Abstract

The urban sprawl index is typically regarded as the most precise indicator of a city's potential to measure the level of sustainable urban growth. Using panel data, we examine the causes of urban sprawl over a twenty-year period in Abomey-Calavi, Benin, and Guangzhou, China. A panel data set might contain group effects, time effects, or both. This study analyzes the typical least squares approach with fixed effects and random effects for panel data analysis. Multiple specification tests, such as the Breusch--Pagan Lagrange Multiplier and Hausman specification tests, are employed to identify the optimal strategy for least square, fixed effect, or random effect models. Using a random effect model to address this scenario appears to be the most successful method for producing statistically significant results. These results could help city planners find ways to make up for the bad effects of urban sprawl on land use.