CORRUPTION

Introduction
The most common definition of corruption is "the abuse of public power for private gain." This
definition can be disingenuously general, depending on how broadly one construes "public
power" and "private gain." By comparison, Shleifer and Vishny's [1993] reference to the "sale of
public assets for private gain'' is more restrictive, since it unduly limits the transactions to those
mediated by money corruption is "the abuse of public power for private gain. n exchange. Rose-
Ackerman [1998] gives a definition that is suited mostly to bribes: "an illegal payment to a
public agent to obtain a benefit that mayor may not be deserved in the absence of payoffs.n
Hutchcroft [1999:227] endorses a definition by Nye [1997] that is more explicit, although less
succinct: "behaviour which deviates from the formal duties of a public role because of private-
regarding (personal, close, family, private clique), pecuniary or status gains; or violates rules
against the exercise of certain types of private-regarding influence".
As is evident from the above examples, the typical definition of corruption involves the notion of
the "public" in a fundamental sense. For this reason, it is customary to regard the main locus of
corruption as government and as invariably involving public officials. Recent usage, however,
also speaks of "private corruption" as occurring in business firms or as it deals with organised
crime. Some examples are overpricing practised by supply managers, excessive charging of
personal bills at company expense and so on. A case for including these acts under corruption
can be made to the extent that they occur in the context of firms having a public dimension, e.g.,
that are widely held publicly, or to the extent that certain rules protecting the general public
proscribe such behaviour, e.g., trading on insider information. Otherwise, these acts would be in
the nature of private damages and the subject of civil liabilities. By extending the concept of
principal and agent defined below, however, a good deal of breach of trust in the private sector
may also be subsumed under corruption. For this paper, however, corruption shall refer only to
cases involving acts of public officials or pertaining to public assets.

Corruption as a breakdown in a principal-agent relation

Public corruption is not a thing but a relation; more precisely, it is the breakdown or rupture
(cor+ rumpere =completely breach) of a presumed relation between social agents. The
fundamental relationship affected in corruption partakes of the nature of the well-known
principal-agent (henceforth, the PA) problem in economics which refers to a situation where one
party (the "agent") is contracted to promote an outcome in behalf of someone else, namely the
"principal" (see, e.g., Arrow [1986]). The agent's action or characteristic, however, potentially
affects not only the principal's but also his own interest. When the agent's action or characteristic
cannot be directly observed by the principal, or where the outcome is affected not only by such
an action or characteristic 1 , there is a problem of ensuring that the agent takes the appropriate
action (or possesses the characteristic) that will promote the principal's interest. In the case of
corruption, the principal is understood to be the government, which embodies public interest in
its laws, policies and strategies. Government performs functions and distributes resources to
attain its goals. The performance of such functions, however, is devolved upon real people,
bureaucrats and politicians (or even private individuals as such concessionaires) who may be
presumed to act in their own interests. A contract implicitly exists between government and its
agents for the latter to discharge their functions in the former's behalf. Since the actions and
qualities of such agents can be observed only imperfectly, however, government cannot be sure
that its agents always perform fully in its interest. Depending on the context, the principal in the
corruption relation may be either the highest echelons of government, with the agents referring to
officials at different levels designated to carry out certain tasks, or the principal can be the
electorate or public at large, with politicians as their agents. It is obvious that the PA problem
exists on several levels, with society at large (the public or the electorate) being the ultimate
principal, and both politicians and bureaucrats existing as agents on the lower levels.
More broadly construed, however, the rules governing the PA relationship between politicians (or
government) and bureaucrats (or the civil service) are found The events of EDSA 2, for example,
must be understood primarily as an informal recallmechanism at work in the entire system
ofbureaucratic recruitment, compensation, promotion and procurement, all of which are designed
by politicians. Included in these formal constraints are the larger laws governing the behaviour of
public officials and the penal system punishing errant behaviour. Finally, one cannot discount the
various informal constraints that internally regulate bureaucratic behaviour such as
professionalism, a moral viewpoint, a sense of duty and esprit de corps, distinct factors that to
some extent transcend pecuniary compensation.
Similarly, the relationship between the public (as embodied by the electorate) and politicians is
governed by the political system as typically set out in a constitution which includes the formal
mechanisms for election and recall. The 1987 Constitution, for example, provides for the conduct
of regular elections and occasional plebiscites, but more innovatively, also for electoral recall
and constitutional amendment under a people's initiative.2 The independence of the government's
branches and the checks and balances among them (e.g., impeachment), the dichotomy between
government and opposition, and laws and the penal system as they relate to corruption and
plunder form part of this formal system. Compared with the bureaucracv, however, the difference
between formal and informal mechanisms is less distinct at the political level. Constitutional
guarantees of the freedom of speech, freedom of the press, freedom of assembly and ultimately,
the right to rebellion do not directly define formal mechanisms to reward or punish erring
politicians, merely leaving open the door for more informal mechanisms to function, including
the threat of exposure to public or media criticism, social ostracism and collective action. The
events of EDSA 2, for example, must be understood primarily as an informal recall-mechanism
at work: large-scale collective action in the wake of a failed formal mechanism (the
impeachment process) turned public opinion, forcing a mass resignation of the cabinet and the
effective resignation of the president. In many ways, the control mechanisms shaping the
relationship between politicians and the electorate at large are among the most weakly defined.
Combined with the inherently large discretion accorded to politicians and the peculiarities of
Philippine political economy, this provides fertile grounds for corruption. Understanding
corruption by proceeding from the principal-agent problem is neither to redefine nor to
decriminalize it, in much the same way that understanding why crime exists is not tantamount to
condoning it. The approach itself does not change the thing viewed but puts its basic elements
and characteristics into sharper focus.

Factors affecting the principal-agent relation

It must be said at the outset, of course, that this paper's use of the PA framework goes beyond its
restricted economic application where it is used primarily to solve for the compensation scheme
that aligns the agent's actions to the principal's objectives. Instead, it will be used here primarily
as a heuristic. By initially positing the requirements for an ideal relationship that ought to exist
between principal and agent and then comparing it with the corrupted or "broken" one that
actually exists, one is able to determine the sources of difficulties that prevent such a relationship
from functioning. 3 What follows describes and analyses some of those factors.

Complex environment and the development context

A complex environment is ubiquitous in any PA problem. If all contingencies could only be
foreseen, then a perfect (iflengthy) contract could be written, specifying exactly what the agent
should do if such and such an event occurred. But contracts cannot be specified too tightly, since
a complex or changing environment may require a different set of actions from what has been
stipulated. A greater or lesser degree of discretion must therefore be typically allowed the agent
or person in charge. This very discretion, however, is what allows the agent to behave in ways
other than those that promote the interests of the principal. The difficulties posed by a complex
environment maybe seen in persistent attempts to prescribe behaviour, as in, say, the voluminous
rules for procurement laid down by the Commission on Audit. These may work well enough to
regulate behaviour in predictable, repetitive contexts (e.g., activities connected with processing
of passports or drivers' licenses). But as one moves to more complex environments, especially as
these pertain to developing countries, it becomes clear that the constant expansion and
elaboration of such rules to prevent abuse cannot go on indefinitely. To do so would seriously
impair the initiative of agents, reducing them to simple robots and becoming barriers to
efficiency itself. In general, therefore, a trade-off exists between the discretion needed by agents
to perform other than merely perfunctorily in a complex world and the specification of behaviour
needed to prevent abuse. The constitutional concern over the possibility of" grave abuse of
discretion" summarises the dilemma.

Information and bounded rationality on the part of principals

Even in a complex environment, the PA problem would be more tractable if the principal was
always able to determine exactly what has occurred and how the agent has acted. There is a limit,
however, to the extent that principals will find it worthwhile to observe the actions of their
agents; agents will typically be better informed of their own actions than the principals are.
Rationality is ultimately bounded or limited because its uses are multiple and competing. The
electorate, for example, does not always invest in the information required to make the best
electoral choices (nor even take the time to vote) and it cannot continuously inform itself of all
that takes place in government. Mter all, people work and consume mostly in a private capacity.
In this sense, exhortations to nationalism and civic duty really represent attempts to persuade the
public to make the investment in political information they would otherwise not make. As will be
argued in the last part, the democratic election of undesirables like Estrada is due in large part to
information failure. Taken together, bounded rationality and a complex environment give rise to
monitoring costs, which simply refer to valuable resources used up whenever principals try to
ascertain the behaviour of agents closely. It is, of course, possible to designate special entities to
monitor the behaviour of agents as part of the arrangement; this is the purpose of entities such as
the COA, the ombudsman (tanodbayan), the anti-graft court (sandigang bayan) and special anti-
corruption commissions. It is equally obvious, however, that such arrangements are costly and
expend scarce resources.

Personalities versus parties-non-shared goals between principal and agent

If actions could only be prescribed beforehand and verified, then the agent's subjective beliefs
and values would become immaterial and irrelevant. As more discretion and authority are
devolved upon an agent, however, (necessitated, say, by an increasingly complex environment),
it becomes more desirable that the agent possess values that reflect those of the principal more
closely. Arrow [1974:23] writes, Trust is an important lubricant of a social system. It is extremely
efficient; it saves a lot of trouble to have a fair degree of reliance on other people's word.
Unfortunately, this is not a commodity which can be bought very easily. Character and values are
more important among politicians than run-of-themill bureaucrats since the former typically
exercise more discretion and scope for judgment than the latter, whose behaviour is usually more
readily prescribed. Much has been written in this country regarding the need to base politics on
issues rather than personalities; it is probably more accurate to say that stances on Character and
values are more important among politicians than run-of-the-mill bureaucrats issues are as
important to the extent that they also reveal character and values. The more complex the
environment, the less easy it is to prescribe definitive issues. To what extent can "character" and
"values" be predicted? The advantage of political parties built on ideologies is that they expressly
contain a well-defined process for sorting out candidates based on the degree to which these
share the party's goals. Political parties place their seal of approval on candidates who have
undergone a process vetted by informed party observers and insiders. This reduces the
information costs on the part of the individual voter who no longer needs to familiarise himself
with the individual candidate but takes the word of the party. An offshoot of mass media and
entertainment, however, is to convey the misimpression that the 8 PUBLIC POLICY Corruption
in the Philippines: Framework and Context electorate can actually get to know candidates'
character and values directly and intimately {witness the Estrada myth of being the defender of
the poor, conveyed through cinema, or the reputation of media personalities developed through
investigative and public assistance programmes on television). This direct appeal to voters and
the false intimacy generated by mass media is an element of modern political culture that has
become particularly conducive to political demagoguery and irresponsibility since first, it can
obviously be false and second, unlike a party system, it provides for no intermediate mechanism
to hold politicians accountable to shared values during periods between elections.

Scope of action and competition

An important condition for the existence of corruption is government's position as a virtual
monopolist in the provision of many goods and services, raising the potential for obtaining rents
by influencing the decisions of government agents. If government were no more than one among
many competitors rather than a monopolist, no premium would attach to the decisions it makes,
and little or no resources devoted to influencing the exercise of its prerogatives. A related point is
that a smaller scale of government would reduce the scope for corruption. This has been
extrapolated into the proposition- discussed under political economy-context below-that a
reduction in the size of government is an indispensable ingredient in the reduction of corruption.
But one must balance the real requirements of development in which a government's role is
indispensable against this. In the same manner, removing all discretion would drastically reduce
opportunities for corruption but would more likely impede efficient responses to complex
environments. Rather than directly reduce public agents' scope of action through closer
prescription, it is also possible to use competition as a means of eliciting appropriate agents'
[Link] can take various forms ranging from combined public and private provision,
contracting and concessions and management contracts to complete privatisation (examples
being water concessions, power sector reforms and various others).

Taxonomy
The explosion of literature on corruption has introduced a plethora of approaches, concepts,
typologies and analyses, as well as corresponding policy recommendations. The following is, to
our knowledge, the first attempt to bring together systematically the largest number of these
approaches.
By type of agent and initiator
Corruption is defined by the location in which it occurs in the public sector, which in turn, is
largely decided by the degree of discretion exercised by public officials. Public-sector corruption
may involve either bureaucrats or politicians. Bureaucrats and administrators are thought to be
interested primarily in pecuniary gain, while politicians are thought to have both pecuniary and
political interests at stake. In the Philippines, earlier literature (typified by Carino et al. [ 1978])
focused on bureaucratic corruption, viewed primarily as a problem of ethos and example. That
literature implicitly adopted the technocratic viewpoint that the political leadership would take an
active interest in reducing or eliminating corruption in the regular bureaucracy as part of
improving the effectiveness of governance. More critical analyses of the quality of
politicalleaderships themselves, however (especially after the fall of dictatorships and exposure
of their excesses), led to an interest in systematic corruption among politicians themselves.
Political corruption is distinct from bureaucratic corruption, partly because of the levels at which
it occurs, and partly because of differences in goals. Hence, for instance, fewer bureaucrats
would be interested in expanding influence and patronage beyond what maximising pecuniary
advantages would dictate. Vote-buying, corruption of the electoral system, political or regulatory
harrassment of opponents are associated with distinctly "political" rather than bureaucratic goals.
A conflation of relationships is also possible when bureaucrats collude with politicians.
Economic impact of corruption
 One of the most contentious issues is the exact economic impact of corruption. The agnostic
view of economists regarding corruption is somewhat startling for many who tend to take a more
ethical view of it. Barro and Sala-i-Martin [ 1995:440, fn. 14] conclude, for example, that "the
theoretical effect of corruption is unclear; in some cases, the economy would operate more
efficiently if govern· mental rules can be readily overcome by cash payments". Even much
earlier, of course, Adam Smith gave the benefit of the doubt to the smuggler and presumably,
would have excused the corrupt customs official. A good deal of literature deals with the
microeconomic analyses of various types of corrupt transaction according to one or more of the
categories cited above. Hence, one may compare petty versus grand or predictable versus
unpredictable corruption. Much of what follows is based on de Dios [ 1999] and enumerates the
various types of micro-analytic arguments seeking to show how one or another type of
corruption has a positive or negative impact on efficiency. It seems fair to say, however, that a
microeconomic analysis alone is usually insufficient to make an airtight conclusion regarding
efficiency, without reference to an analysis of the larger institutional framework, including the
rationale and optimality of the existing laws themselves.

Economic strategy and the scope of government

A central argument is that less government intervention in the economy and a greater reliance on
markets instead is a key ingredient in reducing corruption [World Bank 1997a], a contention that
must be measured against the idea of an activist "developmental state" in the East Asian mode
[Chang 1996]. How countries of the latter sort (e.g., South Korea or Japan) managed to temper
the state interventions that mode called for with the need to restrain bureaucratic and political
self-seeking is an indispensable part of a growth puzzle. It has also often been noted [Chang
forthcoming citing Okimoto 1989 and also Fabella] that corruption is typically less in tradables
sectors where poor performance is more easily exposed than in the nontradables sector. Part of
the reason is that the standards of performance are more easily determined in tradables, where
world prices serve as non manipulable benchmarks and budget constraints are hard. A second
reason is that "locational competition" (Standortswettbewerb) to attract transnational firms and
mobile capital may exert pressure on developing countries to enforce stricter rules against
corruption which discourage such investments. But it has also been maintained that the rush
towards the market, as witness the wave of liberalisations and privatisations, have provided
opportunities for new and largescale forms of corruption (in the Philippines, the BW stock-
market scam) as assets are revalued and large players such as transnational corporations enter the
field. Nonetheless, these moves in favour of privatisation are in the nature of one-off
opportunities for corruption rents and have the advantage that they do not lead to the recurring
economic distortions associated with continuing government economic involvement. So it would
seem that if one would bring down the level of corruption, it is necessary, perhaps sufficient, to
make the state less interventionist. To be sure, the number of objects of corrupt transactions
would be reduced in a minimalist state, apparently bringing the level of corruption down with it.
While it has its merits, however, such an argument is at best incomplete. While less intervention
would decrease the number of transactions that may be corrupted, less intervention may only
result in the increase in proportion of such transactions being corrupted, or a change from being
petty to grand of each incidence of corruption if its root causes remain. Secondly, the reasons for
government intervention in the economy as a rule proceed from the choice of developmental
strategy so that the abandonment of such a strategy because of corruption would occur if
corruption totally negates the benefits from such a strategy. A paradox might in fact arise here.
Cross-country studies have shown, conclusively we believe, that less corruption correlates with
more development. The development of latecomer countries, on the other hand, involved
increased intervention by the state (e.g., Chang and Rowthorn [ 1994]), which was surely
accompanied by an increase in the scope of corruptible transactions. This dilemma can only be
tackled by gaining a deeper understanding of the problems of corruption and economic
development.
Conclusions and recommendations
While corruption has always existed in the Philippines, it has changed in form and magnitude
through time as economic strategies and political systems have also changed. Its significance is
determined primarily by inadequacies in the nation's political life, especially those that weaken
·the bonds of control and monitoring between the principals (society at large) and those who
serve them (politicians and bureaucrats). It must be understood primarily as apolitical
phenomenon; bureaucratic corruption is often merely secondary and instrumental but is of
subordinate importance because (a) it pales in comparison with the grand corruption involved in
corruption among politicians, and (b) it is often abetted and controlled by politicians (e.g.,
through appointments and undue influence on bureaucrats). Thus, even as one may argue for
better design in the incentive mechanisms for politicians and bureaucrats, one must return to
more fundamental questions about why the political relationship is deeply flawed and fails to
serve as an effective check to the behaviour of officials.
A fundamental reason must be the continuing perception among many Filipinos that the
relationship between the government and themselves is at best an abstract one. Many years ago,
Corpuz traced the historical roots of graft and corruption to the "negative image" of government
among Filipinos. Especially during the Spanish period, positions were awarded or sold as a
matter of course to undeserving and often abusive individuals, leading to a government that was
unresponsive to the needs of the inhabitahts:
The institutions of government were agents of abuse and oppression. At its best, government
came to mean for the Filipino an institution that was burdensome; at its worst, it was
predatory .... To the Filipino, government became an institution to be avoided, for its interests
were contradictory to his. People and government were estranged from each other and the bonds
of community were dissolved .... [Corpuz 1965:78]

References
Bardhan, P. [ 1997] "Corruption and development: a review of issues", Journal of Economic
Literature 35(3), September: 1320-1346. Campos,]. E., ed. [forthcoming] Corruption; The boom
and bust of East Asia. Quezon City: Ateneo University Press. Campos,]. E., E. Lien, and S.
Pradhan [forthcoming] "The predictability of corruption: implications for investment" in: J .E.
Campos, ed. Carino, L., ed. [1986] Bureaucratic corruption in Asia: causes, consequences, and
controls. Quezon City: JMC Press. Chang, H-J. [1996] The political economy of industrial
policy. London: Macmillan. Chang, H-J. [forthcoming] "State, capital, and investments in Korea"
in:J.E. Campos, ed. Corpuz, O.D. [1965] The Phzlippines. Englewood, New Jersey: Prentice-
Hall. Della Porta, D. and A. Vannud [1997] "The 'perverse effects' of political corruption", in:
Heywood, ed. Dios, E. de [1999a] "Corruption: a framework" Public Policy, Vol. 3(3).July-
September 1999:35-54. Dios, E. de ( 199b] "Executive-legislative relations in the Philippines:
continuity and change", in C. Barlow, ed. Institutions and economic change in Southeast Asta:
the context of development from the 1960s to the 1990s. London: Edward Elgar. Dios, E. de and
H. Esfahani [forthcoming] "Centralisation, political turnover, and investment in the Philippines",
inJ. E. Campos, ed. Frohlich, N. and]. A. Oppenheimer [1978] Modem political economy. New
Jersey: Prentice-HalL Goudie, A. and D. Stasavage-[199 _]"Corruption: the issues". Technical
Paper No. 122. Gray, C. W. and D. Kauffman [1998]. "Corruption and Development", Finance
and Development, The World Bank. [Link] fand/englishl 0398/ article~
Heywood, P. ed. [ 1997] Political corruption, Oxford: Blackwell Publishers. Jomo, KS.
[forthcoming] "Governance, rent-seeking, and private investment in Malaysia" in:]. E. Campos,
ed. VOLUME V NUMBER 1 January-June 2001 41 De Dios & Ferrer Kaufmann, D. and S. Wei
[ 1999] "Does 'grease money' speed up the wheels of commerce?" World Bank Policy Research
Paper 2254 (December). Krueger, A. [ 197 4] "The political economy of of the rent-seeking
society" American Economic Review, (June) 64:291-303. Laffont, J. -J. and J. Tirole [ 199 3] A
theory o/ incentives in procurement and regulation. Cambridge, Mass.: MIT Press. Lambsdorff, J
[1999] "Corruption in empirical research: a review". Transparency International Working Paper
(November). Leff, N. [ 1964] "Economic development through bureaucratic corruption",
American Behavioral Scientist: 8-14. Li S. and P. Lian [1999] "Governance and investment: why
does China attract largescale FDI despite widespread corruption?" inJ.E. Campos, ed. Mauro
[1995] "Corruption and growth", Quarterly Journal a/Economics August. Mauro, P. [1997].
"Why worry about corruption?", Economic Issues No.6, International Monetary Fund. Mcintyre,
A [1999] "Investment, property rights, and corruption in Indonesia". in J.E. Campos, ed. North,
D. ( 1990) Institutions, Institutional Change and Economic Performance. New York: Cambridge
University Press. Okimoto, D. [1989] Between MIT! and the market: Japanese industrial policy
for high technology. Stanford: Stanford University Press. Rose-Ackerman [1998]. "Corruption
and Development", Annual World Bank Conference on Development Economics, Washington:
The World Bank. Rose-Ackerman, S. [1998] "Corruption and the global econopmy", in G.
Shabbir Cheema, ed. Corruption and integrity improvement initiatives in developing countries,
1998, Management Development and Governance Division, United Nations Development
Programme. Shleifer and Vishny [1993] "Corruption", Quarterly Jourunal a/Economics Taduran,
R [2000] '1\ phenomenon called fraud'', Recent economic indicators, October, School of
Economics, University of Asia and the Pacific: 4-5. World Bank [ 1997 a] World development
report: the state in a changing world. New York: Oxford University Press. World Bank [ 1997b]
Helping countries combat corruption: the role o/ the World Bank. Poverty Reduction and
Economic Management. September.