Working Paper Series: Special Edition of 2016 to 2018 Interns

5.0 Results and Analysis The correlation results indicate a positive relationship for all the independent variables except for government expenditure, consistent with the initial prediction in the model (see table 4 in Appendix). The results also provide strong evidence of a strong correlation between private sector credit to GDP and per capita GDP (0.5941), gross fixed capital formation and per capita GDP (0.5660). However, there was a weaker correlation between point-of-sales terminals and per capita GDP (0.0601); and card transaction and per capita GDP (0.2448). Importantly, given that correlation does not necessarily provide evidence of causation, a regression was done. 5.1 Estimation Results In the study, the models were estimated using the Pooled Ordinary Least Squares (OLS) Method. The summarized regression results are presented in the Table 1. Four models were estimated: • Model 1 : lGDP_PC it = β 0 + β 1 CTR it + β 2 GCF it + β 3 TO it + β 4 GEXP it + µ it (4) • Model 2 : lGDP_PC it = β 0 + β 1 lPOST it + β 2 GCF it + β 3 TO it + β 4 GEXP it + µ it (5) • Model 3: lGDP_PC it = β 0 +β 1 CTR it +β 2 GCF it +β 3 TO it +β 4 GEXP it +β 5 PSC it + µ it (6) • Model 4: lGDP_PC it = β 0 +β 1 lPOST it +β 2 GCF it +β 3 TO it +β 4 GEXP it +β 5 PSC it + µ it (7) The results on card transaction in the first regression (Model 1) are consistent with the prior expectations of the model. The coefficient was positive and significant at the 5.0 per cent level, with a point estimate of 0.027. The significance of the coefficient suggests that a 1.0 per cent increase in card transaction would lead to a 0.0271 per cent increase in per capita GDP on average, holding everything else constant. As more cards are supplied and more merchants accept cards, transaction volume increases. That is because consumers feel more comfortable using their cards for a larger proportion of their overall transactions. At the same time, merchants would desire access to the increasing pool of cardholders with guaranteed payment. In other words, an electronic payment system produces a multiplier effect that can result in significant increases in consumption. However, in model 3 when private sector credit was added to the model, card transactions became insignificant. A possible explanation is that private sector credit and innovation are joint measures of financial deepening however, private sector credit has a larger contribution, therefore, a greater growth effect.

71