A Mathematical Modeling Approach to Household Electricity Consumption: Socioeconomic and Behavioral Evidence from Karachi

Abstract

Household electricity consumption in Karachi has intensified due to rapid urbanization, income inequality, climatic stress  and persistent power supply disruptions. This study employs an integrated mathematical modeling approach to examine the socioeconomic and behavioral determinants of residential electricity demand in Pakistan’s largest metropolitan city. Using household survey data combined with secondary statistics from Karachi Electric and national sources  the analysis applies multiple linear regression and machine learning techniques to identify key drivers of consumption. The results indicate that income, household size, appliance ownership  and seasonal factors significantly increase electricity usage  while education contributes to conservation through improved energy awareness. Importantly behavioral practices particularly cooling habits, lighting routines  and peak hour usage exert a strong independent influence on household electricity demand beyond socioeconomic characteristics. In comparison to classic econometrics, machine learning models can rapidly outperform other techniques when handling complex issues such as multi-faceted and nonlinear behavior between dependent variables. These results illustrate an urgent need for policies that encompass demand-side management and include tariff innovations, behavioral interventions, and energy efficiency measures in order to develop a solution for residents' increasing electricity burden and promote sustainable energy development in Karachi.