Francis Fukuyama

Corruption has in many ways become the defining issue of the 21st century, just as the 20th century was characterised by large ideological struggles between democracy, fascism and communism. Today a majority of the world’s nations accept the legitimacy of democracy and at least pretend to hold competitive elections. What really distinguishes political systems from one another is the degree to which the elites ruling them seek to use their power in the service of a broad public interest or simply to enrich themselves, their friends and their families. Countries from Russia and Venezuela to Afghanistan and Nigeria all hold elections that produce leaders with some degree of democratic legitimacy. What distinguishes them from Norway, Japan or Britain is not so much democracy as the quality of government which, in turn, is greatly affected by levels of corruption.

Corruption hurts life outcomes in a variety of ways. Economically, it diverts resources away from their most productive uses and acts like a regressive tax that supports the lifestyles of elites at the expense of everyone else. Corruption incentivises the best and the brightest to spend their time gaming the system, rather than innovating or creating new wealth. Politically, corruption undermines the legitimacy of political systems by giving elites alternative ways of holding onto power other than genuine democratic choice. It hurts the prospects of democracy when people perceive authoritarian governments to be performing better than corrupt democratic ones and undermines the reality of democratic choice.

However, the phenomenon labelled ‘corruption’ comprises a wide range of behaviours whose economic and political effects vary greatly. It is remarkable that, for all of the academic effort put into the study of corruption, there is still no broadly accepted vocabulary for distinguishing between its different forms. Before we can tackle corruption, we need some conceptual clarity as to what it is and how it relates to the broader problem of good government.

2.1 Corruption as a modern phenomenon
Corruption can exist in many contexts, from bribery in a sports organisation to a secretary stealing from the office pool. I am here going to focus on political corruption, which concerns the abuse of public office for private gain (see Johnston 2005, p. 11).

The first point to note is that corruption is a modern phenomenon. The very terms ‘public’ and ‘private’ did not always exist. In the European medieval era, virtually all regimes were what Max Weber labelled ‘patrimonial’ – that is, political authority was regarded as a species of private property which could be handed down to descendants as part of their patrimony. In dynastic times, a king could give away an entire province with all of its inhabitants to his son or daughter as a wedding present, since he regarded his domain as a private possession. Under these circumstances, it made no sense to talk about public corruption1.

The concept that rulers did not simply own their domains but were custodians of a broader public interest was one that emerged gradually in the 16th and 17th centuries. Theorists such as Hugo Grotius, Jean Bodin, Thomas Hobbes and Samuel von Pufendorf began to argue that a ruler could be legitimately sovereign not by right of ownership, but out of a kind of social contract by which he protected public interest – above all, the common interest in peace and security. The very notion that there was a potential conflict between public and private interest emerged with the rise of modern European states. In this respect, China beat Europe to the punch by nearly 1,800 years, having been one of the earliest civilisations to develop a concept of an impersonal state that was the guardian of a collective public interest.

Today no rulers dare assert publicly that they ‘own’ the territories over which they exercise authority; even traditional monarchs such as those in the Arab world claim to be serving a broader public interest. Hence we have the phenomenon that political scientists label ‘neo-patrimonialism’ – in which political leaders pretend to be modern servants of the common good in political systems with modern trappings like parliaments, ministers and bureaucracies. But the reality is that elites enter politics to extract rents or resources and enrich themselves and their families at the expense of everyone else.

A modern state which seeks to promote public welfare and treats its citizens impersonally is not just a recent phenomenon, but also one that is difficult to achieve and inherently fragile. The reason for this has to do with human nature. Human beings are social creatures, but their sociability takes the very specific forms of favouritism towards family and friends.

The demand that we treat people on an impersonal basis, or hire a stranger who is qualified rather than a relative or a friend, is not something that comes naturally to human beings. Modern political systems set up incentives and try to socialise people into different forms of behaviour. But because favouritism towards friends and family is a natural instinct, there is a constant danger of relapse – something I have elsewhere labelled ‘repatrimonialisation’ (Fukuyama 2011).

People who live in rich developed countries often look down on countries pervaded by systemic corruption as if they are somehow deviant cases. But the truth of the matter is that, up until a few centuries ago, there were virtually no modern uncorrupt states. Making the transition from a patrimonial or neo-patrimonial state to a modern impersonal one is a difficult and historically fraught process, much more difficult in most respects than making the transition from an authoritarian political system to a democratic one.

But if most countries throughout most of human history were patrimonial or neo-patrimonial, there were still large differences between them with regard to the quality of government. So we need to make some finer distinctions between types and levels of corruption.

2.2 Types of corruption
There are two separate phenomena related to corruption that are not identical to it. The first is the creation and extraction of rents, and the second is patronage or clientelism.

In economics, a rent is technically defined as the difference between the cost of keeping a good or service in production and its price. One of the most important sources of rents is scarcity: natural resource rents exist because the selling price of oil far exceeds the cost of pumping it out of the ground.

Rents can also be artificially generated by governments. Many of the most common forms of corruption revolve around a government’s ability to create artificial scarcities through licensing or regulation. Placing tariffs on imports restricts imports and generates rents for the government; one of the most widespread forms of corruption around the world lies in customs agencies, where the customs agent will take a bribe to reduce the duties charged or expedites the clearance process so that the importer will have their goods on time.

The ease with which governments can create rents through their taxation or regulatory powers has led many economists to denounce rents in general as distortions of efficient resource allocation by markets and to see rent creation and distribution as virtually synonymous with corruption. The ability of governments to generate rents means that many ambitious people will choose politics rather than entrepreneurship or the private sector as a route to wealth.

But while rents can be and are abused in the fashion described, they also have perfectly legitimate uses, which complicate any blanket denunciation. The most obvious type of a ‘good’ rent is a patent or copyright – by which the government gives the creator of an idea or creative work the exclusive right to any resulting revenues for some defined period of time. Economists Mushtaq Khan and Jomo Kwame Sundaram (2000) point out that many Asian governments have promoted industrialisation by allowing favoured firms to generate excess profits, provided they were ploughed back into new investment. While this opened the door to considerable corruption and abuse, it also worked as a means of stimulating rapid growth at a rate possibly higher than market forces on their own would have produced.

All government regulatory functions create artificial scarcities and therefore rents. But while we can argue about the appropriate extent of regulation, few people would like to see these functions simply abandoned. The creation and distribution of rents by governments have a high degree of overlap with corruption, but are not simply the same phenomenon.

The second phenomenon that is often identified with corruption is that of patronage or clientelism. A patronage relationship is a reciprocal exchange of favours between two individuals of different status and power, usually involving favours given by the patron to the client in exchange for the client’s loyalty and political support. The favour given to the client must be a good that can be individually appropriated, such as a job in the post office, a Christmas turkey or a get-out-of-jail card for a relative, rather than a public good or policy that applies to a broad class of people (Eisenstadt and Roniger 1984). Patronage is sometimes distinguished from clientelism by scale; patronage relationships are typically face-to-face ones between patrons and clients and exist in all regimes whether authoritarian or democratic, while clientelism involves larger-scale exchanges of favours between patrons and clients, often requiring a hierarchy of intermediaries (see Scott 1972). Clientelism thus exists primarily in democratic countries where large numbers of voters need to be mobilised (Piattoni 2001, pp. 4–7).

Clientelism is considered a bad thing and a deviation from good democratic practice in several respects. In a modern democracy, citizens are supposed to vote based on the politician’s promises of broad public policies or a ‘programmatic’ agenda. Such choices are supposed to reflect general views of what is good for the political community as a whole and not just what is good for one individual voter.

Of course, voters in advanced democracies cast their ballots according to their self-interest; programmes targeted at one group of citizens are nonetheless justified in terms of broad concepts of justice or the general good. Moreover, targeted programmes must apply impartially not to individuals but to broad classes of people. Targeted benefits to individuals are bad from the standpoint of social justice. In clientelistic systems, redistributive programmes that are supposed to help all poor people, for example, end up benefiting only those poor people who support a particular politician. This weakens support for effective universal policies and preserves existing social inequalities.

Nevertheless, there is reason to think that clientelism is actually an early form of democratic participation. In the United States and other countries, it was a way of mobilising poor voters and therefore encouraging them to participate in a democratic political system. It was suboptimal when compared to programmatic voting, yet provided a degree of accountability insofar as the politician still felt obligated to provide some benefits in return for political support. In that respect, clientelism is quite different from a more destructive form of corruption in which a politician simply steals from the public treasury for the benefit of his or her family, without any obligation to provide a public service in return. The problem with clientelism is that it usually does not remain confined to a mechanism for getting out the vote, but morphs into misappropriation.

A final conceptual distinction that needs to be made is between corruption and low state capacity.

‘Anti-corruption and good governance’ has become an often-repeated slogan in the development community and some people treat good governance and the absence of corruption as equivalents. Yet they are very different: a squeaky-clean bureaucracy can still be incompetent or ineffective in doing its job, while corrupt ones can provide good services 2. Beyond low levels of corruption, good governance requires state capacity – that is, the human, material and organisational resources necessary for governments to carry out their mandates effectively and efficiently. It is linked to the skills and knowledge of public officials and whether they are given sufficient autonomy and authority to carry out their tasks.

Corruption, of course, tends to undermine state capacity (for example, by replacing qualified officials with political patronage appointees); conversely, highly professional bureaucracies tend to be less subject to bribery and theft. Low levels of corruption and high state capacity therefore tend to be correlated around the world. But getting to good governance is a much larger task than simply fighting corruption.

The distinction between corruption and low state capacity allows us to better understand differences between the effects of corruption in countries around the world. In the World Bank Institute’s Worldwide Governance Indicators for 2014, China ranks in the 47th percentile with respect to control of corruption, behind Ghana and just ahead of Romania (World Bank 2014). On the other hand, China has a great deal of state capacity. In the government effectiveness category, it is in the 66th percentile, while Romania is in the 55th and Ghana is in the 44th (World Bank 2014). This validates the common perception that the Chinese Government has a great deal of capacity to achieve the ends it sets, despite strong perceptions of pervasive corruption. The predictability and scale of corruption are also important; if a business owner expects to pay 10% of the transaction value in bribes, they can regard that as a kind of tax, which is less damaging to investment than a bribery level of 75% or one that varies arbitrarily from year to year.

2.3 Overcoming corruption
The first generation of anti-corruption measures taken in the mid-1990s by development finance institutions involved ambitious efforts to overhaul civil service systems along Weberian lines: incentivising officials by increasing wage dispersion and setting formal recruitment and promotion criteria. These measures had very little effect; the problem lay in the fact that corrupt governments were expected to police themselves and to implement bureaucratic systems developed over long periods in rich countries with very different histories. More recent efforts have focused on fighting corruption through transparency and accountability measures – that is, increasing the monitoring of agent behaviour and creating positive and negative incentives for better compliance with the institution’s goals. This has taken a variety of forms: cameras placed in classrooms to ensure that teachers show up for work; participatory budgeting where citizens are given a direct voice in budgeting decisions; and websites where citizens can report government officials taking bribes. Since governments cannot be trusted to police themselves, civil society has often been enlisted in a watchdog role and mobilised to demand accountability. Mechanisms like anti-corruption commissions and special prosecutors have, if given enough autonomy, also shown some success in countries such as Indonesia and Romania.

These later efforts, however, have also had uneven success (see, for example, Kolstad and Wiig 2009; Mauro 2002). In particular, transparency initiatives by themselves do not guarantee changes in government behaviour. For example, in countries where clientelism is organised along ethnic lines, co-ethnics are frequently tolerant of leaders who steal. Elsewhere, citizens may be outraged by news of corruption, but then have no clear way of holding individual politicians or bureaucrats accountable. In other cases, successes in punishing individual politicians are not sufficient to shift the normative framework in which virtually everyone in the political class expects to profit from office. Finally, anti-corruption campaigns may disrupt informal understandings and personal relationships that underpin investment and trade: without formal property rights and contract enforcement under a system of independent courts, the paradoxical short-term effect of prosecuting corrupt officials may be to deter new investment and thereby lower growth.

There is a single truth underlying the indifferent success of existing transparency and accountability measures to control corruption. The sources of corruption are deeply political. Without a political strategy for overcoming this problem, any given solution will fail. Corruption in its various forms – patronage, clientelism, rent-seeking and outright theft – all benefit existing stakeholders in the political system, who are generally very powerful players.

Lecturing them about good government or setting up formal systems designed to work in modern political systems will not affect their incentives and therefore will have little transformative effect. That is why transparency initiatives on their own often fail. Citizens may be outraged by news about corruption, but nothing will happen without collective-action mechanisms to bring about change. The mere existence of a democratic political system is no guarantee that citizens’ anger will be translated into action; they need leadership and a strategy for displacing entrenched stakeholders from power. Outside pressure in the form of loan conditionality, technical assistance or moral pressure is almost never sufficient to do the job. Anti- corruption commissions and special prosecutors who have had success in jailing corrupt officials have done so only because they receive strong grassroots political backing from citizens.

2.4 The American experience
The political nature of corruption and the necessarily political nature of the reform process can be illustrated by the experience of the United States in the 19th century (as I describe in Fukuyama 2014, chapters 9–11). American politics in that period was not too different from politics in contemporary developing democratic countries such as India, Brazil or Indonesia. Beginning in the 1820s, American states began extending the franchise to include all white males, vastly expanding the voter base and presenting politicians with the challenge of mobilising relatively poor and poorly educated voters. The solution, which appeared particularly after the 1828 presidential election that brought Andrew Jackson to power, was the creation of a vast clientelistic system. Elected politicians appointed their supporters to positions in the bureaucracy or rewarded them with individual payoffs like Christmas turkeys or bottles of bourbon. This system, known as the spoils or patronage system, characterised American government for the next century, from the highest federal offices down to local postmasters in every American town or city. As with other clientelistic systems, patronage led to astonishing levels of corruption, particularly in cities such as New York, Boston and Chicago where machine politicians ruled for generations.

This system began to change only in the 1880s as a consequence of economic development. New technologies like the railroads were transforming the country from a primarily agrarian society into an urban industrial one. There were increasing demands from business leaders and from a newly emerging civil society for a different, more modern form of government that would prioritise merit and knowledge over political connections. Following the assassination of the newly elected President James A. Garfield in 1881 by a would-be office seeker, Congress was embarrassed into passing the Pendleton Act. It established a US Civil Service Commission for the first time and the principle that public officials should be chosen on the basis of merit. Even so, expanding the number of classified (i.e. merit- based) officials met strong resistance and did not become widespread until after the First World War. Individual municipal political machines such as Tammany Hall in New York were not dismantled completely until the middle of the 20th century.

The American experience highlights a number of features of both corruption and the reform of corrupt systems. First, the incentives that led to the creation of the clientelistic system were deeply political. Politicians got into office via their ability to distribute patronage; they had no incentive to vote in favour of something like the Pendleton Act that would take away those privileges. The only reason it passed was a tragic exogenous event – the Garfield assassination – which mobilised public opinion in favour of a more modern governmental system.

Second, reform of the system was similarly political. The Progressive Era saw the emergence of a vast reform coalition made up of business leaders, urban reformers, farmers and ordinary citizens who were fed up with the existing patronage system. It required strong leadership from politicians like Theodore Roosevelt who was himself head of the US Civil Service Commission. It also required a clear reform agenda pointing towards modern government, formulated by intellectuals such as Frank Goodnow, Dorman Eaton and Woodrow Wilson. Finally, reform was helped along by economic development. Industrialisation in the US produced new social groups such as business leaders who needed efficient government services, a broad and better-educated middle class who could mobilise for reform, and a grassroots organisation of civil society groups.

2.5 Conclusions
The American experience is suggestive of how progress in the fight against corruption may be waged in contemporary societies suffering from it. Reform is always a political matter that will require formation of a broad coalition of groups opposed to an existing system of corrupt politicians. Grassroots activism in favour of reform may emerge spontaneously, but such sentiments will not be translated into real change until it receives good leadership and organisation. Reform also has a socio-economic basis: economic growth often produces new classes and groups that want a different, more modern politics.

America points to another feature of anti-corruption efforts. Control of corruption was very much bound up with efforts to increase state capacity. The period that saw the emergence of an industrial economy was also characterised by huge increases in levels of education – particularly higher education, which produced an entirely new class of professionals who worked for both private businesses and the government. One of the first government agencies to be modernised in the late 19th century was the US Department of Agriculture, which benefited from a generation of professional agronomists trained in the numerous land-grant universities that sprang up around the United States. The latter, in turn, were the product of the far-sighted Morrill Act of 1862 that sought to increase agricultural productivity (among other things) through higher education.

It would not have been possible to reform the old patronage-based bureaucracy without access to the human capital represented by this entire generation of university- educated officials. Every important reform effort undertaken to create modern state bureaucracies – in Germany, Britain, France, Japan and elsewhere – was accompanied by parallel efforts to modernise the higher education system in ways that would benefit public administration. Today development finance institutions focus on helping to provide universal primary and secondary education to poor countries and have largely given up on supporting elite education. The reasons for this are understandable, but do not correspond to the historical experience of state modernisation in countries that became rich in earlier eras.

These general observations about historical efforts to build modern uncorrupt administrations suggest that the process will be an extended one, characterised by prolonged political struggle. Fortunately, having a modern bureaucracy is not a sine qua non of economic development. No existing rich country had a squeaky-clean government in its early stages of economic growth – neither Britain, nor the United States in the 19th century, nor China today. Corruption and weak governance are obstacles to economic growth, but economic growth can happen also in poorly governed societies and will produce, over time, social conditions and resources that will make government reform more feasible. This is perhaps a pessimistic conclusion, given the fact that rentier states and kleptocratic governments are the source of international conflict and instability in today’s world. But it is also a realistic assessment derived from the historical record.