PAEDIATRIC NEW HIV INFECTIONS IN TANZANIA: A BOX-JENKINS ARIMA APPROACH

Abstract

Using annual time series data on the number of children (ages 0 – 14) newly infected with HIV in Tanzania from 1990 – 2018, the study predicts the annual number of children who will be newly infected with HIV over the period 2019 – 2030. The paper applies the Box-Jenkins ARIMA methodology. The diagnostic ADF tests show that, N, the series under consideration is an I (2) variable. Based on the AIC, the research presents the ARIMA (0, 2, 1) model as the best model. The inverse root of the MA polynomial further reveals that the presented model is quite stable. Moreso, the residual correlogram of the model indicates that the residuals are not serially correlated; while the normality test shows that the residuals of the model are also normally distributed. This is overwhelming proof that the applied model stable, adequate and suitable for forecasting new pediatric HIV infections in Tanzania. The results of the study indicate that the number of new HIV infections in children in the country will continue to decline dramatically, over the out of sample period. In fact, our model shows that Tanzania could win the war against pediatric HIV in at least 6 years’ time! Yes, it is possible to end the HIV scourge in children in Tanzania.