Parallel inference for massive distributed spatial data using low-rank models

Due to rapid data growth, statistical analysis of massive datasets often has to be carried out in a distributed fashion, either because several datasets stored in separate physical locations are all relevant to a given problem, or simply to achieve faster (parallel) computation through a divide-and-conquer scheme. In both cases, the challenge is to obtain valid inference that does not require processing all data at a single central computing node. We show that for a very widely used class of spatial low-rank models, which can be written as a linear combination of spatial basis functions plus a fine-scale-variation component, parallel spatial inference and prediction for massive distributed data can be carried out exactly, meaning that the results are the same as for a traditional, non-distributed analysis. The communication cost of our distributed algorithms does not depend on the number of data points. After extending our results to the spatio-temporal case, we illustrate our methodology by carrying out distributed spatio-temporal particle filtering inference on total precipitable water measured by three different satellite sensor systems.     

Employment and earnings of foreign-born scientists and engineers

Population Research and Policy Review -     

Erosion oder Stabilität der Beruflichkeit?

Regarding the changing patterns of mobility processes in the German labor market, this paper takes up the scientific debate on the declining significance of occupational qualifications. By comparing several labor market entry cohorts, the matching of the learned occupation and the practiced occupation is analyzed in both the horizontal and the vertical dimension of occupational mobility. In doing so, the question of a successful utilization of the skills acquired in the educational process can be addressed. Changes of occupations will be operationalized on the basis of a new professional classification by using the data set “Working and Learning in a Changing World” (ALWA). First, the Kaplan Meier survival curves show that the risks of horizontal occupational mobility have increased across cohorts. This indicates that the formerly close relationship between the education system and the employment system has weakened. Furthermore, according to our multivariate Cox regression analysis, the mobility towards inadequate status positions has increased. Thus, due to the expansion of atypical forms of employment and the increased volatility of the markets, altered mobility patterns of the younger cohorts entering the labor market can be shown. Consequently, the increase of unsuccessful transfers of skills in inter-firm mobility suggests a weakened cohesiveness between professionalism and a growing risk to loose professional qualifications.     

Promotions and the pattern of grocery shopping time

The histograms of interpurchase times for frequently purchased packaged goods have consistently shown pronounced seven-day cycles. Evidence supports that the weekly spike phenomenon is the result of consumers' regular shopping trip schedules. We explore the implications of this peculiar regularity on the issue of consumer purchase timing acceleration. Data for five product categories are examined. Promotions are found to have little effect in accelerating purchase timing. In contrast, conventional interpurchase time models are shown to overstate the effect of promotions.     

Bayesian estimation of P (Y<X) from Burr-type X model containing spurious observations

We consider the problem of estimating R=P(Y<X) when X and Y are independent Burr-type X random variables. We assume that the sample from each population contains one spurious observation. Bayes estimates are derived for exchangeable and identifiable cases. Monte Carlo simulation is carried out to compare the bias and the expected loss of R.     

Child mortality and fertility decline: Does the Barro-Becker model fit the facts?

I compare the predictions of three variants of the altruistic parent model of Barro and Becker for the relationship between child mortality and fertility. In the baseline model fertility choice is continuous, and there is no uncertainty over the number of surviving children. The baseline model is contrasted to an extension with discrete fertility choice and stochastic mortality and a setup with sequential fertility choice. The quantitative predictions of the models are remarkably similar. While in each model the total fertility rate falls as child mortality declines, the number of surviving children increases. The results suggest that factors other than declining infant and child mortality are responsible for the large decline in net reproduction rates observed in industrialized countries over the last century. Financial support by the National Science Foundation (grant SES-0217051) and the UCLA Academic Senate is gratefully acknowledged. I thank Sebnem Kalemli-Oczan, Rodrigo Soares, and two anonymous referees for comments that helped to substantially improve the paper. Olesya Baker and Ilya Berger provided excellent research assistance. Responsible editor: Junsen Zhang.     

The Opportunity Cost of Admission Preferences at Elite Universities*

Objective. This study examines how preferences for different types of applicants for admission to elite universities influence the number and composition of admitted students. Methods. Previous research with these NSCE data employed logistic regression analysis to link information on the admission decision for 124,374 applications to applicants' SAT scores, race, athletic ability, and legacy status, among other variables. Here we use micro simulations to illustrate what the effects might be if one were to withdraw preferences for different student groups. Results. Eliminating affirmative action would substantially reduce the share of African Americans and Hispanics among admitted students. Preferences for athletes and legacies, however, only mildly displace members of minority groups. Conclusions. Elite colleges and universities extend preferences to many types of students, yet affirmative action is the one most surrounded by controversy.     

Child Support Obligations and Low‐Income Fathers

Using the 1994–1998 waves of the Current Population Survey—Child Support Supplement (N =5,387), the aims of this study are to document child support obligation rates of nonresident fathers, to examine the effect of the obligation rate on child support compliance, and to calculate the trade‐off between fathers’ financial responsibility and children's well‐being, paying particular attention to low‐income fathers. The results indicate that low‐income fathers have high child support obligation rates, which significantly reduce their child support compliance. Although lowering the obligation rate for these fathers may improve their compliance, it does not fully offset the lowered obligation amounts and leads to a 30% net payment loss for welfare mothers and a 43% loss for nonwelfare mothers. Policy implications are discussed.     

The predictability of stock returns - a nonparametric approach

This paper reexamines the predictability of stock returns with a nonparametric model. We first identify, through a set of diagnostic tests, five lagged predictive factors from a linear model. Using these factors, we predict one-month-ahead stock index returns with a nonparametric approach. We find that our nonparametricmodel. We first identify, through a set of diagnostic tests, five lagged predictive factors from a linear model. Using these factors, we predict on -month-ahead stock index returns with a nonparametric approach. We find that our nonparametric model can correctly predict about 74% of stock index return signs. With various ex ante trading rules based on nonparametric predictions and transaction cost schedules, we then compare the performance of "managed" portfolios with that of the buy and hold portfolios. We fmd that the managed portfolios are mean-variance dominant over the buy-and-hold strategies when no or low transaction costs are assumed. When high transaction costs are assumed instead, the mean-variance dominance diminishes However,the Sharpe index of risk-adjusted portfolio performanceindicates that the managed portfolios significantly outperform the buy-and-hold strategies even for the high-transaction cost scenario. We show that the difference in performance between the managed portfolios and the buy-and-hold strategies can be partially explained by the January effect or the small firm effect. In sum, this paper demonstrates the merits of using a nonparametric approach for predicting stock returns and testing market efficiency.