Bayesian networks for imputation

Summary.  Bayesian networks are particularly useful for dealing with high dimensional statistical problems. They allow a reduction in the complexity of the phenomenon under study by representing joint relationships between a set of variables through conditional relationships between subsets of these variables. Following Thibaudeau and Winkler we use Bayesian networks for imputing missing values. This method is introduced to deal with the problem of the consistency of imputed values: preservation of statistical relationships between variables ( statistical consistency ) and preservation of logical constraints in data ( logical consistency ). We perform some experiments on a subset of anonymous individual records from the 1991 UK population census.     

Will taxpayers ever like taxes?: Responses to the U.S. Tax Reform Act of 1986

Major policy changes like the 1986 Tax Reform Act (TRA) in the United States provide natural experiments to study attitudinal and behavior responses to law. Surprisingly, public evaluations of the TRA became increasingly negative after passage, while general support for the tax system became more positive in response to the changes. To explain this puzzle, we propose a dynamic model of taxpayer attitudes that is consistent with the observed loose linkage between logically-connected concepts. We briefly review the political context, objectives and impacts of the 1986 Tax Reform Act, and present several studies testing the impact of self-interest, prior beliefs, and social discussion on changes in attitudes and behavior.Results suggest that the complexity of taxes and difficulty in knowing objective impacts weakens the role of self-interest in attitudinal changes, thereby strengthening the role of prior beliefs and discussion. Evaluations of tax laws are most responsive to more politically-oriented influences.     

Estimating rating transition probabilites with missing data

In this article we provide a rigorous treatment of one of the central statistical issues of credit risk management. GivenK-1 rating categories, the rating of a corporate bond over a certain horizon may either stay the same or change to one of the remainingK-2 categories; in addition, it is usually the case that the rating of some bonds is withdrawn during the time interval considered in the analysis. When estimating transition probabilities, we have thus to consider aK-th category, called withdrawal, which contains (partially) missing data. We show how maximum likelihood estimation can be performed in this setup; whereas in discrete time our solution gives rigorous support to a solution often used in applications, in continuous time the maximum likelihood estimator of the transition matrix computed by means of the EM algorithm represents a significant improvement over existing methods.     

Bayesian analysis of elapsed times in continuous‐time Markov chains

The authors consider Bayesian analysis for continuous‐time Markov chain models based on a conditional reference prior. For such models, inference of the elapsed time between chain observations depends heavily on the rate of decay of the prior as the elapsed time increases. Moreover, improper priors on the elapsed time may lead to improper posterior distributions. In addition, an infinitesimal rate matrix also characterizes this class of models. Experts often have good prior knowledge about the parameters of this matrix. The authors show that the use of a proper prior for the rate matrix parameters together with the conditional reference prior for the elapsed time yields a proper posterior distribution. The authors also demonstrate that, when compared to analyses based on priors previously proposed in the literature, a Bayesian analysis on the elapsed time based on the conditional reference prior possesses better frequentist properties. The type of prior thus represents a better default prior choice for estimation software.     

An evaluation of the childhood family structure measures from the sixth wave of the British Household Panel Survey

Summary.  The paper performs an evaluation of the data that were collected in the sixth wave of the British Household Panel Survey (BHPS) on childhood family structure. After comparing such data with a large number of studies by using external sources, we find that the BHPS data overestimate the proportion of people who report an experience of life in a non-intact family during childhood by about 10%. Although an explanation based on recall error that deteriorates with the age of the BHPS respondents is possible, the overestimation is likely to be accounted for by non-ignorable attrition that may affect most of the comparison studies based on longitudinal data. Conversely, comparisons with other independent measurements from the BHPS itself reveal that the wave 6 data underestimate the proportion of young people who have lived at least part of their childhood in a non-intact family by about 8%. The probability of disagreement between these two sets of measures is strongly associated with poor interview characteristics, which may affect the comparison measure more than the wave 6 measure. Despite such differences, there is therefore a substantial degree of similarity between the family structure information that was collected in the sixth wave of the BHPS and the host of highly diverse records against which it has been compared.     

ECONOMIC DETERMINANTS OF THE TIMING OF PREFERENTIAL TRADE AGREEMENT FORMATIONS AND ENLARGEMENTS

Preferential trade agreements (PTAs) have proliferated over the past 60 years. While a small number of recent studies have examined empirically the economic determinants of the likelihood of a pair of countries having a PTA, this study explains empirically the timing of all PTA formations and enlargements from 1950 through 2006 using duration analysis. Our main and novel goal is to predict (in‐ and out‐of‐sample) a substantive share of these 1,560 PTA events using a parsimonious model with mainly economic variables, taking selection dynamics into account. Our analysis reveals that we can predict correctly in‐sample the actual year of entry into force for 26% of the 1,560 bilateral PTA formations/enlargements in the period 1950–2006 among 10,518 pairings of 146 countries using only a few economic and political variables. Moreover, we can predict correctly in‐sample 57% of these PTA events within a 10‐year window leading up to the event using this model. The model also performs well out‐of‐sample for the near term (82%), but not if the out‐of‐sample period is very long. We conclude with an evaluation of the model's ability to predict the timing of the North American Free Trade Agreement, the European Union's formation and enlargements, and the model's ten most likely post‐2006 PTA events. (JEL F14, F15)