Search Results

We model school performance as a function of home inputs, school inputs and three levels of instability suffered by children - at home, school, and in their neighborhood. We include these measures of instability in our model under the assumption that a child's school performance will be higher when the circumstances of their lives are relatively stable. We include circumstances of co-residence, where they are living, their parents relationship, and mobility. Of course, the circumstances are largely chosen by parents. Although it is very interesting to consider the impact of stability on performance, to do so, we would need to model the stability itself. Since we are primarily interested in the relationship between age at school entry and subsequent school performance we want to account for as much of the heterogeneity within and across households in factors that also affect school performance. We use our measures of instability in this spirit. Our focus is on the relationship between the age children start formal schooling and their subsequent school performance. One reason to focus on this relationship is that states have a long history of regulating the age at which children enter school. We take advantage of this regulation to predict age of school enrollment that is reasonably orthogonal to the unobserved individual and family background characteristics. To do so, we must assume that each family's choice of state is not determined by compulsory schooling laws. We also take advantage of having observations on multiple children in the same family to estimate how age of school entry affects siblings who were required to enter school at different ages either by virtue of a change in the compulsory school age or because their family moved to a state with a different compulsory school age. We will use the children of the NLSY respondents. We will use performance data on standardized tests administered to each child at age 3, 6, and 9. We model the change in test scores between age 3 and age 6 and between age 6 and age 9. We can hold constant family effects and identify policy effects by looking at children in the same family subject to different policy regimes.

We adopt a life-course perspective to study smoking behavior in Great Britain, Russia, and the United States. Given their different mixes of policies, it is intriguing that the cross-sectional prevalence of smoking in Great Britain and the U.S. is similar, while in Russia men's smoking rates are very high. Our results reveal other similarities and differences that are not apparent in cross-sectional data. For example, we find that the timing of smoking initiation is very similar in most cohorts across the three countries. Another interesting pattern is that the very high smoking prevalence among cohorts of Russian men reflects both high smoking initiation and an almost total lack of smoking cessation. Future research is needed to address a host of questions about the determinants of life-course smoking behavior, including the separate impacts of tobacco control policies on initiation and cessation.

Tobacco control policies range from high excise taxes, to direct restrictions on smoking in public places, to the regulation of pharmaceutical products that aid smoking cessation. In this paper we depart from the standard cross-sectional approach and instead adopt a life course perspective to study the impact of tobacco control policies across countries -- Great Britain, Germany, Russia, and the United States. Of the four countries we study, Great Britain taxes cigarettes most heavily - the price of cigarettes in Great Britain are more than twice the average price in Germany, and are more than six times the average price in the Russian Federation. The U.S. has the most restrictions on smoking in public places and the least restrictions on the sale of smoking cessation products. For example, in the U.S. nicotine patches are allowed to be sold widely over-the-counter, while in Great Britain, Germany and the Russian Federation these products are available only in pharmacies or by prescription. Given the different mixes of tobacco control policies, it is intriguing to note that the prevalence of smoking in Great Britain and the U.S. is fairly similar, while in Germany smoking rates are somewhat higher and in Russia smoking rates are very high for men but much lower for women. According to Corrao et al. (2000), the 1996 British smoking prevalence rate was 29 percent for males and 28 percent for females; the 1997 U.S. smoking prevalence was 28 percent for males and 22 percent for females; the 1997 German smoking prevalence was 43 percent for males and 30 percent for females; and the 1996 Russian smoking prevalence was 63 percent for males and 13 percent for females. While some part of these differences are likely due to differences in cultural norms about smoking, it is likely that the mix of tobacco control policies in each country also plays an important role. In this paper we lay the groundwork for researchers to take advantage of large differences across countries in tobacco control policies. To do so, we first summarize available information on tobacco control policies in force in each country. We then document in several ways the rates of smoking in Germany, Great Britain, Russia, and the U.S. In particular we describe the life course patterns of smoking by men and women in each country over time. Finally, we present preliminary econometric results from a discrete time hazard model of a sample of U.S. women smokers? decisions.

Full-text available on-line at: http://www.diw.de/deutsch/abteilungen/ldm/archiv/ar2002/gsoep2002/abstracts/kenkel_lillard_mathios.pdf