In many economic models a central variable of interest is lifetime or permanent income which is not observed in survey data sets and typically proxied by annual income information. To assess the quality of such approximations, we use a unique source of lifetime earnings — the German pension system — and focus on two important issues that have been largely ignored in the existing literature. The first ...
According to cross-sectional studies in sport science literature, decreasing sports activity with increasing age is generally assumed. In this paper, the validity of this assumption is checked by applying more effective methods of analysis, such as longitudinal and cohort sequence analyses. With the help of 20 years' worth of data records,from the German Socio-Economic Panel, the development of sports ...
This article proposes a test for missing at random (MAR). The MAR assumption is shown to be testable given instrumental variables which are independent of response given potential outcomes. A nonparametric testing procedure based on integrated squared distance is proposed. The statistic?s asymptotic distribution under the MAR hypothesis is derived. In particular, our results can be applied to testing ...
In a recent paper, Pratt and Zeckhauser (JPE, 1996) discuss the measure of individuals' willingness to pay (WTP) for the reduction of risks to their lives which should be used for public decisions on risk-reducing projects. They suggest to correct observed WTP for the "dead-anyway" effect, which says that WTP increases with the level of risk to which the individual is exposed - an effect ...
Objectives: As diary, panel, and experience sampling methods become easier to implement, studies of development and aging are adopting more and more intensive study designs. However, if too many measures are included in such designs, interruptions for measurement may constitute a significant burden for participants. We propose the use of feature selection—a data-driven machine learning process—in study ...