Parameter orthogonalization and Bayesian inference with many instruments

It is examined how Bayesian inference might proceed in models with many instruments. A new prior specification based on Lancaster’s (1997) parameter orthogonalization is developed. This orthogonalization is shown to guarantee that the statistical problems associated with the first stage coefficients are not carried over in estimating the parameter of interest.     

Characterizing life-cycle wealth distributions using statistical inference and dominance criteria

This paper examines how the distribution of household wealth in Canada varies with age over the life cycle. The wealth distribution is characterized in terms of decile means and decile shares for each of six age groups, and comparisons between age-specific distributions are based on first-and second-order stochastic dominance criteria. It is found that (i) mean wealth levels and wealth distributions increase significantly with age in concave quadratic fashion until near-retirement and then decline, and (ii) wealth inequality declines in convex fashion with age, at first steeply and then not significantly. This joint pattern in mean and inequality of wealth holdings across age groups presents a challenge for basic theories to explain.     

Interaction effects in econometrics

We provide practical advice for applied economists regarding robust specification and interpretation of linear regression models with interaction terms. We replicate a number of prominently published results using interaction effects and examine if they are robust to reasonable specification permutations.     

On statistical inference for inequality measures calculated from complex survey data

We examine inference for Generalized Entropy and Atkinson inequality measures with complex survey data, using Wald statistics with variance?Ccovariance matrices estimated from a linearization approximation method. Testing the equality of two or more inequality measures, including sub-group decomposition indices and group shares, are covered. We illustrate with Indian data from three surveys, examining pre-school children??s height, an anthropometric measure that can indicate long-term malnutrition. Sampling involved an urban/rural stratification with clustering before selection of households. We compare the linearization complex survey outcomes with those from an incorrect independently and identically distributed (iid) assumption and a bootstrap that accounts for the survey design. For our samples, the results from the easy to implement linearization method and the more computationally burdensome bootstrap are typically quite similar. This finding is of interest to applied researchers, as bootstrapping is currently the method that is most commonly used for undertaking statistical inference in this literature.     

金融计量经济学:成就和挑战

本文通过对近 2 0年来金融计量经济学主要发展成就的回顾 ,介绍了计量经济学研究的热点领域和有待解决的问题。本文认为 ,ARCH模型以及在此基础上发展起来的其他异方差模型、GMM以及与之相关的参数估计方法的出现以及在金融经济学中成功的运用 ,是金融计量经济学最重要最基本的成就 ;数据及处理方法的独特性使金融计量经济学相对独立于传统计量经济学而发展。     

Presidential Address: Mathematics in economics and econometrics

Abstract The paper discusses the choices of mathematical approaches in economics and econometrics, in particular, approaches that either (a) aim for a sharp result or (b) use the least restrictive assumptions. It is argued that, while the choice (a) often necessitates strong mathematical assumptions, choice (b) may lead to only partial identification and may require using less familiar mathematical techniques. This is discussed in the context of the problem of defining a probability density: existence may fail in function spaces; even after imposing assumptions that ensure existence, the problem is not well posed. A density function may not exist for economic variables as a consequence of institutional rigidity such as an income supplement. The apparatus of generalized functions provides the general solution to identification and well‐posedness of density, but at the cost of less sharp results and greater mathematical complexity.     

Targets,instruments and generalized inverses

The purpose of this paper is to demonstrate the usefulness of generalized inverses for problems associated with macro economic policy. Previous research on the selection of targets and instruments has specified the problem so that both the values of the instruments and the values realized for the targets are a joint outcome of the constrained optimization of an arbitrarily defined welfare function. Two theorems are developed which suggest that: (1) if the number of instruments exceeds the number of targets, a welfare function need not be defined. It is possible to discriminate between policies with a minimum norm generalized inverse which is based on the values of the instruments: (2) given the number of targets exceeds that of instruments the generalized inverse approach reduces to a special case of Theil's certainty equivalence method. Klein's six equation model from ‘Economic Fluctuations in the United States’ is used to illustrate these approaches.     

Some remarks on the undersized sample problem in econometrics

This paper discusses three approaches to the problem of undersized samples: the estimable function, the generalized inverse and the principal component method. It turns out that the k-class estimation methods amount to ordinary least squares if use is made of estimable function or generalized inverses. The generalized inverse approach is a special case of the estimable function approach and of the principal component solution.     

Endogeneity,exogeneity and economic policy empirical evidence on the econometrics of new classical macroeconomics

Zusammenfassung Die Neue Klassische Makroökonomik provozierte nicht nur die ökonomische Theorie, sondern brachte auch eine neue Herausforderung für die ökonometrische Methodologie im Zusammenhang mit der Parameterschätzung unter der Hypothese der rationalen Erwartungsbildung und unter dem Einfluß von Reaktionen auf Maßnahmen der Wirtschaftspolitik, vor allem aber wurde eine neue Diskussion über Kausalität, Endogenität, Exogenität und die Grenzen der traditionellen ökonometrischen Strukturmodelle initiiert.Anhand von Wiener-Granger-Kausalitätstests wird die Grenze der Aussagefähigkeit dieser Testverfahren zur Identifikation von Abhängigkeiten zwischen ökonomischen Variablen diskutiert. Mittels des datenanalytischen Verfahrens der Hauptkomponentenanalyse wird die Möglichkeit einer altenativen Parametrisierung einer ökonomischen Datenbasis demonstriert, aus der ein stochastisches Maß für die Endogenität bzw. Exogenität einer Variablen bezüglich eines Datensatzes entwickelt wird.     

Forecast accuracy and uncertainty in applied econometrics: a recommendation of specific-to-general predictor selection

This article provides empirical evidence on the performance of general-to-specific and specific-to-general predictor selection in linear autoregressions. A pseudo ex-ante forecasting comparison is conducted for a set of 494 macroeconomic time series as collected recently by Dees et al. (J Appl Econom 22:22–38, 2007). In total 10,374 time series realizations are contrasted against competing predictions founding on the two search principles. Overall, specific-to-general modeling is characterized by superior performance in terms of smaller absolute forecast errors and better matching of nominal and empirical coverage levels of ex-ante prediction intervals.     

Reaggregation and firm-level inference in multiplant technologies

We examine conditions under which plant-level data can be used to make firm-level inferences about technology, supply, and demand. Global conditions for such reaggregation involve nonjointness restrictions in both the plant and product dimensions. In the neighborhood of the firm’s fixed overheads, however, restrictions may be eased by appealing to a multiplant generalization of Kohli’s notion of almost nonjointness, which we term “almost reaggregation.” While global conditions for almost reaggregation are the same as for full reaggregation, the local conditions outlined are more easily satisfied. Analysts thus are less likely to commit reaggregation bias when firm-level overheads are constant, and can take advantage of this conclusion through judicious model design.      

Statistical inference in cross-country demand systems

Consumption data on eight goods in 15 countries are used for testing and estimating a system of demand equations across countries. The estimated income elasticities are compared with those of Clements et al. (1979) and Houthakker (1957).