Analysis of the determinants of imports in Kenya: An autoregressive distributed lag bounds testing approach

Abstract

This study examined the determinants of aggregate imports in Kenya using time series data from 1980 to 2017. The theoretical underpinning of this study was the benchmark model which postulates that an aggregate import demand equation relates imports demanded by a country to the ratio of import prices to domestic prices and domestic income. The thrust of this model is in consumer theory. The study augmented the model by introducing additional explanatory variables and hypothesized that import price index (IPI), Domestic price index (DPI), real income (RINC), exchange rate (ER) and real foreign reserves (RFR) were determinants of imports in Kenya. The study used the Autoregressive bounds testing approach to test for cointegration. The bounds test of cointegration revealed that there was a long run association among the variables in the import demand function. The short run coefficients indicate that DPI was the main determinant of imports in Kenya with a coefficient of 0.925 indicating that significant reduction in prices of domestically produced goods will lead to a fall in imports holding other factors constant. The coefficient of IPI was negative and significant while real income did not significantly influence the country’s imports. Real foreign reserves and lagged exchange rate also had a significant effect on imports in the short run. The coefficient of the error correction term was -0.652 and was significant at 1 per cent implying that the model will settle at long run equilibrium. The most important determinant of Kenya’s imports in the long run was DPI as was the case in the short run. This was followed by IPI whose coefficient was negative and significant at 1 per cent. Real income and exchange rate do not significantly affect imports in the long run. The estimated ECM was subjected to stability test using CUSUM test, autocorrelation test using the Breusch-Godfrey Serial Correlation LM test and Heteroscedasticity test using Breusch–Pagan–Godfrey test. Therefore, the study concluded that Kenya’s imports are mainly determined by prices of domestic goods, import prices, foreign reserves, and exchange rate. Consequently, the study recommends that policies aimed at managing (or reducing) the level of imports should target price of domestic goods as well as price of imports. Trade restrictions such as tariffs and import quotas that would affect import prices would be effective in managing the level of imports.