You are here: Home / Understanding Corruption Begins with Defining “Corruption”

Understanding Corruption Begins with Defining “Corruption”

Posted by Ben Wheatland at Oct 02, 2015 08:48 AM |
Filed under:
ACRN guest blogger Brigitte Zimmerman provides an overview of the American Political Science Association's annual meeting from September 2015.

For the second year in a row, Transparency International has asked me to review the corruption research presented at the American Political Science Association (APSA)’s Annual Meeting. Political scientists from around the world gathered in San Francisco in early September, and corruption was again a widely discussed topic.

Last year, the common thread I observed across papers and panels was examining the role of citizens in mitigating corruption. This year, a group of scholars that presented at APSA focused on the need to move away from researching “corruption.” Instead, explicitly or implicitly, they urged scholars and policymakers to precisely identify a dimension, sector, process or form of corruption, and then attempt to explain and understand the more precise version of corruption rather than discussing “corruption” generally.

Jennifer Bussell makes this argument in her chapter in a newly released edited volume by Susan Rose-Ackerman and Paul Lagunes. She advocates for a pragmatic approach to typologies of corruption generally and provides an exhaustive dissection of the concepts and measures of corruption employed in the past. She then proposes a practical typology of corruption that focuses on corruption regarding different types of state resources, and uses this typology to identify the parties that directly control and indirectly influence these resources (and therefore have opportunities for corruption). The new typology strikes me as useful, though it is the broader point of the chapter regarding the importance of being more specific and practical in defining corruption that seems most impactful.

Corruption Occurs in Different Branches of Government

Two papers at APSA focused on differences in corruption patterns within different branches of government. In her presentation, Vineeta Yadav argued that predominant cross-national surveys measuring “corruption” do not necessarily offer comparable measures. She illustrated the need to anchor subjects in specific definitions of corruption. She also discussed how differences in the sensitivity of questions about corruption might affect corruption data; citizens may be more hesitant to answer questions about certain forms of corruption or about corruption in certain branches of government. Drawing from surveys about corruption conducted in Brazil and India, she illustrates this issue and shows that the sensitivity of the question varies across context as well as across corruption area. Her work demonstrates the importance of carefully using corruption data in cross-context analysis.

Similarly, Kelly McMann and Jan Teorell presented a paper (co-authored with myself) using the Varieties of Democracy (V-Dem) dataset to examine the relationship between democratic institutions and corruption. The V-Dem data include observations from all countries for over 100 years. One contribution of the paper is to re-examine our understanding of how democracy affects corruption using the new data. However, another strength of V-Dem is that it offers measures of corruption by branch of government: legislative, executive, and judicial. This allows us to more precisely identify causal mechanisms. For example, we find evidence that introducing free and fair elections more dramatically decreases corruption within elected branches of government. There is still an effect of free and fair elections on corruption overall, but it is weaker.

Corruption Is Multi-Dimensional

A set of papers at APSA defined specific dimensions of the corruption space, and suggested mapping specific forms of corruption on these dimensions. Related to the chapter discussed above, Jennifer Bussell presented a paper examining the role of middlemen in corruption. She proposed two dimensions of the corruption space that are theoretically relevant for understanding when middlemen might broker corrupt exchanges. She argues that middlemen are used most when corruption is repetitive and there is a need for matchmaking between the corrupt official and the citizen party (what she calls “high volume” corruption). Conducting a survey experiment among Indian officials at different levels, she shows these dimensions predict the percentage of respondents that allocate a portion of corruption earnings to a middleman.

In a paper on the same panel as Dr. Bussell, I presented a paper on how public officials choose among corruption forms. In the paper, I argue that officials consider two dimensions of corruption when making their choice: the detectability of the corruption and the impact of the corruption on citizens. If corruption is likely to be discovered, they will bear the costs of hiding it. Further, if corruption is likely to be discovered and they believe citizens will receive this information, they will bear the costs of hiding it and they will also avoid forms of corruption citizens are likely to sanction. I find evidence for these expectations in a survey experiment of district officials in Malawi. I also conducted extensive interviews with government officials and anti-corruption stakeholders in Malawi, who validated the salience of the detectability and impact dimensions.

Corruption Permeates Many Government Processes and Functions

Another group of papers at APSA considered how corruption varies depending on the government process in which it is found. Scholars seem to be moving away from examining the relationship between corruption and democratic institutions and towards examining how specific rules and regulations in government processes affect corruption dynamics.

First, Carl Dahlström and Victor Lapuente (together with coauthors Nicholas Charron and Mihály Fazekas) presented a paper regarding meritocratic bureaucracy and corruption in procurement. Using a rich panel dataset documenting the number of bidders in procurement processes, they show that single bidder procurements are less likely when the bureaucracy is seen as meritocratic.

Feng Yang made a similar argument in a paper regarding bureaucratic rotation in China. Observing that rotating bureaucratic officials is believed to be associated with lower levels of corruption, Yang tests two possible mechanisms underlying this relationship. He finds evidence for the following mechanism: bureaucratic rotation shortens the time horizon of officials, making fostering a relationship by spending money in “entertainment and travelling” (Yang’s indicator of petty corruption) less appealing for local businesses. However, this relationship between shorter terms and lower corruption does not hold at the level of the local leader, where such costs are lower with longer-serving mayors.

Collectively, these papers illuminate the frontier of empirical corruption research. Rather than seeking effective anti-corruption approaches, political scientists are seeking effective approaches for mitigating one form of corruption, corruption with certain attributes, or corruption in one branch of government or government procedure.

Brigitte Zimmerman ( is an Assistant Professor in the Department of Public Policy at the University of North Carolina, Chapel Hill. Her work focuses on issues of corruption, accountability, and transparency in developing countries.

Filed under:
Add comment

You can add a comment by filling out the form below. Plain text formatting. Comments are moderated.

Our partner