Abstract: An inappropriate treatment of heterogeneity can make the estimate of structural parameters inaccurate and uninformative. The present paper reviews how the direct modeling of cross- sectional differences via Generalized Structured Models represents a useful bridge between a statistical approach, where the conditional distribution of the dependent variable returns any value of the outcome given any value of the explanatory variables, and an econometric analysis, where parameters have policy implications. The explicit modeling of heterogeneity across groups improves the quality of the estimates, mitigates their dependence upon the chosen instrumental variable and diminishes the self-selection problem.