David W. Harless

The Predictive Utility of Generalized Expected Utility Theories

Recent experimental choice studies compare expected utility with competing theories of decision-making under risk. Formal tests used to judge the theories usually count the number of consistent responses, ignoring systematic variation in inconsistent responses. A maximum-likelihood estimation method is developed that extracts more information from the data and enables one to judge the predictive utility--fit and parsimony--of utility theories. Analyses of twenty-three data sets suggest a menu of theories that sacrifice the least parsimony for the biggest improvement in fit. The menu is mixed fanning, prospect theory, expected utility, and expected value. Copyright 1994 by The Econometric Society.

Nurse staffing and adverse events in hospitalized children

This study determined whether the number of hours of care provided by RNs was related to mortality and complications in hospitalized children. Administrative data (1996-2001) were used to examine discharges of 3.65 million pediatric patients in 286 general and childrens hospitals in California. A greater number of resource-adjusted hours of care provided by RNs was related to significantly reduced occurrences of postoperative pulmonary complications, postoperative pneumonia, and postoperative septicemia; the positive impact of increases in nurse staffing was of greater magnitude at institutions providing fewer resource-adjusted hours of RN care. There was also evidence of an impact of increases in nurse staffing on urinary tract infections, but it was statistically significant only for institutions with higher resource-adjusted hours of RN care. There was no statistically significant relationship between RN staffing and mortality. More hours of care provided by RNs was associated with improved quality of care for hospitalized pediatric patients.

Investor behavior and the persistence of poorly-performing mutual funds

Abstract Recent studies tentatively suggest some mutual funds have near-term, persistent positive performance, but reinforce earlier findings that other funds have long-term, persistent negative performance. In this paper we ask how consistently underperforming mutual funds are able to persist nevertheless. We compare two models: One model incorporates the assumption that investors choose funds on the basis of past risk-adjusted returns. The other model, motivated by the representativeness heuristic, assumes investors react to recent returns without considering the predictive validity of returns. Our analysis of a sample of no-load growth funds supports a model of investor behavior combining two extremes: when choosing among funds investors respond to returns ignoring differences in systematic risk and expenses, but upon choosing a fund they stick fast forsaking the sensitivity to returns displayed in evaluating funds.

Variations in inpatient mortality among hospitals in different system types, 1995 to 2000.

Background:Relatively few studies focused on the impact of system formation and hospital merger on quality, and these studies reported typically little or no quality effect. Objective:To study associations among 5 main types of health systems–centralized, centralized physician/insurance, moderately centralized, decentralized, and independent–and inpatient mortality from acute myocardial infarction (AMI), congestive heart failure, stroke, and pneumonia. Data and Methods:Panel data (1995–2000) were assembled from 11 states and multiple sources: Agency for Healthcare Research and Quality State Inpatient Database, American Hospital Association Annual Surveys, Area Resource File, HMO InterStudy, and the Centers for Medicare and Medicaid Services. We applied a panel study design with fixed effects models using information on variation within hospitals. Results:We found that centralized health systems are associated with lower AMI, congestive heart failure, and pneumonia mortality. Independent hospital systems had better AMI quality outcomes than centralized physician/insurance and moderately centralized health systems. We found no difference in inpatient mortality among system types for the stroke outcome. Thus, for certain types of clinical service lines and patients, hospital system type matters. Research that focuses only on system membership may mask the impact of system type on the quality of care.

Using Minimum Nurse Staffing Regulations to Measure the Relationship Between Nursing and Hospital Quality of Care

This study tests whether changes in licensed nurse staffing led to changes in patient safety, using the natural experiment of 2004 California implementation of minimum staffing ratios. We calculated counts of six patient safety outcomes from California Patient Discharge Data from 2000 through 2006, using the Agency for Healthcare Research and Quality Patient Safety Indicators (PSI) software. For patients experiencing nonmortality-related PSIs, we measured mean lengths of stay. We estimated difference-in-difference equations of changes in PSIs using Poisson models and calculated the marginal impact of nurse staffing on outcomes from fixed-effect Poisson regressions. Licensed nurse staffing increased in the postregulation period, except for hospitals in the highest quartile of preregulation staffing. Growth in registered nurse staffing was associated with improvement for only one PSI and reduced length of stay for one PSI. Higher registered nurse staffing per patient day had a limited impact on adverse events in California hospitals.

Delivery system characteristics and their association with quality and costs of care: implications for accountable care organizations.

Background: Implementation of accountable care organizations (ACOs) is currently underway, but there is limited empirical evidence on the merits of the ACO model. Purpose: The aim was to study the associations between delivery system characteristics and ACO competencies, including centralization strategies to manage organizations, hospital integration with physicians and outpatient facilities, health information technology, infrastructure to monitor community health and report quality, and risk-adjusted 30-day all-cause mortality and case-mixed-adjusted inpatient costs for the Medicare population. Methodology: Panel data (2006–2009) were assembled from Florida and multiple sources: inpatient hospital discharge, vital statistics, the American Hospital Association, the Healthcare Information and Management Systems Society, and other databases. We applied a panel study design, controlling for hospital and market characteristics. Principal Findings: Hospitals that were in centralized health systems or became more centralized over the study period had significantly larger reductions in mortality compared with hospitals that remained freestanding. Surprisingly, tightly integrated hospital–physician arrangements were associated with increased mortality; as such, hospitals may wish to proceed cautiously when developing specific types of alignment with local physician organizations. We observed no statistically significant differences in the growth rate of costs across hospitals in any of the health systems studied relative to freestanding hospitals. Although we observed quality improvement in some organizational types, these outcome improvements were not coupled with the additional desired objective of lower cost growth. This implies that additional changes not present during our study period, potentially changes in provider payment approaches, are essential for achieving the ACO objectives of higher quality of care at lower costs. Practice Implications: Provider organizations implementing ACOs should consider centralizing service delivery as a viable strategy to improve quality of care, although the strategy did not result in lower cost growth.

Nurse staffing and post-surgical complications using the present on admission indicator

We evaluated the relationship between registered nurse (RN) staffing and six post-surgical complications: pneumonia, septicemia, urinary tract infections, thrombophlebitis, fluid overload, and decubitus ulcers, in a dataset that contained the present on admission (POA) indicator. We analyzed a longitudinal panel of 283 acute care hospitals in California from 1996 to 2001. Using an adaptation of the Quality Health Outcomes Model, we found no statistically significant relationships between RN staffing and the complications. In addition, the signs of the relationships were opposite to those expected. That is, as staffing increased, so did some of the complications. We discuss potential reasons for these anomalous results, including the possibility that increases in RN staffing may result in earlier detection of complications. Other explanations include issues with risk adjustment, the lack of nurse level variables in the model, and issues with the POA indicator itself.

California's Minimum-Nurse-Staffing Legislation And Nurses' Wages

In 2004, California became the first state to implement minimum-nurse-staffing ratios in acute care hospitals. We examined the wages of registered nurses (RNs) before and after the legislation was enacted. Using four data sets-the National Sample Survey of Registered Nurses, the Current Population Survey, the National Compensation Survey, and the Occupational Employment Statistics Survey-we found that from 2000 through 2006, RNs in California metropolitan areas experienced real wage growth as much as twelve percentage points higher than the growth in the wages of nurses employed in metropolitan areas outside of California.

Do Women Pay More for New Vehicles? Evidence from Transaction Price Data

Recently, there has been a debate concerning whether price discrimination by new vehicle dealers results in systematically higher prices for minority and female buyers. Ian Ayres and Peter Siegelman (1995) and Pinelopi Koujianoi Goldberg (1996) use different methodologies and different data sets and come to different conclusions. Ayres and Siegelman use testergenerated paired audits and find that black and female test buyers were quoted higher prices than white males. Goldberg uses Consumer Expenditure Survey (CES) data and finds no systematic evidence of price discrimination in transaction price discounts by race or sex (but greater variation in discounts for blacks). Our paper contributes to the debate on price discrimination in new car purchases using a new J.D. Power and Associates database that reports gross profit directly from dealer financial software systems. Although our data set does not distinguish the race of the buyer, we do get direct evidence on dealer profit for male and female buyers in a sample reflecting over 4,000 new vehicle transactions. Our paper addresses the previously obtained contradictory results and differs from the previous studies in several ways. The database we use permits us to measure directly dealers’ self-reported gross profit per vehicle—the best measure of differential treatment. The database reflects contemporaneously collected dealer data: it comes from the same financial software programs that are used to automatically report data to the franchiser/ manufacturer, financial institutions, and motor vehicle bureaus as well as being used to calculate sales tax liabilities. Further, the data set allows us to test for price discrimination in the profit contributions of ancillary tie-ins such as service contracts, credit life insurance, and dealer residuals from financing the vehicle purchase. In the Ayres and Siegelman study, black auditors received initial offers or negotiated prices that were much higher than those for the paired white male auditors at the same dealerships—

Nurse Staffing, Mortality, and Length of Stay in For-profit and Not-for-profit Hospitals

400 more for black females and