Estimating Intergenerational Returns to Medical Care: New Evidence from At Risk Newborns

Perhaps its main contribution is in documenting that the marginal returns to medical care in one generation can have considerable long term implications, improving the outcomes of individuals in the first generation, while at the same time bequeathing weaker health stocks to the following generation, at least when considering measures of health at birth.3 The structure of the paper is as follows. [...] 15 birth weight of the second generation, we can write the expected birth weight of the second generation conditional on the child appearing in the second generation in terms of the well known Heckman (1974) selection equation. [...] For example, in the case of birth weight, there is a clear upward shift when crossing from below the cut off (more treated) to above the cut off (less treated) individuals in Figure 3(a), and the point estimate in Table 2 suggests a return of −149 grams on the next generation. [...] While there are a large number of inter generational links covered in these 30 years of data (421,382 in the case of birth weight for the 32+ week sample, when focusing on the optimal bandwidth of 171 grams either side of the cut off, this is reduced to 811 effective observations.12 Nevertheless, despite noisy estimates, in all measures observed in Panel A point estimates are consistent with there. [...] Each point refers to the average in 50 gram bins, with the size of the point representing the number of individuals on which the average is based.