Do higher civil servant salaries limit corruption? Evidence from Hong Kong
Project Description
Does paying civil servants higher salaries prevent them from becoming corrupt? Policymakers have been grappling with this question for at least a thousand years and academic studies are inconclusive. In 1070s China, Chief Minister Wang Anshi increased the salaries of government officials in an effort to prevent embezzlement of state funds, and around 200 years later, the Kublai Khan (Emperor Shizu) did the same. In Hong Kong, Governors MacDonnell and Des Voeux held opposing views. More recently, Singapore’s Prime Minister Lee Kuan Yew emphasised that paying high salaries to government officials is crucial in ensuring probity in government.
In this project, we analyse a hand-collected sample of 41,540 corruption complaints by the public that have led to 2,196 prosecutions and 933 convictions of government officials across 90 government departments and public agencies in Hong Kong during 1974-2019. The project has been conducted with the assistance of the Independent Commission Against Corruption (ICAC) and the Civil Service Bureau, which have provided us with data, some of which is not in the public domain.
We measure civil service corruption and civil service salaries at the government department level. Our analysis allows us to compare across departments at the same point in time, thus keeping market-level and time-varying enforcement and institutional factors constant. We can also control corruption opportunities across different departments. Our research design improves significantly on past international studies that analyse real-world data, which generally use more crude estimates of corruption and salaries, and find inconclusive results.
We find a strong negative relationship between civil service salaries and corruption, which is robust to different estimation approaches. Our results are economically significant. We find that a 10% increase in salary levels leads to a 4.1% reduction in prosecutions, between 7.5% and 19.4% reduction in convictions (three years later), and between 3.4% to 4.9% reduction in aggregate bribe amounts solicited or accepted by perpetrators (after one year).
Unlike the inconclusive results obtained in previous studies that use real-world data, our results using our more granular data are strongly in line with the theoretical and controlled laboratory experimental evidence.
Project Investigator
Professor Aristotelis STOURAITIS (Department of Accountancy, Economics and Finance)
Project Collaborator
International Institute for Management Development (IMD) & IMD World Competitiveness Centre (Lausanne, Switzerland)
