Honey and Poison: On Corruption
I
For as long as human beings have had governments, they have worried about public corruption. The Hebrew Bible warns repeatedly that those in authority — especially judges — should not take bribes, “for bribes blind the clear-sighted and upset the pleas of those in the right.” The Arthashastra, a third-century Indian text on the art of statecraft, cautioned that just as one cannot avoid “tasting honey or poison on the tip of the tongue,” government officials will inevitably be tempted to steal public money for themselves. Countless other examples — from classical Greece and Rome to Imperial China to the Islamic empires of the Near East — testify to the pervasiveness of public corruption across cultures and across time. Indeed, from the ancient world up through today, corruption has been a central concern of statesmen, philosophers, and journalists — and the undoing of powerful figures and the catalyst for major reform movements. Anxiety over corruption also figures prominently in culture across the centuries — from Shakespeare’s Brutus accusing his friend Cassius of having “an itching palm, to sell and mart your offices for gold to undeservers” to Lin-Manuel Miranda’s Alexander Hamilton rapping that “corruption’s such an old song that we can sing along in harmony.”
And yet the problem of corruption, for all its ubiquity, is often neglected. Perhaps most strikingly, for a very long time the international development community — a shorthand term for the various government agencies, multilateral institutions, and non-governmental organizations focused on improving the well-being and opportunities of the residents of poorer countries — paid scant attention to corruption. This may have been due in part to the belief that corruption, while immoral and unjust, was only marginally relevant to economic development. The comparative lack of attention to corruption was also related to concerns about the political sensitivity of the issue: to talk about corruption is almost always to talk about politics. Indeed, at the World Bank in the 1980s and early 1990s, officials rarely uttered the word “corruption” in public, and referred to it behind closed doors as “the C-word” — a nod to the fact that this was a problem that everyone knew existed but agreed should not be discussed openly.
Roughly a quarter-century ago, this began to change. As is often the case, the process was gradual and the causes complex, so it would be a mistake to attribute the emergence of anticorruption as a central international development issue to any one person or event. Still, at least symbolically, a breakthrough moment occurred in October 1996, when James Wolfensohn, then president of the World Bank, gave what came to be known as the “cancer of corruption” speech. Addressing the annual meeting of the World Bank and International Monetary Fund, Wolfensohn declared in no uncertain terms that to fight global poverty, organizations such as the Bank needed to promote “transparency, accountability, and institutional capacity,” and, more specifically, “to deal with the cancer of corruption.” Though Wolfensohn did not dwell on the issue — his remarks on corruption took up less than two minutes of his address — he did provide a succinct explanation of why corruption was a development issue: “Corruption diverts resources from the poor to the rich, increases the cost of running businesses, distorts public expenditures, and deters foreign investors.” Today such a statement would be unremarkable. But back in 1996 it was a big deal, especially since, as Wolfensohn later recounted, he had been warned shortly after the start of his presidency not to talk about “the C-word.” He ignored that warning — and, crucially, he did so by reframing corruption not as a purely political or moral issue, but as an issue that directly affected economic development. Corruption was now squarely on the international development agenda, and it remains so to this day.
Over the generation since Wolfensohn’s speech, leading multilateral organizations, including the World Bank, IMF, United Nations, and OECD, have paid increasing attention to this issue, forming divisions and sponsoring projects devoted to anticorruption activities. We now have an international anticorruption agreement, the UN Convention Against Corruption (UNCAC), to which most countries in the world are parties (even if compliance is uneven at best); there are also regional anticorruption agreements in the Americas, Europe, Africa, and elsewhere. Donor agencies, such as USAID, the UK’s Department for International Development, Germany’s GIZ, Sweden’s SIDA, and many others, support extensive anticorruption programming. Anticorruption, in short, is on the map.
But despite this progress in the cause of anticorruption — the more sophisticated conception of it, the protocols and agreements — the idea that the international development community should make the fight against corruption a high priority is not universally accepted. Among the many objections to the emphasis on anticorruption as an integral part of international development, I want to highlight — and debunk — three quasi-myths that have gained more traction in these debates than they deserve. I call these ideas quasi-myths, rather than simply myths, because each of these them does have a kernel of truth. But each of these three arguments is, on the whole, more false than true, and more misleading than helpful.
The first quasi-myth is that corruption is a culturally relative concept, such that practices that wealthy Western countries consider corrupt are acceptable in other societies. The familiar refrain here is that in some cultures what “we” would consider a bribe, “they” would consider a gift — or more generally that Western norms regarding the line between the public sphere and the private sphere do not apply in many non-Western societies. Thus, the argument continues, when organizations such as the World Bank or USAID or the OECD promote an anticorruption agenda in developing countries, they are in fact imposing a set of values that are inconsistent with local customs and traditions. In its strongest form, the argument accuses those promoting an international anticorruption agenda of engaging in a form of “moral imperialism,” or even that these efforts are intended to advance Western economic interests (say, in lowering trade and investment costs) while stigmatizing non-Western modes of government and social practices as morally inferior.
The idea that “corruption” is a culturally specific concept, such that practices that would be seen as outrageously corrupt in the West are considered legitimate elsewhere, has a long history. Consider, as one particularly infamous example, the impeachment trial of Warren Hastings, the first British Governor-General of India, which began in 1788 (and dragged on, with frequent delays and interruptions, until 1795). Hastings was impeached for mismanagement and corruption. In his defense, he argued, among other things, that it would be inappropriate to apply British moral standards to his conduct in India, because practices that Englishmen would deem corrupt were part of the normal operation of government in Asia. Edmund Burke, who served as chief prosecutor at Hastings’ trial, denounced this argument as “geographical morality.” To the contrary, Burke insisted, “there is no action which would pass for an action of extortion, of peculation, of bribery and of oppression, in England, that is not an act of extortion, or peculation, of bribery and oppression, in Europe, Asia, Africa, and all the world over.” Despite Burke’s pleas, the House of Lords acquitted Hastings. To be sure, many of those who accuse modern anticorruption campaigners of moral imperialism would condemn Hastings for his conduct in India — he was, after all, a literal imperialist. Yet Hastings’ appeal to “geographical morality” — what we might today call “moral relativism” — has a strong family resemblance to this more modern critique of the international anticorruption agenda.
Is there any truth to the argument that the international anticorruption campaign seeks to impose — deliberately or unintentionally — a set of values, practices, and institutions that are inconsistent with local norms and cultures in non-Western countries? The short answer is no. The overwhelming weight of the empirical evidence — derived from surveys, interviews, in-depth case studies, and other sources — indicates quite clearly that Burke was right and Hastings was wrong. At least when it comes to what we might think of as the “core” forms of corruption — bribery, embezzlement (what Burke called “peculation”), and the like — there is actually remarkably little variation in attitudes and moral evaluations across societies and cultures: these practices are broadly understood as corrupt and wrong, and are roundly and nearly universally condemned.
To be sure, there is more variation across societies with respect to what we might think of as “grey area” corruption, as well as with respect to when certain kinds of corrupt acts might be justifiable given the circumstances. This is the kernel of truth to the argument that different cultures have different attitudes toward what counts as (wrongful) corruption. Even here, though, we need to be careful about the implicit cultural condescension of assuming that what “we” consider corrupt, “they” would consider appropriate. Very often the difference runs in the other direction. After all, campaign contributions and lobbying activities that many countries would consider blatantly corrupt are treated in the United States as not only lawful but as constitutionally protected. More importantly, the extent of cultural variation in the understanding of corruption is relatively modest, and not pertinent to the forms of corruption that Wolfensohn and others like him have in mind. The idea that bribery and embezzlement are the concerns only of wealthy Western countries, and that the prominence of an anticorruption agenda is therefore a form of Western neo-imperialism, finds essentially no support in the extensive research on what the residents of non-Western countries actually think.
Where, then, does this myth come from, and why does it persist? I have three conjectures. First, some who push this idea are as blatantly self-serving as Warren Hastings. The employees of Western multinationals who pay bribes to government officials in non-Western developing countries, and the officials who take those bribes, have an incentive to suggest that in the countries in question these so-called “bribes” are actually a manifestation of a rich and longstanding cultural tradition of offering gifts as a sign of respect. Second, some of those who advance the “moral imperialism” critique of the anticorruption agenda seem to harbor a deep (and perhaps understandable) mistrust of Western governments and multilateral institutions generally; these skeptics are primed to be receptive to the idea that an anticorruption campaign spearheaded by such entities is likely to have a hidden agenda.
The third possible explanation, which I would guess to be the most important, is that it is easy to mistake cynical or fatalistic resignation about corruption — an attitude that is quite widespread in much of the developing world — for the cultural legitimacy of corrupt practices. But these are not at all the same thing. Many people tolerate corruption, or even participate in petty corruption themselves, out of a feeling that they are trapped, that corruption is inevitable, that the system is rigged and there is nothing they can do about it. Such tolerance and participation can be misperceived as “cultural acceptance,” especially when accompanied by rationalizations that invoke venerable cultural tropes. But grudging tolerance and rationalization are not the same thing as moral assent and legitimacy. And when it comes to corruption—at least to core forms of corruption like bribery and embezzlement — there is far less variation in attitudes across countries and cultures than one might expect.
The second quasi-myth that is sometimes offered up as a reason to object to the international development community’s focus on anticorruption is the idea that, at least in developing countries, corruption can actually help the process of economic development. Rather than being sand in the wheels of the economy, the argument goes, corruption may instead grease those wheels, enabling entrepreneurs and investors to cut through burdensome red tape and enter markets that would otherwise be inaccessible. This idea was nicely captured by Samuel Huntington, in 1968 in his book Political Order in Changing Societies. “In terms of economic growth,” Huntington wrote, “the only thing worse than a society with a rigid, over-centralized, dishonest bureaucracy is one with a rigid, over-centralized, honest bureaucracy.” What he meant was this: if the government has put in place excessive and inefficient rules and regulations that stifle economic activity, then the economy will be better off if the public officials charged with enforcing those regulations take “grease” payments to look the other way, rather than insisting on rigorous and scrupulous enforcement of the misguided rules. It follows from this that vigorous action to suppress corruption — for example, by monitoring bureaucrats more closely and imposing stiffer penalties on those caught offering or accepting bribes — may, if unaccompanied by other reforms to the regulatory system, actually worsen a society’s economic prospects.
We should give the argument its due, because it too contains a kernel of truth. When the formal rules are inefficient — for example, when securing a business operating permit through the normal channels would take an inordinate amount of time and expense — then corruption may indeed function as an efficiency-enhancing grease. That said, even proponents of this view would acknowledge that in such circumstances corruption is at most what economists would call a “second-best” solution. Those who endorse the efficient grease hypothesis would also presumably acknowledge that other forms of corruption can have substantial negative impacts on the economy — for example, when corruption facilitates the subversion of government programs that enhance productivity and welfare. So the question whether corruption is, on the whole, more likely to grease or to sand the wheels of economic development is ultimately an empirical question.
There has been quite a bit of research on this empirical question since Wolfensohn delivered his “cancer of corruption” speech, and while the issue is not entirely settled (issues like this rarely are, given the challenges of isolating causal relationships in the available data), the overwhelming weight of the evidence suggests that Wolfensohn was more correct than Huntington. Corruption is far more often an impediment to economic development than a facilitator of economic development. One of the reasons for this was identified by Gunnar Myrdal in 1968 in his classic Asian Drama: An Inquiry into the Poverty of Nations. Writing in response to Huntington and others who had argued that corruption was a way for entrepreneurs to cut through bureaucratic red tape, Myrdal pointed out that much of this red tape had been deliberately imposed precisely to create more opportunities for extracting bribes. While excessive red tape may lead to corruption, corruption also leads to the proliferation of red tape — which suggests that effective anticorruption measures can make it politically easier to eliminate needless regulations.
An even more important reason why corruption is more often associated with worse economic outcomes is that, while it may be true in certain contexts that corruption enables the circumvention of excessive business licensing requirements and other inefficient rules, there are a whole lot of other things that governments do that are important to economic development — things such as investing in infrastructure, supporting health and education, maintaining order, providing impartial courts and dispute resolution services, and enforcing rules that protect the integrity and efficiency of markets — that are undermined by widespread corruption. So while we should not dismiss out of hand the idea that corruption can sometimes function as an efficient grease, and we should certainly be mindful of the fact that excessive, inefficient regulations can both encourage corruption and inhibit economic growth, the idea that fighting corruption in developing countries will prove counterproductive because of corruption’s supposed efficiency-enhancing properties seems by and large inconsistent with the best available evidence.
The third quasi-myth that sometimes comes up in debates over whether anticorruption should be a high priority for the international development community suggests a quite different reason for skepticism. Even if one believes that endemic corruption in developing countries is both immoral and economically detrimental in those countries, some critics contend that there is still no point in making anticorruption a central agenda item, because there is nothing that can realistically be done about corruption, at least in the short to medium term. On this view, cultures of corruption are so deeply embedded in certain societies that corruption should be treated as an unfortunate but unavoidable constraint on a society’s development prospects — like being landlocked, or located in the tropics, or having deep ethnic cleavages. In this pessimistic view, even though widespread corruption is a problem, few of the reforms or initiatives championed by anticorruption advocates have a realistic chance of making more than a trivial difference. It is just not a problem that can be effectively addressed through new policies or institutional reforms; the best one can do is to hope that long-term historical trends eventually produce a cultural change — but even that might be optimistic, given the alleged persistence of cultures of corruption across extended historical time periods.
Here again, there is an element of truth to the argument. Systemic corruption often does have self-reinforcing and self-perpetuating tendencies — corruption begets corruption, creating a vicious cycle that can make entrenched corruption very difficult to dislodge. But the idea that corruption is the inevitable product of some deep cultural tradition that developed centuries ago, and that only those countries lucky enough to have inherited a “good” cultural tradition (say, from northern Europe) have much hope of making headway against the corruption problem, is not only tinged with racism, but is inconsistent with most of the available evidence.
For starters, there is no systematic evidence — when one controls for other factors (like per capita GDP) — that particular cultural traditions have a robust correlation with present-day corruption. I should acknowledge an interesting caveat: some researchers have found that, all else equal, majority-Protestant countries have lower levels of perceived corruption than majority-Catholic or majority-Muslim countries. But when one looks more closely at individual-level data, there is no strong evidence that individual Protestants have different attitudes toward corruption than do individual Catholics, Muslims, or others, and it seems much more likely that the apparent correlation between Protestantism and perceived corruption at the country level is spurious.
But the more important evidence against the notion that countries are locked into particular levels of corruption by their cultural heritage is that this view is inconsistent with the historical experience of those countries that today are viewed as relatively less corrupt. Consider Scandinavia. Today we have a stereotype of the Scandinavian countries as being very clean — which is fairly accurate, at least if we compare Scandinavia to other parts of the world. We sometimes imagine that this is because of some longstanding and essential feature of Scandinavian culture. But if we were to step into a time machine and go back a couple of centuries, things would look quite different. Things were rotten in the state of Denmark not just in Shakespeare’s imagined medieval period, but up through much of the eighteenth century and beyond. The fight against corruption in Denmark got underway after the establishment of an absolute monarchy in 1660, but the Danish state did not get corruption under control until a series of reforms adopted over the course of the eighteenth and early nineteenth centuries. As for Sweden, at the turn of the nineteenth century the Swedish state was extremely corrupt, particularly with respect to rampant nepotism and the purchase and sale of offices (and there was a fair amount of garden-variety bribery as well). To observers at the time, before the significant reform processes that took place in these and other countries, it might well have seemed that corruption was deeply embedded in these countries’ cultures. The “clean” Scandinavian culture that informs our contemporary stereotypes really only emerged in the mid-to-late nineteenth century.
The extent to which modern “good performers” did not simply inherit a cultural tradition of clean government is illustrated even more vividly by the United States. To be sure, even today the United States is no paragon of government integrity. In comparison to most other countries in the world, though, American government has relatively low levels of bribery, embezzlement, and similar forms of corruption. But it was not always thus. Throughout much of the nineteenth century, corruption in the United States was rampant — especially at the state and local level, but at the national level as well. We do not possess international corruption indexes that go back to the nineteenth century, but if we did, and if such indexes were on a common scale, it is quite likely that the United States in the 1840s and 1850s, and perhaps as late as the 1890s, would receive corruption scores comparable to the scores that developing democracies like India, Brazil, South Africa, and Ukraine receive today.
Certainly it seemed that way to domestic and foreign observers at the time. In 1849, in his book describing his travels in the United States, the Scottish journalist Alexander Mackay remarked that he had heard several Americans declare “that they believe their own government to be the most corrupt on earth.” Nearly a decade later, another foreign observer, the British MP William Edward Baxter, expressed his shock at the level of corruption in New York, reporting that “as great corruption exists [there] as was ever brought to light in the days of the Stuarts.” And it was not just foreign observers who made such damning comparisons. In 1858, the same year Baxter published his book, Senator Robert Toombs of Georgia lamented that while Americans may “speak of the corruptions of Mexico, of Spain, [and] of France, … I do not believe today that there is as corrupt a Government under the heavens as these United States.” A small-town newspaper editor who visited Washington D.C. that same year expressed his shock at the brazen purchase and sale of offices brokered by party leaders in the Senate, the White House, and various government bureaus, with “the actual sum of money to be paid for an office … as publicly named … as the price of dry goods are named between a dealer … and his customers.” These anecdotal observations, though perhaps a bit hyperbolic, have been largely corroborated by historians. The United States was, for much of the nineteenth and early twentieth century, mired in forms of systemic corruption not so different from those afflicting much of the developing world today.
Why does this matter? It matters because if countries such as Denmark, Sweden, the United States, and others suffered from widespread and systemic government corruption at an earlier point in their histories, then we should question the notion that countries that today have corruption more under control have achieved this because of some deep and immutable cultural inheritance, by a lucky historical break, and that countries where corruption is entrenched and systemic are likely trapped in that state due to inherited norms that are too deeply ingrained in their cultures to be dislodged through institutional and political reform. If countries such as the United States managed to make the transition from a state of endemic corruption to one in which corruption, while of course still present, is aberrational and manageable, then this warrants the hope that other countries that today face a seemingly intractable regime of corruption might be able to make a similar transition.
II
But how can modern developing countries make such a transition? The historical examples are encouraging illustrations that such changes are possible and help to debunk the view that cultures of corruption are immutable — but they are less useful in supplying a template for modern reformers, given the dramatic differences in historical, economic, and political context. Even if we endorse the view that James Wolfensohn laid out back in 1996 — that anticorruption should be front and center in the international development agenda — we still need to ascertain what institutional reforms and policy initiatives can help to effect, under modern conditions in the contemporary developing world, the sort of transition that took place in wealthy Western countries much earlier.
One popular prescription — though not one that institutions like the World Bank can openly advocate — is democratization. The logic here is straightforward and compelling. Corruption thrives when those who wield power are not accountable to those they are supposed to serve; and since democracy makes public officials accountable to the populace through regular elections — and because people in just about every country find corruption objectionable — more democratic countries are likely to have substantially lower levels of corruption than less democratic countries, all else equal. The global democratization project, on this view, is also a global anticorruption project.
Matters are more complicated, however. It turns out that the evidence that democratization regularly results in improvements in government integrity is not terribly strong. It is certainly the case that, even when one controls for national wealth and other confounding factors, countries that have been strong democracies for quite a long time are less corrupt (or at least are perceived as less corrupt) than other countries. But if one excludes from the analysis the relatively small number of countries that have been full democracies for over forty years (mostly though not exclusively Western European countries and the former European settler colonies in North America and Oceania), there does not appear to be a robust and consistent relationship between the level of democracy and the level of (perceived) corruption. In other words, while old, established democracies do seem to be notably cleaner than other countries, new democracies and partial democracies do not appear less corrupt, on average, than non-democracies. (Some studies have even suggested that new democracies and partial democracies might be somewhat more corrupt than autocracies, though the evidence here is not as strong.) More anecdotally, it is not hard to come up with numerous examples of politicians suspected or known to be quite corrupt who regularly win elections.
What explains the puzzling lack of strong empirical support for the view that democratization reduces corruption? One possibility is that democracy does indeed have corruption-suppressing effects, but only when democracy has become fully entrenched and institutionalized. If that is so, then we would have cause for optimism that some of the newer democracies will eventually see significant improvements in public integrity; we just need to be patient, and work to further strengthen democratic institutions. It could also be the case that new and partial democracies actually do have lower levels of corruption than autocracies, but autocracies are, on average, better able to suppress evidence of corruption, thus leading the international corruption indexes, which rely substantially on perceptions, to systematically underestimate the extent of corruption in autocracies relative to (newer) democracies.
Maybe. But there is also a more pessimistic reading of the available evidence. It may be that “first wave” democracies are systematically different from second- and third-wave democracies, perhaps because the defining characteristic of a “democracy” (reasonably free and fair elections held on a regular basis) is not, by itself, sufficient to promote clean government. Without other norms and institutions in place, elections may become little more than competitions between rival patronage networks. Moreover, the expense of running a modern election campaign, coupled with the fact that electoral victory may be the key to (illicit) access to state resources, may actually encourage certain kinds of corruption. Indeed, many new democracies have seen an upsurge in corruption that is directly associated with the democratic process. Even though most citizens claim to find corruption reprehensible — and this claim may well be sincere — it may nevertheless be the case that in newer democracies in less developed countries, certain forms of corruption are helpful, perhaps essential, to winning elections and maintaining coalitions. If this more pessimistic view is accurate, then even though we can and should support democratization, we need to be more circumspect about the impact of democratization upon the attempt to eliminate corruption, at least in the short to medium term.
Another common prescription for addressing systemic corruption through broad institutional reform focuses on shrinking the size and the scope of government. The idea here is that a root cause of much of the corruption that afflicts developing and developed countries alike is a bloated state sector, and an outsize role for the government — rather than impersonal market forces — in allocating resources. This is allegedly a recipe for corruption, because those with power over the allocation of resources (politicians and bureaucrats) will seek to leverage that power to obtain wealth, while those with wealth or connections will leverage those advantages to influence the public officials responsible for allocating public resources. Perhaps the most well-known proponent of the view that reducing government size is the key to taming systemic corruption was the economist Gary Becker, whose views were succinctly captured in the titles of two opinion pieces he penned in the mid-1990s: “To Root Out Corruption, Boot Out Big Government,” and “If You Want To Cut Corruption, Cut Government.”
Yet this hypothesis fares even worse than the hypothesis that democratization is the key to reducing corruption. It turns out that government size (typically measured by government consumption spending and/or government revenue as a percentage of gross domestic product) is strongly and consistently correlated with lower levels of perceived corruption, even if one controls for other factors like per capita GDP. In other words, those countries with governments that do more taxing and spending, relative to the overall size of the national economy, are generally perceived as cleaner than otherwise similar countries with smaller governments.
While this relationship is quite well established, the reasons for it remain unclear. One possibility is that a larger government is associated with a stronger social safety net and lower levels of economic inequality, and this may reduce forms of corruption that are driven by economic insecurity. Another possibility is that larger governments tend to spend more on things such as education, which may have corruption-reducing effects. Larger governments may also have better paid, more professional, and more effective bureaucracies, and may invest more in effective law enforcement and judicial institutions. These and other hypotheses essentially suggest that the projects on which larger governments are spending public money are often things with corruption reducing-effects that tend to outweigh whatever increase in corruption may be associated with giving government officials a greater role in resource allocation.
Alternatively, or in addition, perhaps citizens demand more integrity from their governments when those governments are doing more, and when a broader and more affluent set of citizens are the beneficiaries of government programs. Put another way, larger governments may be associated with stronger citizen demand for clean government. That possibility also suggests that the relationship between larger governments and lower levels of public corruption may be due, at least in part, to a kind of reverse causation: perhaps citizens are only willing to support significant growth in the size of government when the public sector has a reputation for integrity; where corruption is widespread, expanding government programs may be less popular and therefore less likely. Relatedly, government decision-makers may not try to tax or spend as much if they expect a greater share of the revenue or expenditures to be stolen or misappropriated. On this account, political support for government expansion tends to rise as (perceived) public corruption declines.
While there remains considerable uncertainty about the mechanisms and direction of causation — whether expanding government tends to reduce corruption, or whether cleaner governments are more likely to expand — the simple hypothesis, propounded by Becker and others, that cutting government size is likely to be associated with significant reductions in public corruption finds little support in the available evidence. That prescription also does not accord well with the historical experience of countries such as the United States, where the period of greatest success in fighting corruption — roughly from the turn of the twentieth century through the end of the New Deal — was also a period of extraordinary expansion in the size and scope of government, especially at the national level. None of this is to deny that there are some governments, and some specific government programs, that are bloated and consequently vulnerable to corruption. It is also not to deny that other forms of excessive government intervention — such as the needless red tape and inefficient regulations that Huntington and others worried about — may contribute to corruption, and that certain forms of deregulation might therefore have corruption-reducing effects. But as a general matter, the neo-libertarian view that smaller governments are more honest governments finds precious little empirical support.
If democratizing the political system and shrinking the state are not the magic keys to promoting government integrity, what sorts of measures are more promising? There is probably no single answer. Corruption is not one thing — it is an umbrella term that covers many related but distinct forms of misconduct, which may require different remedial approaches. Corruption takes different forms, and has different roots, in different countries. Therefore, as is true in so many areas of public policy, effective anticorruption efforts must be appropriately tailored to the local context. Those who work in this field are well aware of all this. Indeed, the assertion that there is no “one-size-fits-all” solution to corruption is repeated so often that it has become a truism.
Still, while we should be cautious about making strong and unqualified claims about “what works” in fighting corruption, it would be a mistake to err in the opposite direction by neglecting the lessons that we can draw from the last twenty-five-odd years of academic research and practical experience in this field. And what emerges from that combination of research and experience is that, while there is not One Big Thing that can transform a corrupt system, there are a number of smaller things that often help.
The items on this list are not all that surprising. Strong laws, enforced by efficient, effective, and nonpartisan prosecutors and courts, while not sufficient, are very important. (A brief digression here: Many countries have created specialized anticorruption agencies with investigative and prosecutorial powers; some countries have even created specialized anticorruption courts or judicial divisions. These measures may sometimes be useful, if the creation of a separate entity helps ensure both institutional independence and sufficient capacity, but the track record of specialized anticorruption bodies is mixed. Operational independence and capacity seem to be what’s most important; formal specialization is, at most, a means to those ends.) Effective law enforcement is essential not just with respect to those laws specifically targeting corrupt acts such as bribery and embezzlement, but also in what we might call corruption-adjacent areas such as money laundering and corporate secrecy. “Follow the money” is especially good advice when it comes to addressing grand corruption, as it is often easier for bad actors to cover up evidence of their underlying corruption — or to escape legal accountability in their home countries — than it is for them to hide or explain away the proceeds of their illicit activity. So, some of the most effective anticorruption measures of the last decade or so target the money, seeking to locate, freeze, seize, and eventually return or otherwise redistribute stolen assets.
Within the bureaucracy, in addition to the enforcement of criminal laws and ethical rules, audits of government programs turn out to be quite effective in reducing theft and other forms of misappropriation. This might not sound that surprising, but it is nonetheless important. Rigorous empirical research has found that even in what we might think of as challenging environments for fighting bureaucratic corruption — such as Indonesia, Brazil, and Mexico — the knowledge that a local government or department’s accounts will be subject to an audit substantially reduces irregularities that are likely attributable to corruption. More generally, the promotion of a professionalized, semi-autonomous civil service, with personnel decisions insulated from partisan political actors, appears to be consistently associated with lower levels of corruption. (Interestingly, civil service salaries do not seem to have as clear or strong an association with corruption levels, though at the extremes — when civil servants are paid very well or very poorly — there is anecdotal evidence of an effect on integrity.) What we might think of as the classic Weberian vision of bureaucracy — characterized by autonomy (at least from partisan politics), norms of professionalism, regular monitoring and oversight, and merit-based personnel decisions — is associated with less corruption.
Other measures that can help reduce corruption are those that facilitate the discovery of what would otherwise remain hidden misconduct. In this regard, legal and institutional protections for whistleblowers, both inside and outside of government, are extremely important. At the very least, potential whistleblowers need to have access to reliable reporting channels that can credibly guarantee confidentiality, and whistleblowers must be protected against retaliation. Some countries, including the United States, have gone further, instituting schemes for paying whistleblowers whose tips lead to significant fines or other monetary recoveries. While systematic evidence on the effectiveness of such reward programs is not yet available, the anecdotal evidence thus far is encouraging.
In addition to protecting and rewarding whistleblowers, governments can help to reduce corruption by promoting transparency more generally. For instance, appropriately designed freedom of information laws, though by themselves insufficient, can facilitate monitoring by outside groups, such as the press and civil society organizations. Speaking of which, while the evidence that democracy reduces corruption is mixed, there does seem to be fairly strong evidence that within democracies, a freer and denser media environment, with more newspapers and radio and TV stations, tends to be associated with lower levels of corruption, and higher probabilities that corrupt incumbent politicians will be punished at the polls. (Perhaps unsurprisingly, there has been a fair bit of optimism in some quarters that modern information technology — especially the internet and social media — will have an even greater corruption-reducing effects, but the impact so far seems relatively modest.) Transparency may be especially important with respect to government management of valuable resources, such as oil and mineral wealth.
These, then, are some of the ingredients for an effective anticorruption strategy. Strong laws enforced by effective and impartial prosecutors and courts; a professional Weberian bureaucracy subject to regular oversight but insulated from partisan politicians; and measures that promote transparency and facilitate monitoring and accountability. This list is incomplete, of course, and there is still a lot that we do not understand, but the basic components sketched above provide the appropriate foundation for an effective anticorruption framework.
III
If we know, at least in broad terms, a fair amount about the kinds of tools and techniques that are effective in reducing corruption, why does corruption remain such a systematic and pervasive scourge in so much of the world?
Though part of the problem may be a lack of capacity, the fundamental problem is political rather than technical. On the one hand, those with the greatest ability to change a corrupt system — those in positions of political and economic power — have the weakest incentives to do so. After all, these political and economic elites are, almost by definition, the winners under the prevailing system. On the other hand, those with the strongest incentive to change a corrupt system have the least ability to do so, because — again, almost by definition — these are the people who have been denied access to political and economic power. This is a familiar problem, hardly unique to the fight against corruption. There is an inherent small-c conservative bias built into most political systems, insofar as those who achieve power typically benefit more from the status quo than those who do not. But the problem may be especially acute with respect to endemic corruption, not only because many of the “winners” owe their power and wealth to the corrupt system that they have mastered, but also because they (and their supporters and associates) might be held personally accountable for their misdeeds if that system changes. Convincing the elites in a corrupt system to support genuine and effective anticorruption reform is a bit like trying to convince turkeys to support Thanksgiving.
We know far less about how to overcome this fundamental political challenge than we do about the tools and techniques that, if fully and faithfully implemented, can substantially mitigate corruption. That said, the political hurdles to anticorruption reform are not insurmountable — after all, some countries have surmounted them, at least in part. A look at those (partial) success stories suggests at least three models for overcoming the inherent political resistance to genuine anticorruption reform.
One possibility, which we might call the “wise king” approach, is to centralize power in a strong and far-sighted leader who can push through dramatic and comprehensive anticorruption reforms without much resistance. If this leader is both a person of high integrity (or at least someone who wants to have that reputation) and sufficiently secure in his or her power, then the leader may be willing to take drastic action to transform a corrupt system. Such a leader would gain much more — in reputation, authority, and legacy — from fighting corruption than from letting corruption persist. Former Singaporean Prime Minister Lee Kwan Yew, who for all his faults deserves credit for the clean-up of Singapore in the 1950s and 1960s, is the most well-known modern example of this model in action. President Xi Jinping of China, who has made anticorruption a central theme of his presidency, seems to be trying to follow in Lee’s footsteps, though his methods are unacceptable to supporters of a liberal order. Yet this approach is not limited to autocratic countries. Some democratic systems concentrate substantial power in the chief executive, giving that person the ability to mandate sweeping anticorruption reforms that might be difficult or impossible in a system with more dispersed power and greater checks and balances. (Mikhail Saakashvili, who became president of the Republic of Georgia in 2004 after the Rose Revolution, is one of the most well-known examples.) And many a populist politician has campaigned on the pledge that, if the voters just give him power — and allow him to sideline or co-opt other institutions that might obstruct his initiatives — then he will clean up the system from top to bottom.
The “wise king” model is understandably attractive to those who are frustrated with the ability of corrupt elites to work the system to block anticorruption reforms. But it is an awfully high-risk strategy. Monarchy under a wise king may be among the best systems of government (at least if one cares only about outcomes rather than process), but monarchy under a bad king, or a mad king, is catastrophic. And the track record of powerful chief executives who have pledged to clean up corruption is not great. Many of the populists who emphasized the fight against corruption both as a campaign theme and as a justification for eroding institutional checks on their authority have proven not only ineffective in fighting corruption, but just as corrupt, or worse, than their predecessors. (Jair Bolsonaro in Brazil and Viktor Orbán in Hungary are particularly prominent examples, though there are plenty of others.) And when anticorruption policy is driven solely by the chief executive and his or her inner circle, there is a substantially greater risk that the fight against corruption will be weaponized, disproportionately targeting political opponents of the regime. So while it is possible that the political impediments to meaningful anticorruption reform can be overcome by a sufficiently far-sighted and skillful leader, the risks of reliance on a powerful leader likely outweigh the benefits.
A second way in which it may be possible to overcome the political headwinds that make it so hard, under ordinary conditions, to sustain effective action against entrenched corruption is to take advantage of moments of crisis that give outsiders, or previously overmatched internal reformers, the opportunity to effect genuine change. We have a number of historical and modern examples of this model in action. The catalyst for Sweden’s significant efforts to promote clean government over the course of the nineteenth century was Sweden’s military defeat by Russia in 1909, and the resulting fear that Sweden’s very existence was in peril. The widespread view that the country’s disastrous military performance was due in large part to the corruption and dysfunction of the Swedish state (especially the role that nepotism and sale of offices played in placing incompetent people in key positions) inspired Sweden’s liberal reformers and gave them more influence with the monarchy, and produced opportunities for reforms that threatened the interests of the entrenched nobility. A more recent example comes from Indonesia, where the Asian financial crisis of 1997 finally brought down Suharto, the corrupt autocrat who had ruled Indonesia for three decades. Suharto’s fall ushered in the so-called reformasi period, which saw not only the democratization of Indonesia’s government, but also numerous significant anticorruption reforms, including the passage of new laws and the creation of an unusually powerful and independent anticorruption commission, as well as a specialized anticorruption court. When a popular movement manages to topple the government, this may also be a moment when good governance reforms that were previously politically unthinkable become feasible.
Of course, for reformers desperate to take on the corruption that is strangling their economies and undermining their governments, it is not so helpful to be told to wait for a major disruptive event like a disastrous military defeat or a financial crisis or a popular uprising. Still, there is an important lesson here, one that recalls the adage that fortune favors the prepared. Pressing for anticorruption reforms in a politically inhospitable environment can be frustrating, and developing carefully crafted legal or institutional reform proposals can seem pointless when the powers that be have little appetite for doing anything that could threaten the sources of their wealth and privilege. But one never knows when a moment of crisis — and opportunity — will emerge. What may have seemed like fruitless efforts to design and to promote politically infeasible changes can pay enormous dividends when the window of opportunity suddenly opens.
The third model for achieving a transition from endemic corruption to manageable corruption might be termed the Long Slow Slog. Rather than a single transformative moment — driven by a wise and powerful leader or a disruptive crisis — a movement away from endemic corruption can result from the slow accumulation of smaller victories and incremental reforms that gradually squeeze pervasive corruption out of the system and alter the norms of politics in a healthier direction. This process typically involves a combination of top-down and bottom-up efforts, with shifting coalitions of activists, journalists, professional elites, and business interests making common cause with reformist politicians who — out of genuine interest, strategic calculation, or some combination — make cleaning up corruption, or improving government more generally, a high priority. The process can be frustratingly uneven and slow, with periods of progress followed by periods of stagnation or backsliding. And it is not really one political struggle, but rather a series of struggles involving various reforms, not all of which are explicitly or primarily about corruption. Yet the cumulative effect of these reforms, if they are sustained and expanded over a sufficiently long period, can be extraordinary.
There are those who are skeptical that the Long Slow Slog model for fighting systemic corruption can possibly work, given how corrupt systems tend to be self-perpetuating. But in fact this model may be the one that holds the most long-term promise. For what it’s worth, the Long Slow Slog is the model that best captures the transformation that occurred in the United States starting around the end of the Civil War in 1865. As noted above, the United States in the mid-nineteenth century was, in many respects, a developing country, with levels and forms of political corruption not too different from modern-day developing democracies such as India or Brazil. By the start of World War II, corruption in the United States — though still very much a problem — was much less pervasive. And in the post-war decades, though corruption scandals continued to make headlines, matters continued to improve. This was not a quick or easy process — there was certainly no “big bang” moment. True, there were a few periods of particularly intense reform activity, especially the Progressive Era at the dawn of the twentieth century; there were also a handful of especially influential reformist leaders, including Theodore Roosevelt, Woodrow Wilson, and Franklin Roosevelt at the national level and Governors Charles Evans Hughes of New York and Robert La Follette of Wisconsin at the state level. But there was no dramatic moment of radical change comparable to Indonesia’s reformasi period in the late 1990s, nor a single transformative leader comparable to Singapore’s Lee Kwan Yew. Rather, the reform process in the United States was a struggle on many fronts, one that was spread out over at least three generations.
Take, as one example, the struggle for civil service reform. Civil service reform is of course not only about fighting corruption, but it is an important aspect of the anticorruption agenda. The so-called “spoils system” — a form of what political scientists sometimes call “clientalism” — is arguably corrupt in itself, and indisputably facilitates and encourages other forms of corruption. In the United States, the first serious civil service reform bill was introduced in Congress in December 1865, by Representative Thomas Jenckes of Rhode Island. It went nowhere. President Grant did implement some modest internal reforms to the executive branch during his first term, and he formed a commission to recommend broader changes to the civil service, but the commission’s ambitious recommendations provoked vigorous resistance, and by 1875 the push for federal civil service reform looked dead. But it rebounded, thanks in large part to a revitalized coalition of reformist activists and sympathetic politicians. The reformers’ efforts received an unexpected boost in 1881 from the assassination of President Garfield by a disappointed campaign worker bitter over the fact that he had not received a government job in return for his political services. The political momentum for reform — boosted by the rallying cry that President Garfield had been, in effect, murdered by the spoils system — finally led to the passage of the landmark Pendleton Civil Service Reform Act in January 1883, just over seventeen years since Representative Jenckes had introduced his first bill.
Yet the Pendleton Act, while hugely significant in its creation of a “merit system” for a portion of the federal civil service, was quite limited in its coverage, applying to only about ten percent of federal employees — and even that limited reform provoked a great deal of hostility from politicians and party operatives. Notwithstanding this resistance, the merit system gradually expanded, albeit in fits and starts, over the next several decades, thanks to a combination of ongoing pressure from reformers and the political calculations of successive administrations. By the time Theodore Roosevelt took office in 1901, the merit system covered roughly 46 percent of the federal civilian workforce; by the time he left office in 1909, that figure had risen to roughly 66 percent. And by the time the United States entered World War II in December 1941 — almost exactly seventy-six years since Representative Jenckes had introduced his first civil service reform bill — the merit system covered approximately 90 percent of federal civilian employees, and almost every state had adopted and implemented a comparable system that covered most state employees. The U.S. civil service merit system was certainly not perfect, but if one compares the shameless and corrupt clientalism of the 1840s and 1850s to the much more professional administration that was established a century later, even the most jaded cynic would be hard-pressed to deny the extent or the significance of the progress that had taken place.
This is but one example of something we see in many countries and in many historical periods. Measures to fight corruption, and to improve the integrity and performance of government more broadly, may take a long time. But not only is incremental progress often possible, these incremental changes can also add up to a meaningful transition away from a system that runs on endemic corruption to one in which corruption, though always a serious problem, is at least manageable. That observation — that the Long Slow Slog approach to fighting systemic corruption can be effective — may seem obvious, even banal. But this conclusion contradicts both the widespread fatalism about the impossibility of making significant progress against entrenched corruption and the view that making progress against entrenched corruption requires a “big bang.” And though the conclusion that the fight against entrenched corruption may often require a Long Slow Slog might seem disheartening, the fact that such a strategy can ultimately bear fruit should be cause for optimism. Those who are engaged in the battle against systemic corruption — including both those who are fighting for change within their own countries and communities, and also those who are seeking to support change from outside — can take heart from the fact that although the problem may often seem intractable, a series of small victories, which may not seem by themselves to do much to change the fundamentals of a corrupt system, can add up to something bigger and more transformational.
This is not to say that corruption can ever be defeated. The temptations to use one’s entrusted power for private advantage, or to use one’s wealth and influence to improperly influence public decisions, are just too strong. And even as the “core” forms of corruption, such as bribery and embezzlement, become harder and riskier, those who want to use their power to acquire wealth, or to use their wealth to exercise power, will find other ways to do so — a point emphasized by those who criticize the political finance and lobbying systems in the United States and elsewhere as forms of “legal corruption.” Yet it would be a mistake to be fatalistic or cynical. The battle against corruption may not be a battle in which we will ever declare final victory — this cancer is not one that can be cured — but progress against this chronic disease of the body politic is possible, so long as those engaged in the fight do not lose heart.