Multilevel or mixed effects models are commonly applied to hierarchical data. The level 2 residuals, which are otherwise known as random effects, are often of both substantive and diagnostic interest. Substantively, they are frequently used for institutional comparisons or rankings. Diagnostically, they are used to assess the model assumptions at the group level. Inference on the level 2 residuals, however, typically does not account for "data snooping", i.e. for the harmful effects of carrying out a multitude of hypothesis tests at the same time. We provide a very general framework that encompasses both of the following inference problems: inference on the "absolute" level 2 residuals to determine which are significantly different from 0, and inference on any prespecified number of pairwise comparisons. Thus, the user has the choice of testing the comparisons of interest. As our methods are flexible with respect to the estimation method that is invoked, the user may choose the desired estimation method accordingly. We demonstrate the methods with the London education authority data, the wafer data and the National Educational Longitudinal Study data.
This paper analyzes the relationship between the diffusion of new technologies and the decentralization of firms. Centralized control relies on the information of the principal, which we equate with publicly available information. Decentralized control, on the other hand, delegates authority to a manager with superior information. However, the manager can use her informational advantage to make choices that are not in the best interest of the principal. As the available public information about the specific technology increases, the trade-off shifts in favor of centralization. We show that firms closer to the technological frontier, firms in more heterogeneous environments and younger firms are more likely to choose decentralization. Using three datasets of French and British firms in the 1990s, we report robust correlations consistent with these predictions.
The impact of aging on healthcare expenditure (HCE) has been at the center of a prolonged debate. This paper purports to shed light on several issues of this debate by presenting new evidence on the "red herring" hypothesis advanced by Zweifel, Felder and Meier (1999). This hypothesis amounts to distinguishing a mortality from a morbidity component in healthcare expenditure (HCE) and claiming that failure to make this distinction results in excessive estimates of future growth of HCE. A re-estimation based on a much larger data set is performed, using the refined econometric methodology. The main contribution is consistency, however. Rather than treating the mortality component as a residual in forecasting, its dynamics are analyzed in the same detail as that of the morbidity component when predicting the impact of population aging on the future growth of HCE. For the case of Switzerland, it finds this impact to be relatively small regardless of whether or not the mortality component is accounted for, thus qualifying the "red herring" hypothesis.
We investigate empirically how industrialized countries and US states share consumption risk at horizons between 1 and 30 years. US federal states share about 50% of their permanent idiosyncratic risk through cross-state capital income flows. While insurance against transitory fluctuations in output is virtually complete, OECD countries do not share any of their permanent idiosyncratic risk. Our results suggest that purely transaction cost based theories cannot explain the home bias, since the potential welfare gains from insurance against permanent shocks would by far outweigh that of insuring against transitory variation. We conclude that permanent and transitory shocks constitute two qualitatively different kinds of risk and that various forms of endogenous market incompleteness may render permanent shocks a lot harder to insure, in particular at the international level.
The idea to learn about international capital mobility from saving and investment data remains appealing. Our approach is based on VAR methods and overcomes some of the problems associated with saving–investment regressions when the data are non-stationary. We propose a new measure of long-run capital mobility that can be easily calculated as a by-product of the estimation procedure of a cointegrated VAR. In an application to historical US and British data, we find long-run capital mobility to have been remarkably stable over the century whereas variations in the mobility of capital primarily seem to have affected short-run capital flows.
Informal evidence suggests that individuals are willing to pay only a finite and, typically, very low price for a specific lottery that converges to an infinite payment with probability one. The established decision theories (expected value, expected utility theory, cumulative prospect theory) cannot satisfactorily explain this low willingness to pay. The presented paradox strengthens the original and the super St. Petersburg paradox.
We study the employment and distributional effects of regulating (reducing) working time in a general equilibrium model with search-matching frictions. Job creation entails fixed costs, but existing jobs are subject to diminishing returns. We characterize the equilibrium in the de-regulated economy where firms and individual workers freely negotiate wages and hours. Then, we consider the effects of a legislation restricting the maximum working time, while we let wages respond endogenously. Employment effects are sensitive to the representation of preferences. In our benchmark, small reductions in working time, starting from the laissez-faire equilibrium solution, always result in a small increase in the equilibrium employment, while larger reductions reduce employment. The regulation benefits workers, both unemployed and employed (even if wages decrease and even in cases where employment falls), but reduces profits and output.
We introduce non-homothetic preferences into an innovation-based growth model and study how income and wealth inequality affect economic growth. We identify a (positive) price effect -- where increasing inequality allows innovators to charge higher prices and (negative) market-size effects -- with higher inequality implying smaller markets for new goods and/or a slower transition of new goods into mass markets. It turns out that price effects dominate market-size effects. We also show that a redistribution from the poor to the rich may be Pareto improving for low levels of inequality.
We analyze a macroeconomic model of monopolistic competition in which consumers earn unequal incomes. When preferences are nonhomothetic, the distribution of income affects equilibrium markups and equilibrium product diversity.
I give necessary and sufficient conditions on the payoff set that guarantee uniqueness of the equilibrium in the Rubinstein bargaining model. The conditions encompass a class of non-convex or disconnected payoff sets with discontinuous Pareto frontiers. Roughly speaking, the equilibrium is unique if the objective function of the corresponding Nash-bargaining game has a unique maximum. I extend the analysis to games where the time between offers is not constant.
