Prediction of Unemployment Rate in Malaysia Based on Macroeconomic Factors
Keywords:Unemployment Rate, Macroeconomic Factors, LASSO Regression, Data Transformation, Mean Square Error of Prediction
The unemployment rate has become a critical issue, not only in Malaysia but a worldwide phenomenon. Hence, the macroeconomic factors that significantly affect the unemployment rate in Malaysia were investigated in this paper. The data used was obtained from Trading Economics and the Central Bank of Malaysia. At first, the influential points were detected and removed using Cook’s Distance. The correlation and multicollinearity were then tested to investigate the relationships among the variables such as unemployment rate, gross domestic product (GDP) growth rate, inflation rate, foreign direct investment (FDI), population growth rate and exchange rate. The LASSO regression method was applied to determine the significant macroeconomic factors that affect the unemployment rate in Malaysia. Three different LASSO models were formed under different conditions, which included the model without data transformation (Model A), the model with data transformation (Model B) and the model with data transformation except for the population growth rate (Model C). In conclusion, Model A was chosen as the best LASSO model as it has the smallest value of MSE(P) compared to Model B and Model C. The inflation rate, FDI, population growth rate and exchange rate were the significant macroeconomic factors that causing an increment or decrement of the unemployment rate in our country. Therefore, fiscal and monetary policy should be enforced by the government and policymakers to improve the issue of unemployment thus stabilizing the economy of Malaysia.