Analysis and Forecasting of the Time Series Data on Births and Deaths in Azerbaijan Using ARIMA Model

Forecasts of births and deaths play an important role in determining the dynamics of both population size and gender-age structure.Since population forecasts are the basis of long-term planning of socio-economic development, the statistical accuracy of forecasts is particularly important, and the applied methods play a special role here.The purpose of this study is to evaluate Autoregressive Integrated Moving Average (ARIMA) model ability to forecast the yearly number of births and deaths in Azerbaijan.In the analysis, the Box-Jenkins methodology was followed when building the suggested model.Besides, Akaike's information criterion (AIC) and Bayesian Information Criteria (BIC) are used to select the best ARIMA model, compared to another estimated models.The prediction results of the models are evaluated using the mean absolute percentage error (MAPE) and the root mean square error (RMSE) .Comparing the predicted data from the ARIMA models shows that the correct selection of model parameters, it possible to fairly accurately predict the yearly number of births and deaths.Thus, using the advantages of the ARIMA model, it is possible to obtain forecasts of birth and death rates for the near future and it possible to observe changes that will occur in the age structure of the population.And these interpretations can guide policymakers to focus on socio-economic development and comprehensive healthcare system strengthening as crucial strategies for raising the fertility level and further reducing mortality rate.