After two decades of global anti-corruption efforts, it is quite clear that deterring and tackling corruption is a key challenge for any country. Throughout, scholars and practitioners of corruption have broadly agreed on a working definition of corruption – the misuse of public office for private gain. Scholars and practitioners have also agreed on its nature and prevalence, as well as on the dramatic economic, political and social costs it imposes in both the public and private sectors. Yet, the measurement of corruption itself remains a major challenge. Corruption is said to be an unobservable phenomenon due to its informal and hidden nature, and often has diverse meanings in different cultural and historical contexts.
How accurate and reliable are existing – and commonly used – measures of corruption?
Since the 1990s, anti-corruption research has widely used perception-based corruption measures. Transparency International’s Corruption Perception Index (CPI) and the World Bank Governance Indicators (WGI) are surely the most well-known indices based on surveyed respondents’ perceptions, attitudes, and opinions towards corruption. Despite the widespread use, the reliability of the indicators have been seriously questioned. CPI and WGI have been criticized for not measuring the actual level of corruption in a given country.
The more recent trend is to rely less on perception and focus on experience-based measures like the World Bank’s Enterprise Surveys (WBES) and the International Crime Victims Survey (ICVS). Individuals and firms are asked whether they have engaged in illicit activity – the supply side of bribery. Despite advantages in identifying sector specific corruption, the accuracy and reliability of the data depends heavily on the formulation of the survey questions and the respondents’ willingness to be forthright. An increasingly common empirical approach for measuring corruption is based on comparing primary and secondary sources of data that suggest corrupt behavior. For instance, measuring leakage of funds between the central government of Uganda and local school districts by comparing reported and received budget data. Nevertheless, some gaps are due to bureaucratic accounting errors or passive waste, rather than direct illicit behavior – calling this method’s accuracy into question as well.
Despite significant advances in the measurement of corruption over the last two decades, it remains necessary to combine multiple measurements and sources of data. To understand and combat illicit behavior empirically, a growing body of evidence shows it is essential to bring insights from the supply side of corruption. Businesses operating in corrupt environments have a clear incentive to report solicitations and payments of bribes – as identification is the first step towards overhauling regulatory inconsistencies that engender corruption. Providing accurate data can help assess collective action opportunities within business communities and allow for comprehensive analysis of perverse incentives; perception alone simply does not capture the scope or scale of corruption.
Elisabeth de Vega Alavedra is an Anti-Corruption/Conflict Fellow at CIPE.