Gauging Corruption's Impact

Corruption — a major barrier to growth and development in poor countries — is difficult to define and even more difficult to measure. Creative research methods are shedding new light on its effects.
January 25, 2006
Email this page Print this page

Why are poor countries poor? Researchers have long sought geographic, sociological, political and economic explanations for the unequal distribution of wealth among nations. Professor Ray Fisman believes corruption is a leading factor in the cycle of poverty in the developing world. But because corruption is so difficult to define and measure, little research has been done on its causes, effects and possible solutions.

“One definition of corruption is illegal use of public office for private gain,” says Fisman. “A slightly different version is the use of public office in a manner that runs contrary to accepted practices. The second definition is highly ambiguous: acceptable to whom?” In the United States, some people might view large campaign donations as a bribe, while others would argue that such donations fall into the realm of accepted practices. The definition becomes even murkier when you consider managers who use company funds for their own benefit. Some people call that corruption, but Fisman makes a distinction between ethical violations in the private and public sectors. “I think of Enron and Tyco as cases of corporate misgovernance, which is a first cousin to corruption,” he says. “I think of corruption as something that takes place in the public sphere.”

Since bribes and kickbacks are virtually impossible to track, Fisman has devised creative methods for measuring corruption. In the 1990s, he set out to assess the value of political connections in Suharto’s Indonesia. “How do you measure the extent to which political ties matter for firms?” asks Fisman. “You look at shocks to market value: one minute a firm has valuable political connections, and the next minute it doesn’t. You can look at changes in political power after elections, but the problem is that you don’t get a lot of electoral surprises. So I looked at the value of well-connected companies on the days when Suharto got sick.”

Fisman hypothesized that on days when Suharto’s health was brought into question — for example, by reports that he would be traveling to Germany for a “health checkup” — firms with close ties to him would see their stock prices drop. The study’s findings suggest that political connections accounted for a quarter of the value of well-connected Indonesian firms during the Suharto era. Since the study was published in 2001, other researchers have adapted Fisman’s method, using surprise election outcomes to study corruption in various parts of the world.

In a study of tax evasion in China, Fisman and Shang-Jin Wei documented the smuggling of goods from Hong Kong by looking at the gap between the declared value of goods leaving Hong Kong and those arriving in China. They found that the size of the gap depended on the tax rate for a particular item: a 1 percent increase in the tax rate corresponded to a 3 percent increase in this “evasion gap.” Further, they found that much of this effect could be traced to the relabeling of high-tariff products as low-tariff goods. “If chickens are high-tariff goods and turkeys are low-tariff goods, all your chickens would become turkeys,” says Fisman. “We’ve been able to say something about not just whether illicit activity exists but about the channel through which it takes place.”

One commonly proposed solution to the problem of corruption is raising civil servants’ salaries. “There’s a psychic cost of corruption, and if you’re making more than $20 a month, maybe you can afford to have a clear conscience,” Fisman says. “But if the society is fairly habituated to corruption so that there’s no social sanction or self-immolation, raising salaries won’t necessarily remove the temptation to accept bribes. In this case, higher salaries would have to be bundled with enforcement in order to be effective.”

The impact of such a solution is difficult to predict. “The challenge is to come up with a clever way of looking at the effects of wages on bribe taking,” says Fisman. “For example, you could double policemen’s wages in one town and not another, and then drive around the two places and see how many times you get pulled over and asked for a bribe. You need a roundabout way of making an assessment.”

Some people have argued that corruption is not always detrimental to development. The Asian Tigers, for example, have grown rapidly for decades, crony capitalism notwithstanding. More research is needed to determine what types of corruption are more or less damaging than others.

“If there’s a list of prices and everyone knows how much it costs to get something done, that’s probably less damaging than a case where corruption generates uncertainty,” says Fisman. “We don’t know a lot about where corruption is most damaging. Much of Southeast Asia and East Asia has high levels of corruption, but in spite of that, those countries are generally economic success stories, and they have respectable distributions of wealth.”

The greatest challenge for researchers in the area of corruption is to develop practical methods for measuring it and assessing its impact. “This whole literature is in its infancy,” Fisman says. “There’s a need to build on the methodology to answer the questions we care about: What are the effects of corruption and what can we do about it?”


Ray Fisman is associate professor of finance and economics at Columbia Business School.

Raymond Fisman

Raymond Fisman was a Columbia Business School faculty member from 1999 to 2015.

Read the Research

Raymond Fisman, Shang-Jin Wei

"Tax Rates and Tax Evasion: Evidence from 'Missing Imports' in China"


Download PDF View abstract/citation