The relationship between inflation and unemployment in Botswana: analysis of the Phillips curve

Abstract

The study conducts a Phillip’s curve analysis of the relationship between unemployment and inflation in Botswana. In investigating inflation-unemployment nexus, cointegration is tested using autoregressive distributive (ARDL) bounds testing approach on the yearly data from 1980 to 2017. The paper initially investigates a bivariate regression between inflation and unemployment, where the proxies used are change in price levels and output gap respectively. To make the investigation robust and yield relevant results, control variables are added; real GDP, real exchange rate, nominal interest rates, import prices and lastly broad money supply. Post establishing presence of strong cointegration effects between inflation and unemployment, ARDL coefficient diagnostics for cointegration and long run form are used to find the impact of unemployment and other control variables on inflation. Unemployment is found to not have a significant impact on inflation, failure to identify a significant relationship between inflation and unemployment implies that existing high levels of unemployment has little to do with the current low inflation rate experienced in Botswana. Import prices are found to have a significant impact on inflation. The last interesting finding of the study is that depreciation of the Pula against the Rand leads to inflation while interest rates have no significant impact on inflation. This is then perhaps a call for monetary policy authorities to focus more on the crawling peg between Botswana and its trading partners. To increase employment Authorities should consider lowering the crawling peg (depreciating the pula) against its trading partners. While this may lead to a modest increase in inflation, this will have a significant impact since Botswana is a net importer.