Stuart A. Low

An econometric analysis of the bank credit scoring problem

Abstract Most credit assessment models used in practice are based on simple credit scoring functions estimated by discriminant analysis. These functions are designed to distinguish whether or not applicants belong to the population of ‘would be’ defaulters. We suggest that the traditional view that emphasizes default probability is too narrow. Our model of credit assessment focuses on expected earnings. We demonstrate how maximum likelihood estimates of default probabilities can be obtained from a bivariate ‘censored probit’ framework using a ‘choice-based’ sample originally intended for discriminant analysis. The paper concludes with recommendations for combining these default probability estimates with other parameters of the loan earnings process to obtain a more meaningful model of credit assessment.

Tests of rationality, neutrality and market efficiency. A Monte Carlo analysis of alternative test statistics

Abstract This paper examines the small sample properties of three testing strategies used to analyze the rationality, monetary neutrality and market efficiency hypotheses. We focus on the original ‘two-step’ Barro test of the MRE hypothesis formed entirely from OLS results, a test that employs the correct variance-covariance formulae for these ‘two-step’ estimates, and Mishkins FIMLE testing framework. Each test is examined under likely model respecifications. The findings highlight the extensive bias incurred by drawing inferences from simple unadjusted ‘two-step’ estimates and reveal the relative power of all tests in identifying alternatives to the null hypotheses.

A reformulation of the Hausman test for regression models with pooled cross-section-time-series data

Abstract A Hausman test has been typically used to determine the consistency of the GLS estimator in static models with pooled cross-section-time-series data. Based on a GMM approach, we reformulate the Hausman test and find that it incorporates and tests only a limited set of moment restrictions. We also consider an alternative GMM statistic incorporating additional restrictions, which has power toward additional sources of model misspecification. Our Monte Carlo experiments demonstrate that while both the Hausman test and the alternative have good power detecting endogenous regressors, the alternative dominates if coefficients of regressors are nonstationary.

Employment Bonuses and Labor Turnover

The purpose of this paper is to illustrate how a two-part compensation system composed of a rigid base salary and a flexible bonus can reduce turnover. It is shown that bonus pay is an effective retention device if it is risk reducing and is correlated with outside contract offers. For the first time, to the best of our knowledge, a model of bonus payments is tested with U.S. instead of Japanese data. The empirical results are suggestive of the conditions that give bonuses an important, even dominant, role in worker retention.

An Application of the Probit Transformation to Tourism Survey Data

An important question facing the tourism industry is the identification of those factors responsible for generating current and continuing visitor interest in a particular region. This study begins by examining various estimation techniques which may be applied to models in which an individual either possesses an attribute (in this case interest in an area) or does not. The advantages of the probit technique are detailed and an illustrative application of this technique is considered.

Preunemployment Job Search and Advance Job Loss Notice

Preunemployment search is the fundamental labor market process generating beneficial effects of advance notice. Yet theory indicates that workers receiving notice may not search, whereas others may search even without advance notice. Our weighted results indicate that over one-third of all nonnotified workers still search and over 40% of workers receiving notice do not respond by searching. Further, preunemployment search determinants differ for notified (nonnotified) workers and men (women). For notified men, search is strongly increased by longer notice and strongly decreased by higher unemployment insurance benefits. But neither factor affects the employed search decisions of notified women.

Employer Tax Evasion in the Unemployment Insurance Program

We use unique data to analyze employer tax compliance with Unemployment Insurance (UI) provisions. The data indicate that employers may have underreported

How do Unemployment Insurance and Recall Expectations Affect on-the-job Search among Workers Who Receive Advance Notice of Layoff?

728 million of UI taxes nationally in 1987 alone. To formally examine this noncompliance, a theoretical model of payroll tax evasion is developed showing that increasing payroll tax rates, among other things, likely increases noncompliance by risk-neutral firms. This prediction is empirically verified. The finding that UI tax evasion is systematically related to various firm characteristics suggests that UI audits may be effectively targeted by statistical profiles derived from our model, thereby improving compliance.

DECISIONS TO RETAIN ATTORNEYS AND FILE LAWSUITS: AN EXAMINATION OF THE COMPARATIVE NEGLIGENCE RULE IN ACCIDENT LAW

This paper explores how advance notice of layoffs, recall (rehiring) expectations, and unemployment insurance (UI) benefits affected on-the-job search among a random sample of Arizona UI recipients in 1975–76. The analysis, which includes extensive controls for the characteristics of workers and their jobs, indicates that pre-unemployment search increased with length of notice and decreased with expected recall. Also, among workers not expecting recall, pre-unemployment search decreased with the level of UI benefits available after layoff. The authors argue that improved experience rating would encourage firms to give employees advance notice when layoffs are imminent, and re-employment bonuses for workers with zero or short unemployment spells would encourage early search.

Stochastic Earnings Functions, Risk, and the Rate of Return to Schooling

The article examines the impact on litigation behavior of a basic feature of tort reform--replacement of the standard of contributory negligence with comparative negligence. We focus on the differential impact of the two negligence standards on incentives to hire attorneys and litigate claims. The evidence is based on a large sample of auto injury accident claims during a period of considerable cross-state variation in liability rules. We find that, on average, comparative negligence provides stronger incentives to hire attorneys and file lawsuits and is associated with higher dollar awards. The average claimant in a contributory negligence state is predicted to have a 36.0 percent likelihood of obtaining representation compared to 47.3 percent in a comparative state. The joint probabilities of representation and filing are 12.5 percent in contributory negligence states versus 21.2 in comparative states. Other findings support the view that litigation is sometimes used as a variance-increasing strategy.