CHAS 2005-2006 Seed-Grant Awards

Beth Angell

Discriminatory Employment Decision Making Towards Applicants with Behaviorally-Driven Health Conditions

The pernicious stigma associated with behaviorally-driven health conditions such as mental illness, substance abuse problems, and HIV/AIDS, is believed to contribute to discriminatory treatment in processes of hiring, retention, and promotion. Growing evidence suggests the prevalence of negative public stereotypes about these conditions, which may prime employers to reject applicants who are labeled mentally ill. In addition, employers who exercise a great deal of discretion over personnel decisions may reject applicants on the basis of physical or behavioral markers of behavioral disorder, even when applicants do not disclose their conditions. This study explores these propositions through an examination of the process and context of employer decision-making in relation to mental illness and related co-occurring conditions (substance abuse and AIDS), using data collected in a currently-funded cross-cultural survey comparing employer attitudes in urban China, Hong Kong, and the United States. Extending the ongoing study, this proposal uses data from qualitative interviews with 100 firm owners to examine 1) the ways that employers may explicitly (in response to labels) and implicitly (in judgments of soft skill) reject applicants via decisions not to hire, retain, or promote; and 2) cultural variations in these decisions that relate to labor market factors, workplace variation, and culturally-mediated attitudes about work, disability, and inclusion.

Anirban Basu

Causal estimation of intention-to-treat effects in randomized clinical trials

Clinicians often aim to estimate intention-to treat (ITT) effects within randomized clinical trials. ITT represents the expected change in outcomes for a randomly chosen person in the population who has been prescribed the new treatment rather than standard care. However, non-compliance and non-response pose problems in the causal estimation of this parameter using available data from clinical trials. Recently, researchers have proposed a latent preference framework that allows causal estimation of ITT (Frangakis & Rubin, 1999). Although this constitutes a significant improvement over traditional methods, it restricts its application to scenarios with all-or-none noncompliance and does not allow for outcome based noncompliance or non-response.  We propose a new estimator that will extend this current framework to not only account for preference-based compliance and drop-outs but also for outcome-based noncompliance and drop-outs, a phenomenon that is common in many clinical trials. We will study the consistency of this proposed estimator using simulation techniques. We hope that this estimator will further extend the applicability of these estimation methods in clinical trials where subjects do drop out of the trial due to lack of efficacy from the current treatment.

Julia Henly and Susan Lambert

The Scheduling Pilot: An Experimental Workplace Intervention to Reduce Instability in Retail Jobs.

Nonstandard work schedules that involve unpredictable and variable work hours are a source of instability common to lower-level jobs in the retail industry. Existing survey and qualitative research suggests that the instability resulting from nonstandard work schedules can have negative consequences for firms, workers, and their families (Presser, 2003; Henly & Lambert, 2005). This experimental study tests a workplace intervention designed to improve the predictability of entry-level retail workers’ schedules. The study will examine the effect of the intervention on worker performance, work-family conflict, and worker and family health and well-being. The study will also develop and test a return-on-investment (ROI) model that estimates the effects of the intervention on store performance, including sales levels and labor costs and savings.

Judith Levine

Women, Work, and Family: Labor Market Decision-Making in the Welfare Reform Era

The project is an examination through semi-structured, in-depth qualitative interviews of how women on the border between welfare and low-wage work – both before and after welfare reform – perceive and manage incentives, costs, and barriers of entering the labor market. It explores what keeps women out of the labor market, how women who become employed fare, and how policies might be designed and implemented better to limit the cost of employment on poor families. The approach of the study is to take a more holistic, ground-level view of poor families’ actual experiences negotiating welfare reform. In so doing, a broad-range of factors both weighing on work decisions and on the possible effects of employment will be explored. One set of these factors of particular interest involve health status and health insurance. We know from past work that poor health lengthens welfare spells and that welfare leavers frequently lose insurance coverage. This study will explore the role health status and insurance coverage play in poor women’s calculations about the costs and benefits of going to work, the actual ability to enter and maintain employment, and the health-related hardships associated with exits from welfare to work.

Blase Polite

Perceptions of Chemotheraphy between African Americans and Whites: Are There Differences and Do They Matter?

African Americans have higher age-adjusted mortality rates from breast, colon, and lung cancer than whites.  Much of this difference can be attributed to the fact that African Americans present with more advanced disease than do whites; however, even when the advanced stage at presentation is controlled for, African Americans continue to have higher mortality rates.  One possible reason for this stage-specific mortality difference is that African Americans are less likely to receive recommended adjuvant chemotherapy than whites.  The reasons for this are not clear and are likely multi-factorial.  The specific aims of this study are two-fold: (1) To describe how African Americans and whites with surgically resected breast, lung, and colon cancer treated at an urban tertiary medical center perceive chemotherapy; (2) To describe potential associations between these perceptions and the subsequent receipt of chemotherapy in the adjuvant (after curative surgery) setting.

Michael Sosin

Understanding the Effects of Cutbacks in Substance Abuse Services

The proposed pilot study seeks to understand the reactions of states, communities, and providers to those recent state cutbacks in substance abuse services that stem from the general state fiscal crisis. Relying on interview methods, the study assesses the effects on provider financing and client services of the interplay of the general social environment, state and community policies, and provider strategies and structures. This assessment makes use of theories of organizational rationing, cross-subsidization, and organizational ecology. Interviews are conducted with public officials, advocates, and representatives of providers in Oregon and Illinois.

Diane Whitmore

Measuring the Impact of School Policies on Chilhood Obesity

Diane Whitmore is collaborating on a project to measure the impact of school policies on childhood obesity and is engaged in a data collection effort to study these effects. Today’s movement toward stricter school accountability standards has been changing the inner workings of elementary schools in the United States, raising test scores in the process. But in order to ensure that children spend more time engaged in learning, other school-time activities – many of which may affect childhood obesity, like physical education and recess, and required summer school attendance – have been altered. These changes have been particularly abrupt in schools labeled as failing under their states’ accountability regime. Under the assumption that children in schools just above and below the cutoff are similar, the impact of changes induced by the accountability measures on children’s health can be identified.