Switching Regression Estimates of the Intergenerational Persistence of Consumption

The influential economic theory of intergenerational transfers predicts a negative connection between credit constraints and intergenerational mobility of consumption. Existing work has used bequest receipt to signal a parent's access to credit markets when investing in his children's human capital. Yet measurement error in bequest receipt generates misclassification error and thus attenuation bias. Employing switching regressions with imperfect sample separation to deal with this error, we show that the intergenerational persistence of consumption in the U.S. for credit constrained families is much higher than for unconstrained families, contrary to what the theory implies. This means children from constrained families are more likely to have consumption levels similar to those of their parents than children from unconstrained families. Our results are robust to the choice of bequest variables and other predictive variables in the switching equation