationships is extended to include situations where the relationship may be an ‘implicit’ one in that the agent acquires the decision-making power without an explicit consent of the principal. The analytics of agency theory are extended in this paper to derive the effects of principal’s lack of control over the agent.1
We show that over-borrowing can take place when the agents are corrupt and the principal lacks control over the agent. Creditors who should normally be expected to discourage over-borrowing by agents may fail to correct for the bias because of certain imperfections in the international markets. A solution of the crisis would, thus, require that changes be made in the political structure within the borrowing countries, or at least that the lenders take cognizance of the biases created by the self-interest of the borrowing agents.
It is important that this ‘cause’ of the debt crisis be seen as only one of the factors that triggered the debt crisis. It is widely recognised that inflationary shocks in the world economy increased the interest burden of the borrowing countries just when their export revenues were declining due to a recession in most of the industrialised world and declining commodity prices. Internal mismanagement in the form of over-valued exchange rates and inability to adjust the domestic economy in response to global shocks exacerbated the effects of the external shocks. It is our contention that the borrowing countries could have faced these external and internal shocks more easily, and the crisis would have been less severe, had the borrowed funds been used for more productive investments within their economies.
We develop the relationship between the political structure and the debt crisis in three stages to demonstrate the economic consequences of political corruption. The first section below presents some evidence on the use of borrowed funds as well as the consequences of the debt crisis in the borrowing countries to highlight the division that exists in these countries between those who govern and those who are governed. The second section describes the political system in a country as an agency relationship between the population and the leadership and explains the problems that can arise in such a relationship. How the borrowing decision is made by the agent within such a relationship is explained in the third section. The last section of this essay offers some suggestions for solving the crisis and the dilemmas it poses for the lending countries.
Consequences of the crisis since 1982
The onslaught of the debt crisis in 1982 has renewed some interest in investigations of the wealth of some of the leaders of countries facing the debt crisis. One also hears about the burden of adjustments due to debt repayments being born by populations in these countries. Most writers, however, fail to link the two phenomenon. We present anecdotal evidence on these two aspects of the foreign debt in this section and provide a conceptual link between the two in the subsequent sections.
Perhaps the worst example of corruption is provided by Zaire’s President Mobutu who is said to have a personal wealth that approaches the entire external debt of the country. The ex-president of the Philippines is assumed to have taken $2.5 to $3.0 billion out of the country. Haiti’s ousted leader, Jean Claude Duvalier, is known to have amassed a fortune of between $100 million and $800 million. In the Dominican Republic, a former president was formally charged with illicitly acquiring wealth while in office and permitting close associates to amass personal fortunes at the expense of the national treasury. The evidence from other Latin American countries indicates a pattern of widespread corruption among a large number of public officials. In most of these countries, funds earned through corruption were saved outside the country, resulting in capital flight.2
Table 4.1 provides some estimates of the extent of capital flight from the major borrowing countries in the period 1976-82. In spite of the problems of defining ‘capital flight,’ the magnitudes in Table 4.1 represent the extent to which domestic investors preferred to export their capital, partly in response to the perceived ‘safety’ of the capital. Much of the capital thus exported was considered unsafe within the country for political reasons although market distortions can be blamed for some of this outflow.
Large variations in the estimates for each country’s capital flight do not distract from the fact that for some of these countries, for example, Argentina, Mexico, Philippines and Venezuela, capital flight – in key years – represented a large proportion of the external debt. Significant proportions of external debt were thus never available for investment in the domestic economy. However, there is evidence more recently that capital outflows – or capital flight – have been reversed in a number of countries, such as Mexico.
Like the evidence for corruption, the evidence for the decline in the population’s income is scattered and only partially documented. On a global scale, even leading bankers recognise that the population in the debtor countries has suffered income losses since the debt crisis. According to one such estimate, the per capita consumption of an average citizen in the heavily indebted countries has been declining at an average rate of 1.6 per cent a year since 1980. One estimate of twenty countries in Latin America and the Caribbean shows that per capita income between 1980 and 1987 has declined in seventeen countries and increased in only three.3 It is on the basis of such evidence that IFIs, such as the UNDP, and bodies like UNICEF, have argued for a more human form of response to the debt crisis, calling for adjustment ‘with a human face.’ The country-wide figures on per capita income, however, hide even larger declines in the incomes of some segments of these societies.
Two of the most powerful groups in debtor states, the ruling elites and wage earners, have been affected in completely opposite manners by the external debt of these countries. The ruling elites gained from the inflow of funds and the wage earners have seen their income decrease so that the debt can be serviced. To understand how this could have been possible, we turn to the dynamics of the relationship between these two groups with one acting as an agent and the other as the principal.
The political system as a principal-agent relationship
A principal-agent relationship arises ‘whenever one individual depends on the actions of another’4 and can be defined ‘as a contract under which one or more persons (the principals) engage another person (the agent) to perform some service on their behalf which involves delegating some decision-making authority to the agent.’5
While an agency relationship makes for an easier decision-making process, it is not free of problems. The main problem in such a relationship arises due to the agent’s tendency to maximise its own interests rather than those of the principal. The agent may wish to divert more resources for its own use than contracted for, and may not wish to put in all the necessary effort required to manage the principal’s resources. Solutions of these problems are complicated by information asymmetry: although the agent has full knowledge of its actions, the principal is not always in a position to observe, and hence control, the actions of the agent. These problems can be best explained in the context of a private corporation.
In a firm owned by one individual who is also the manager, the owner/manager will spend effort to maximise the value of the firm. Since the manager receives all the profits of the firm as the owner, there is no need for anyone to monitor or motivate the manager. An agency conflict, however, begins to arise if the owner/manager sells part of his equity to outsiders and all the shareholders together appoint a manager as their agent. The conflict arises from ‘the manager’s tendency to appropriate perquisites out of the firm’s resources for his own consumption . . . [and] . . . from the fact that as the manager’s ownership claim falls, his incentive to devote significant effort to creative activities . . . falls.’6 In the simple case when the principal is not allowed monitoring or bonding activities, sale of equity to non-managers results in welfare losses.
The incentive and perquisite problem is resolved in private markets through two specific mechanisms and through functioning of free markets, all of which presume the existence of a legal framework in which the contracts can be enforced. The two mechanisms consist of monitoring and bonding activities ‘(I)t is usually possible by expanding resources to alter the opportunity the owner-manager has for capturing non-pecuniary benefits. These methods include auditing, formal control systems, budget restrictions, and the establishment of incentive compensation systems which serve to more closely identify the manager’s interests with those of the outside equity holders.’7 In addition, Fama has emphasised the importance of ‘managerial labour markets: in providing incentives to the managers to provide the firm with their best efforts.’8 Perhaps the best testimonial for the effectiveness of the monitoring and bonding activities, and of the functioning of the managerial labour markets, is the fact that, as Jensen and Meckling argue, ‘the corporation has thus far survived in market tests against potential alternatives.’
The political system in a country can also be viewed as a principal-agent relationship. Political leaders act as agents of the population to manage the country’s resources, such that the welfare of the resource owners, that is the population, is maximised. It can be assumed that the principal will replace the agents if the latter’s efforts are viewed as unsatisfactory. The delegation of the resources management responsibilities by the principal to the agents facilitates the decision-making because it would be impossible for every member of the population to take part in every decision. Since all members of the population are unlikely to agree even on minor decisions, rules are developed for choosing agents, the most common of which is election by a simple majority.
From the point of view of the agent as well as the principal, the problems in this agency relationship may be more serious than in the case of a private corporation. Unlike the agents in a private corporation, political leaders as agents face the problem of a multiplicity of goals, with various members of the population placing dissimilar weight on different goals. Agents have to learn to balance the conflicting goals of various members of the population over time.
The principal’s problems of providing appropriate incentives to the agent and of monitoring and controlling the agent’s behaviour are also more difficult than in the case of a private corporation. Difficulties arise partly due to information asymmetry in that the principal is not always able to observe the agent’s behaviour. More serious problems, however, may arise because this relationship sometimes may be an ‘implicit’ one in that the agent may have acquired its position either through an established mechanism in which large sections of the principal have no voice in the choice of the agent, or in some cases even by being able to ignore the established mechanism altogether. The study of such ‘implicit’ agency relationships is important because such a relationship may represent the reality more accurately for many developing countries than the ‘explicit’ agency relationship in which the principal chooses its agent. The analytics of a relationship in which the principal may not have a legal control over the actions of the agent may be quite different from those of an agency relationship in which the principal-agent interaction is governed largely by information aymmetry.
Although we are witnessing a revival of democratic politics in a number of countries across the Third World, the effectiveness and sustainability of such democratic revivals can be questioned. Moreover, few developing countries can be said to have political systems in which large sections of the principal are able to exercise genuine control over the behaviour of their agents.9 Many political leaders acquire power with the support of the elite which represents only a small segment of the population. The mechanism for the selection of the leader in these cases may not allow for participation of all segments of the society. In other cases, especially military dictatorships, the leaders may have acquired power by ignoring or suspending the established mechanism for the selection of the political leadership. Whatever their route to power, leaders who acquire power without the support of large segments of the population do not feel any need to represent the interests of the excluded segments.
The most serious problem in such a relationship is that the agent will have an incentive as well as opportunities to allocate resources either for its own benefit, or for the benefit of its supporters, who are only a small segment of the population. For ease of exposition, we use the term ‘population’ or ‘principal’ to designate the large majority which does not directly support the leader.
An agent with control of all government resources can appropriate the principal’s wealth either directly by way of corruption or indirectly by reallocating resources to favour its own supporters.10 The agent takes actions which favour its supporters because, presumably, it wants to be re-elected.11 The excluded segment of the public or the principal, however, must rely on indirect mechanisms that exist in a political system for exercising influences over the agent. It can exert influence over the leader either by providing information through family, friendship or patronage ties. The effectiveness of this mechanism depends upon elite-mass congruence. The public can also try to control the behaviour of the leaders through legal actions. Threat of legal actions, when effective, will make corruption more difficult. The effectiveness of this mechanism depends upon the independence and efficiency of the judicial system in the country. The public can also threaten to remove the leaders from their position through physically violent means. Such a threat, when credible, may force the leaders to act in the interests of the public. Finally, the public may control some resources, denial of which may force the leaders to change their behaviour. Farmers, for example, may control production of a crop required to fulfil government plans, thus giving them some means to influence the agent’s behaviour.
Politicians, on the other hand, can reduce the effectiveness of the public’s control because they, as agents, control all the national resources. Even in advanced democratic countries, ‘growing political participation by the public has been offset by growing (or more effective) political manipulation of the public by the ruling elites.’12 The leaders, given their power over the election apparatus, can either eliminate elections or make them ineffective as means of true expression of the public’s opinion. The leaders may, through legislative and police/military power, suspend the fundamental rights of the public. Such actions reduce the public’s ability to take concerted action against the leadership by making it difficult to collect and disseminate information about the latter’s actions, and to interact with other segments of the public. The leaders may thus use the government apparatus to prevent the public from bringing about a change, violent or otherwise, of the leadership. As part of coercion, they may acquire control of news and information media which will reduce the effectiveness of public’s control mechanisms. The leaders may also use the government resources to reduce the independence of the judicial system which will reduce the public’s ability to take legal action. In addition, the leaders, having a monopoly over the administrative apparatus of the government, may threaten the public with denial of access to resources available through the administration. The leaders are likely to use the inherent powers in the control of the administration to influence the public’s behaviour. Finally, the leaders may attempt to influence the public through information, patronage, or notions of authority embedded in the culture of the country.
The leaders, having access to these mechanisms, are able to reduce the impact of public’s control mechanisms. The balance of these two sets of controls, one by the public on the leader and the other by the leaders on the public, will determine the extent to which the agents in any country will feel obliged to follow the wishes of the public. Among the various borrowing countries, it is possible to find countries that lie at each end of the spectrum which goes from a complete dictatorship to an effective democracy. An ‘effective democracy’ may be defined, for the purpose of this analysis, as one in which all segments of the public have sufficient and effective control mechanisms, so that if the leaders do not act in the best interests of the public, they can be removed from power.
Setting aside various shades of non-democratic systems, we define a ‘complete dictatorship’ as a system in which the public has no control over the leaders, as the other end of the spectrum. The leaders in this system can ignore the wishes of the public because they have sufficient control over the administrative apparatus of the government not to have to be concerned about any actions that the public might take.13
The foreign debt decisions of an agent
To understand how leaders in many developing countries could have borrowed amounts which they are unable to repay now, consider the following simplified portrayal of the foreign borrowing decision under the two extreme political systems identified in the previous section. The process described here views the level of external debt as a decision made by the agents, that is, the political leaders. These leaders are assumed to have preference for decisions that would maximise their own welfare. However, the public may be able to limit the independence of action of governments. The creditors may also try and reduce their discretion or their sovereign power. The effectiveness of the constraints, of course, depends upon the political system under which decisions are being made. The level of debt that the agents choose, therefore, becomes a function of the political system in which they operate.
The decision-making process described here differs from traditional analysis of the behaviour of decision makers in two respects. First, we allow for the existence of some conflict of interest between the political leaders and the population within a borrowing country. Traditional analyses treat borrowing countries as monolithic entities in which a unanimity of goals among all the different groups is assumed to exist. In the model presented below, we assume unanimity only within each interest group. Second, we allow the decision makers to be corrupt. Traditional economic analysis treats all participants as honest, responding passively to economic incentives and the policies they face. In the discussion that follows, the term agent always refers to the political leadership of a country and the term principal to the population. For ease of discussion, we treat all the political leaders as united and willing to work together as one agent.
The agent negotiates loans with foreign creditors on behalf of the principal. The ostensible use of the funds is to supplement domestic savings to increase the level of domestic investment.14 The output from the investment of domestic savings and borrowed funds (that is, external savings), after the repayment of the loans, belongs to the principal. If the decision on the level of foreign debt had to be made by the principal, its optimal point would be where the marginal rate of return on investment equals zero. Borrowing beyond this point would reduce the principal’s net income (that is, output less debt repayments) below that at the optimal point. The principal would require that, ideally, its agent negotiates external debt at the same level that the principal itself would have negotiated.
The agent, however, is interested in maximising the share of the principal’s wealth that it is able to appropriate. The agent’s ability to appropriate wealth arises from two sources. One of these appropriations arises due to information asymmetry and the other due to lack of control
First, the agent has an ability to create, and appropriate, a ‘slack’ in the use of principal’s resources. This slack results from the principal’s inability to observe all the actions of the agent. Due to unobservability, the principal is also unable to prevent the agent from appropriating the slack.
Second, the agent may openly appropriate the principal’s wealth. The principal can observe this action of the agent but can prevent the appropriation only if the principal’s control mechanisms over the agent are effective. In what we have labelled as democratic societies, such an appropriation will tend to be avoided whereas under the dictatorship, the principal will not be able to control the agent at all.
How does an agent that has these two possibilities to appropriate the principal’s wealth decide on the level of external debt?
Assume that the agent’s appropriations, whether directly or through the slack, are fixed proportions of the principal’s total initial wealth, which is the sum of domestic savings and externally borrowed funds. The agent’s self interest is, therefore, served by increasing the principal’s initial wealth by borrowing as large an amount as possible. There are, however, two constraints on the amount that the agent can borrow. First, the principal will use all its influence and controls over the agent to force the agent to borrow at a level which the principal would consider optimal for itself. Second, the creditors would lend funds only when the probability of these funds being repaid is sufficiently high.
These funds will be paid out of the principal’s income and therefore it is necessary for the agent to invest at least a part of the initial wealth on behalf of the principal. This constraint ensures that the agent will not appropriate the entire initial wealth of the principal. The creditors will lend funds only up to the level where the income of the principal exceeds or at least equals the required debt repayments.
Consider the effects of the constraint of the principal and that of the creditors separately. Whether or not the principal’s constraint will bind the agent’s actions depends upon the nature of the political system. The principal is assumed to have full control over the agent in a ‘functioning democracy.’ In such a system, the principal can force the agent to borrow funds abroad up to the point at which the marginal output received by the principal just equals the required debt repayments. The rate of return on all domestic investments, whether financed by domestic savings or by external funds, will be positive and the principal’s income would be higher than in the absence of foreign borrowings. The agent’s income consists only of appropriation of the ‘slack’ which is a fixed portion of the initial wealth. The creditors have no hesitation in lending the amounts required because the principal’s income is far above the amounts required to pay off the debts. Whether or not the creditors would have liked to lend more is not relevant to our analyses.
To continue reading “Dictatorship, Democracies and the Debt Crisis” by Arvind K. Jain, please see: https://journal.probeinternational.org/1993/06/01/dictatorship-democracies-and-the-debt-crisis-part-ii/
Arvind K. Jain, The Politics of Global Debt (ed.) Stephen P. Riley, St. Martin’s Press, June 1, 1993