A Researcher Turned Politicians Faces Into Economic Data Asking Whether Their Obesity Could Serve As An Unlikely Warning Sign For Something Governments Rarely Reveal Corruption

. Measuring it across countries is therefore notoriously difficult, often requiring surveys, expert assessments and institutional indicators.Pavlo Blavatskyy, an economist at Montpellier Business School in France, approached the problem from a decidedly unusual direction: could the physical appearance of a country's political elite contain a statistical clue about corruption?His 2021 study, Obesity of Politicians and Corruption in Post-Soviet Countries, examined cabinet ministers across 15 former Soviet republics and found a strong association between the estimated body mass of politicians and established measures of corruption. The work was later recognised with the 2021 Ig Nobel Economics Prize.The premise sounds almost like a joke. The research was not.From photographs to body massBlavatskyy used frontal photographs of cabinet ministers and applied a computer-vision method developed in earlier research to estimate their body-mass index, or BMI, from facial features.The study covered ministers from Armenia, Azerbaijan, Belarus, Estonia, Georgia, Kazakhstan, Kyrgyzstan, Latvia, Lithuania, Moldova, Russia, Tajikistan, Turkmenistan, Ukraine and Uzbekistan. The published paper reports 299 frontal face images representing 2,017 cabinet ministers across those countries.The researcher then calculated the median estimated BMI of politicians in each country and compared those figures with three established measures: Transparency International's Corruption Perceptions Index, the World Bank's Control of Corruption indicator and the Index of Public Integrity. According to the study, the relationship was sufficiently strong to suggest that politicians' physical characteristics could potentially act as a proxy for corruption where conventional data are unavailable, such as at very local levels.Why would obesity have anything to do with corruption?Blavatskyy proposed a possible behavioural explanation involving the social environment surrounding political power. His paper discusses the idea that corruption and lobbying can involve lavish meals and excessive consumption of food and drink. If corrupt officials participate disproportionately in such settings, their physical condition could, theoretically, become associated with the wider corruption environment.But this is a hypothesis, not evidence that eating more makes someone corrupt. The study does not establish that an overweight politician is personally corrupt. Nor does it show that obesity causes corruption. Its finding concerns a statistical relationship between the median estimated BMI of political elites and broader country-level corruption indicators.The researcher tested the idea againThe unusual hypothesis did not end with the cross-country comparison.In a follow-up paper published in 2021, Blavatskyy examined 469 frontal photographs of Ukrainian cabinet ministers who served between 2000 and 2020. Again, computer vision was used to estimate BMI from photographs. He then compared changes in the estimated median BMI of ministers over time with corruption-related measures.The Ukrainian analysis found that the estimated BMI series moved over time with the World Bank's Control of Corruption measure. Blavatskyy also compared it with a curious proxy involving discrepancies between Swiss exports of luxury wristwatches to Ukraine and Ukraine's reported imports of those watches.He acknowledged that simple correlation has limitations, including the possibility of time lags: changes in political behaviour would not necessarily appear immediately in politicians' body weight.But the idea was challengedThe research also attracted methodological criticism. A 2021 comment by György Márk Kis questioned aspects of the original approach, including whether the selection of 15 former Soviet republics could produce a misleading correlation. Blavatskyy subsequently responded with the Ukraine analysis, arguing that the relationship could also be observed within a single country over time.The evidence supports an intriguing association, not a corruption detector that could reliably identify dishonest officials by looking at their photographs.