We examine the effect of radiation exposure in utero, resulting from nuclear weapon testing in the 1950s and early 1960s, on long-run outcomes of Norwegian children. Exposure to low-dose radiation, specifically during months 3 and 4 in utero, leads to lower IQ scores for men and lower education attainment and earnings among men and women. Children of persons affected in utero also have lower cognitive scores, suggesting a persistent intergenerational effect of the shock to endowments. Given the lack of awareness about nuclear testing in Norway at this time, our estimates are likely unaffected by avoidance behavior or stress effects.

Given the wide use of child care subsidies across countries, it is surprising how little we know about the effect of these subsidies on children's longer-run outcomes. Using a sharp discontinuity in the price of child care in Norway, we are able to isolate the effects of child care subsidies on both parental and student outcomes. We find very small and statistically insignificant effects of child care subsidies on child care utilization and parental labor force participation. Despite this, we find significant positive effect of the subsidies on children's academic performance in junior high school, suggesting that the positive shock to disposable income provided by the subsidies may be helping to improve children's scholastic aptitude.

Using Norwegian data, we examine effects of school starting age (SSA). Unlike much recent literature, we can separate SSA from test age effects using scores from IQ tests taken outside school at about age 18. We find a small, negative effect of starting school older but much larger positive effects of age at test. Also, starting older leads to lower earnings until about age 30. We find little impact of SSA on educational attainment, but boys who start older are less likely to have poor mental health at age 18. Additionally, starting school older has a negative effect on the probability of teenage pregnancy.

We introduce two simple new variants of the jackknife instrumental variables (JIVE) estimator for overidentified linear models and show that they are superior to the existing JIVE estimator, significantly improving on its small-sample-bias properties. We also compare our new estimators to existing Nagar (1959) type estimators. We show that, in models with heteroskedasticity, our estimators have superior properties to both the Nagar estimator and the related B2SLS estimator suggested in Donald and Newey (2001). These theoretical results are verified in a set of Monte Carlo experiments and then applied to estimating the returns to schooling using actual data.