The relation between microfinacing and corruption by country: An analysis of an open source dataset

Abstract

Examining the relation between global microlending and corruption may inform how trust and influence propogate through crowds. Building this understanding may help U.S. Army intelligence officers leverage crowds for humanitarian efforts as well as, to detect signs of adversarial influence. A dataset was created combining open source data from Kiva, a non-profit microfinancing institution, and Transparency International, a global coalition against corruption that publishes an annual Corruption Perceptions Index (CPI). The CPI was merged with Kiva microfinancing variables related to Kiva field partners. A preliminary analysis was conducted on a subset of the data in an effort to determine a near real-time microfinancing proxy for the CPI using the Kiva microfinancing data. Results suggest that when controlling for time on Kiva, the average loan size in dollars, delinquency rate, average loan size per GDP, and average time to fund loan all significantly predict CPI.