Aug 14th 1997
From The Economist print edition
Can governments reduce corruption by paying more to public servants?
CORRUPTION in government is an age-old problem, but a new subject of study for economists concerned with economic growth. Within the past year, organisations such as the World Bank and the IMF have begun to argue that corruption can deter investment and retard growth. Last week, the Fund suspended a $220m loan programme for Kenya, to punish the government for its failure to root out graft. In future, according to a newly adopted IMF policy, its aid may be “suspended or delayed on account of poor governance, if there is reason to believe it could have significant macroeconomic implication”.
How, though, are governments to curb “poor governance”? One common recommendation for poor countries is to increase the pay of civil servants. Clerks and police officers earning $50 a month, the reasoning goes, have no choice but to take bribes in order to feed their families. If governments paid their public officials better, such people would have less need to supplement their incomes illicitly, and more to lose by getting caught.
The fly in this ointment is the assumption that officials are driven to corruption by necessity. They may simply be driven by greed. And the balance between these motives will affect the amount by which their pay needs to be raised in order to make them less corruptible. In a recent paper* two IMF economists, Caroline Van Rijckeghem and Beatrice Weder, try to work through these complexities. One surprising implication of their findings is that while higher pay for public servants may indeed reduce corruption, the benefits may be smaller than the added costs.
Quantifying costs and benefits in this murky domain is difficult, so the researchers have drawn on a multinational corruption index devised by a company called Political Risk Services. The index is based on surveys of business opinion, and tries to assess the relative likelihood in different countries of high government officials demanding back-handers in return for favours, as well as the extortion of smaller bribes by petty officials. Regrettably, the authors refuse to disclose how individual countries rank on the corruption index, lest the IMF be accused of labelling some of its members corrupt.
The economists then rank bureaucrats’ pay in 25 developing countries by the ratio of government wages to manufacturing wages (see chart). This provides a rough estimate of the difference in rewards available to workers in the private and public sectors. And they adjust for national differences in the chances of a corrupt official being caught and punished by using an index of the effectiveness of national legal systems developed by the same risk-assessment firm. In addition, they factor in Political Risk Services’ attempt to measure the relative quality of public administration in the different developing countries.
The results of all this are not entirely reassuring. One encouraging finding is that there does indeed appear to be less corruption in countries where public servants are paid relatively well compared with other workers. The relationship is strong enough, Ms Van Rijckeghem and Ms Weder reckon, to suggest that raising average civil-service pay from 100% to 200% of the manufacturing wage reduces corruption by about one unit in the corruption index. This would be a substantial improvement, equivalent broadly to reducing the amount of corruption in China to that of Hong Kong, or from that in Argentina to the amount in Spain.
And yet this numerical correlation does not prove that governments can merely increase pay to eliminate corruption. For one thing, the inverse relationship between wages and corruption does not always hold; El Salvador, for example, pays its bureaucrats far more than the average factory worker but still ranks towards the bottom of the corruption index.
Moreover, paying civil servants enough to persuade them to reject bribes and kickbacks is itself costly. Ms Van Rijckeghem and Ms Weder estimate that government wages need to be between 2.8 and 7.4 times the wages paid in manufacturing, depending on the other factors in each country, before corruption would become negligible. In some countries, that could mean a three-or fourfold increase in public-secstor pay. Singapore has the best ranking in the corruption index, but it is the only one among the 25 countries in the sample that pays its bureaucrats more than twice the average manufacturing wage.
Establishing a statistical link between low wages and corruption does not show which way causation flows. Extensive corruption might reduce a government’s tax take, limiting the amount it can pay employees. Or perhaps some governments deliberately keep wages low because they take it for granted that customs inspectors and driving-licence clerks will supplement their official pay.
Although it is plausible to suspect that low pay encourages government officials to seek extra income, a strong correlation is hard to draw. A telling finding from the research is that when its authors looked at trends in individual countries over a 12-year period, changes in public-sector wages had no significant effect on the reported extent of corruption.
This may be, the authors say, because relative pay levels are unlikely to have changed enough even over a dozen years to show up on an index of corruption that is broad and subjective to begin with. The study, remember, argues that a doubling of the relative wages of civil servants is needed to reduce a country’s score on the corruption index by a single unit.
Still, there is another reason to question the efficacy of paying officials more. Arguably, the type of corruption that most damages economic growth is not petty bribery but rather the massive theft of state funds by venal politicians and their senior henchmen. It is hard to believe that corruption at this level is much influenced by official pay scales. And the pettier forms of corruption are anyway much harder to control when office juniors see their masters filching vast quantities of other people’s money.
* “Corruption and the Rate of Temptation: Do Low Wages in the Civil Service Cause Corruption?” IMF Working Paper. May 1997.
|OPINION | WORLD | BUSINESS | FINANCE & ECONOMICS | SCIENCE &
PEOPLE | BOOKS & ARTS | MARKETS & DATA | DIVERSIONS | PRINT EDITION
Copyright © The Economist Newspaper Limited 2005. All rights reserved.