Regional Dollarization Differences and Determinants: The Case of Turkey

Abstract

This study investigates the determinants of deposit dollarization within 81 cities of Turkey for the period of 2007:Q4 – 2019:Q1. Within this scope, exchange rate, CDS premiums, city inflation, city export and import are used to model deposit dollarization. As an econometric approach both fixed and random effect panel data methods are implemented using relevant libraries in Python programming language to measure the impact of independent variables. As a result, all independent variables have statistically significant and positive effect on deposit dollarization within cities. When evaluated according to the effect sizes, inflation, import, export, exchange rate and CDS premiums are considered to be determinant, respectively, on deposit dollarization in Turkish cities. On the other hand, when the two models are tested with the Hausman test, which has a null hypothesis “random effect estimator is true”, it is concluded that the random effect model is more suitable for our model. In addition, to see if citizens living abroad that have registered in the cities have an impact on the dollarization, Pearson correlation is conducted and it is found that there is relatively strong correlation between the number of expatriate citizens and deposit dollarization in the city.