The determinants of access to credit for Small and Medium Enterprises (SMEs): Evidence from Uzbekistan

This study examines the determinants of access to bank financing for enterprises in Uzbekistan, addressing both the decision to apply for credit and the probability of approval conditional on applying. Using World Bank Enterprise Survey data (n=1,008 enterprises), we employ a twostage analytical framework: binary logit regression models examine factors associated with having an existing loan, and Heckman probit selection models jointly estimate the loan application decision and approval probability, accounting for potential selection bias. The study reveals severe credit rationing in Uzbekistan, with only 13.3% of enterprises holding bank loans and 10.1% applying for new credit. The most striking finding is the dominant effect of existing banking relationships: enterprises with current loans achieve 87.0% approval rates compared to 41.7% for first-time applicants. The Heckman outcome equation confirms this relationship banking effect, representing approximately 30-35 percentage point higher approval probability. Medium-sized enterprises enjoy substantial advantages in both application propensity and approval probability. Export activity and checking account ownership significantly enhance credit access. Contrary to international evidence, female-managed enterprises show positive approval coefficients, though statistical significance is marginal. The highly significant selection parameter confirms substantial selection bias, validating the Heckman approach.