Heterogeneous determinants of profitability among household-based agri-food processing enterprises in Malawi: evidence from quantile regression

In developing economies, household-based agri-food processing enterprises play an important role in rural income generation, job creation and agricultural value addition. However, many of these enterprises experience persistent profitability challenges despite operating within similar market environments. Existing studies mainly use average-effect approaches to examine enterprise performance, potentially overlooking important differences in how profitability is influenced by productive resources across enterprises with different performance levels. The study addresses this limitation by investigating heterogeneous drivers of profitability among household-based agri-food processing enterprises in Malawi. The study combines survey-weighted quantile regression as the principal empirical analysis, survey-weighted ordinary least squares (OLS) regression as a benchmark, and variance inflation factor diagnostics, coefficient equality tests across quantiles, extended quantile estimation and winsorized re-estimation for robustness assessment, to successfully examine whether the effects of enterprise characteristics and productive inputs vary across the profitability distribution. The findings reveal substantial heterogeneity in enterprise profitability. Survey-weighted Quantile regression results show that the determinants of profitability differ across enterprise performance levels. By going beyond average effects, the study reveals the importance of distribution-sensitive approaches for understanding enterprise performance in resource-constrained agricultural economies. The findings demonstrate that policies based solely on average effects may overlook important differences among enterprises.