Properties of Model-Averaged BMDLs: A Study of Model Averaging in Dichotomous Response Risk Estimation

Model averaging (MA) has been proposed as a method of accounting for model uncertainty in benchmark dose (BMD) estimation. The technique has been used to average BMD dose estimates derived from dichotomous dose-response experiments, microbial dose-response experiments, as well as observational epidemiological studies. While MA is a promising tool for the risk assessor, a previous study suggested that the simple strategy of averaging individual models' BMD lower limits did not yield interval estimators that met nominal coverage levels in certain situations, and this performance was very sensitive to the underlying model space chosen. We present a different, more computationally intensive, approach in which the BMD is estimated using the average dose-response model and the corresponding benchmark dose lower bound (BMDL) is computed by bootstrapping. This method is illustrated with TiO(2) dose-response rat lung cancer data, and then systematically studied through an extensive Monte Carlo simulation. The results of this study suggest that the MA-BMD, estimated using this technique, performs better, in terms of bias and coverage, than the previous MA methodology. Further, the MA-BMDL achieves nominal coverage in most cases, and is superior to picking the "best fitting model" when estimating the benchmark dose. Although these results show utility of MA for benchmark dose risk estimation, they continue to highlight the importance of choosing an adequate model space as well as proper model fit diagnostics.     

An Equilibrium Model of Health Insurance Provision and Wage Determination

We investigate the effect of employer‐provided health insurance on job mobility rates and economic welfare using a search, matching, and bargaining framework. In our model, health insurance coverage decisions are made in a cooperative manner that recognizes the productivity effects of health insurance as well as its nonpecuniary value to the employee. The resulting equilibrium is one in which not all employment matches are covered by health insurance, wages at jobs providing health insurance are larger (in a stochastic sense) than those at jobs without health insurance, and workers at jobs with health insurance are less likely to leave those jobs, even after conditioning on the wage rate. We estimate the model using the 1996 panel of the Survey of Income and Program Participation, and find that the employer‐provided health insurance system does not lead to any serious inefficiencies in mobility decisions.     

Work Accident Insurance and Prevention in Southern Africa: The Case of Zimbabwe

The paper uses Zimbabwe as a case study to depict the situation in southern Africa with respect to work accident insurance and prevention, provided by workers' compensation schemes which are among the earliest forms of social security to be introduced in the region. There has been an unfortunate tendency to concentrate on compensation issues at the expense of prevention initiatives. Zimbabwe, however, has attempted to integrate the two, notwithstanding the fact that the coverage of the scheme is narrow and the benefits inadequate.     

Segmented Assimilation,Local Context and Determinants of Drug Violence in Miami and San Diego: Does Ethnicity and Immigration Matter?1

Does the ethnic and immigrant composition of a community and existence of immigrant enclaves or barrios influence community level drug violence? This study explores the relationship between these and other factors in Miami and San Diego census tracts. We employ data about the distribution of Cubans, Central Americans, Haitians, Mexicans and Southeast Asians, controlling for social and economic influences of drug versus non‐drug violence. We also analyze the impact of various waves of immigration and immigrant communities to understand the circumstances under which drug violence occurs or is limited at the census tract level. The findings lend some support to the positive and negative aspects of Portes and Rumbaut's (2001) segmented assimilation hypothesis in Miami and San Diego neighborhoods. The strength of this conclusion varies and is contingent upon ethnic composition, new versus old immigration, and the all‐encompassing effects of economic deprivation     

A DIAGNOSIS FOR OUR TIMES:

This paper explores two submerged networks in the Pittsburgh area alternative health movement through the use of thirty two intensive interviews and ethnographic research. Utilizing Jürgen Habermas's colonization theory as well as Alberto Melucci's emphasis on the process of collective identity formation, the paper describes a "conservative Christian' and a culturally progressive network demonstrating how these networks differ and what they have in common. The paper examines how new oppositional identities and meanings emerge from face to face interactions between alternative practitioners, patients and activists in a variety of settings. Overall, the paper argues that resistance to the "medical industrial complex,' one leading instrument of "colonization,' is coming from the Christian right as well as elements of the cultural left.     

Deserts and oases: the continuing concentration of population in the American Mountain West

The Mountain West is a region that seems to be simultaneously rural and urban. With its wide-open spaces, many national parks, monuments, and forests, and high degree of federal land ownership the West appears as the quintessential rural area. However, over 70 percent of the West's population live in metropolitan areas. This simultaneous rural and urban nature of the West is important in understanding the changing population geography of the region. We examine this by focusing on changing patterns of population concentration among metro and nonmetro counties. Unlike other regions in the US, the Mountain West has never experienced a period of counterurbanization or population deconcentration. Not only is current in-migration to the region increasingly concentrated in old and new metro areas, it is also concentrated in a select number of nonmetro areas as well—particularly nonmetro counties adjacent to metro areas, in retirement destinations, and in recreation centers.     

CONCENTRATED POVERTY, RACE, AND HOMICIDE

Research on the link between levels of poverty and homicide in urban areas has persistently reported the existence of a relationship for whites but not for blacks. This is despite the fact that most analysts expect that the higher levels of urban black homicide are due in part to the higher levels of urban black poverty. The present research introduces a more meaningful, spatially based measure of concentrated poverty and argues that the effect of concentrated poverty on homicide rates should be the same for both racial groups. The hypotheses are tested with race-disaggregated data for a sample of central cities circa 1990. The results suggest that, when poverty is measured as a linear spatially based phenomenon, in is a more important determinat of race-specific homicide rates than overall city levels of disadvantage and that the concentration of poverty increases both black and white homicide rather equally.     

USING LINEAR PREDICTORS TO IMPUTE ALLELE FREQUENCIES FROM SUMMARY OR POOLED GENOTYPE DATA

Recently-developed genotype imputation methods are a powerful tool for detecting untyped genetic variants that affect disease susceptibility in genetic association studies. However, existing imputation methods require individual-level genotype data, whereas in practice it is often the case that only summary data are available. For example this may occur because, for reasons of privacy or politics, only summary data are made available to the research community at large; or because only summary data are collected, as in DNA pooling experiments. In this article, we introduce a new statistical method that can accurately infer the frequencies of untyped genetic variants in these settings, and indeed substantially improve frequency estimates at typed variants in pooling experiments where observations are noisy. Our approach, which predicts each allele frequency using a linear combination of observed frequencies, is statistically straight-forward, and related to a long history of the use of linear methods for estimating missing values (e.g. Kriging). The main statistical novelty is our approach to regularizing the covariance matrix estimates, and the resulting linear predictors, which is based on methods from population genetics. We find that, besides being both fast and flexible - allowing new problems to be tackled that cannot be handled by existing imputation approaches purpose-built for the genetic context - these linear methods are also very accurate. Indeed, imputation accuracy using this approach is similar to that obtained by state-of-the art imputation methods that use individual-level data, but at a fraction of the computational cost.