ESTIMATING AND PROJECTING PAKISTAN’S POPULATION GROWTH: A LOG - LINEAR REGRESSION MODEL APPROACH WITH CONFIDENCE INTERVALS.

Abstract

This study analyzes the long-term population growth of Pakistan using a log-linear regression model estimated through Ordinary Least Squares (OLS). Annual population data are employed to capture the exponential growth pattern by modeling the natural logarithm of population as a function of time. The estimated time coefficient is positive and highly significant, confirming a persistent growth trend in Pakistan’s population. The model demonstrates an excellent fit, as reflected by a high coefficient of determination () and strong correlation between time and population. Diagnostic statistics, including the F-statistic and t-tests, indicate that the regression model and its parameters are statistically significant at the 1% level. Residual analysis further supports the adequacy of the model assumptions. Based on the estimated growth rate, population forecasts are generated for the period . The projections indicate continued population expansion, posing challenges related to resource allocation and urbanization. The findings highlight the effectiveness of log-linear regression for demographic forecasting. This study provides valuable evidence to support population planning and sustainable development policies in Pakistan.