Ted Mouw

Residential segregation and interracial friendship in schools

This article uses social network and spatial data from the National Longitudinal Study of Adolescent Health (Add Health) to examine the effect of racial residential segregation on school friendship segregation in the United States. The use of hierarchical models allows the authors to simultaneously estimate the effects of race, within‐school residential segregation, and school diversity on friendship choice using the Add Health data. The authors use these results to predict the decline in friendship segregation that would occur if across‐ and within‐school residential segregation were eliminated in U.S. metropolitan areas. The results suggest that about a third of the level of racial friendship segregation in schools is attributable to residential segregation. Most of this effect is the result of residential segregation across schools rather than within them.

Occupations and the Structure of Wage Inequality in the United States, 1980s to 2000s

Occupations are central to the stratification systems of industrial countries, but they have played little role in empirical attempts to explain the well-documented increase in wage inequality that occurred in the United States in the 1980s and 1990s. We address this deficiency by assessing occupation-level effects on wage inequality using data from the Current Population Survey for 1983 through 2008. We model the mean and variance of wages for each occupation, controlling for education and demographic factors at the individual level to test three competing explanations for the increase in wage inequality: (1) the growth of between-occupation polarization, (2) changes in education and labor force composition, and (3) residual inequality unaccounted for by occupations and demographic characteristics. After correcting for a problem with imputed data that biased Kim and Sakamoto’s (2008) results, we find that between-occupation changes explain 66 percent of the increase in wage inequality from 1992 to 2008, although 23 percent of this is due to the switch to the 2000 occupation codes in 2003. Sensitivity analysis reveals that 18 percent of the increase in inequality from 1983 to 2002 is due to changes in just three occupations: managers “not elsewhere classified,” secretaries, and computer systems analysts.

JOB RELOCATION AND THE RACIAL GAP IN UNEMPLOYMENT IN DETROIT AND CHICAGO, 1980 TO 1990

The spatial mismatch hypothesis argues that residential segregation and job decentralization combine to adversely affect the employment opportunities of minorities. While employment is increasingly located outside of central cities, residential segregation prevents minorities from moving closer to suburban jobs. Although this hypothesis has intuitive appeal, there is little consensus regarding its empirical validity. This study (1) constructs detailed geographic measures of changes in employment opportunities, (2) estimates a fixed-effects model of changes in the unemployment rate over time, and (3) accounts for spatial correlation in the error term. Neighborhood-level employment data from 1980 and 1990 are used to measure changes in the distance to jobs from census tracts in the Detroit and Chicago metropolitan areas. In both cities, the decentralization of employment and the loss of manufacturing jobs resulted in substantial changes in the spatial distribution of employment. The empirical results indicate that a decline in the spatial proximity to employment is associated with an increase in the unemployment rate for blacks.

Culture wars and opinion polarization: The case of abortion

Recent observers have pointed to a growing polarization within the U.S. public over politicized moral issues—the so‐called culture wars. DiMaggio, Evans, and Bryson studied trends over the past 25 years in American opinion on a number of critical social issues, finding little evidence of increased polarization; abortion is the primary exception. However, their conclusions are suspect because they treat ordinal or nominal scales as interval data. This article proposes new methods for studying polarization using ordinal data and uses these to model the National Election Study (NES) abortion item. Whereas the analysis of this item by DiMaggio et al. points to increasing polarization of abortion attitudes between 1972 and 1994, this articles analyses of these data offers little support for this conclusion and lends weight to their view that recent concerns over polarization are overstated.

Racial differences in the effects of job contacts: Conflicting evidence from cross-sectional and longitudinal data

Abstract Recent research has argued that using job contacts to find work restricts the opportunities of Black workers. Although this makes sense in light of literature on urban poverty that contends that inner-city Blacks are isolated from effective job networks, this paper argues that there are two reasons why these findings may be misleading. First, the existence of discrimination in the labor market means that contacts may be an efficient method of job search for Black workers. Second, because the use of contacts may be an endogenous variable, we need to look at longitudinal data. In the analysis, I compare cross-sectional and longitudinal results on the relationship between contacts and wages. While the cross-sectional data indicate a negative relationship between contacts and wages, the longitudinal data suggest this merely reflects lower levels of opportunity among workers who use contacts rather than the effect of contacts per se.

ARE BLACK WORKERS MISSING THE CONNECTION? THE EFFECT OF SPATIAL DISTANCE AND EMPLOYEE REFERRALS ON INTERFIRM RACIAL SEGREGATION*

I use data on the hiring practices and spatial location of firms in four cities to model the process of interfirm racial segregation. When I control for the spatial location of the firm, the use of employee referrals reduced the probability of hiring a black worker by 75% in firms that are less than 10% black. Among all firms, the results suggest that employee referrals are just as important as the geographic location of the firm in generating employment segregation: both increase the predicted level of interfirm racial segregation among blue-collar workers in the cities studied by about 10%.

Network Sampling with Memory: A proposal for more efficient sampling from social networks.

Techniques for sampling from networks have grown into an important area of research across several fields. For sociologists, the possibility of sampling from a network is appealing for two reasons: (1) A network sample can yield substantively interesting data about network structures and social interactions, and (2) it is useful in situations in which study populations are difficult or impossible to survey with traditional sampling approaches because of the lack of a sampling frame. Despite its appeal, methodological concerns about the precision and accuracy of network-based sampling methods remain. In particular, recent research has shown that sampling from a network using a random walk–based approach such as respondent-driven sampling (RDS) can result in a high design effect (DE): the ratio of the sampling variance to the sampling variance of simple random sampling (SRS). A high DE means that more cases must be collected to achieve the same level of precision as SRS. In this article, we propose an alternative strategy, network sampling with memory (NSM), which collects network data from respondents to reduce DEs and, correspondingly, the number of interviews needed to achieve a given level of statistical power. NSM combines a “list” mode, in which all individuals on the revealed network list are sampled with the same cumulative probability, with a “search” mode, which gives priority to bridge nodes connecting the current sample to unexplored parts of the network. We test the relative efficiency of NSM compared with RDS and SRS on 162 school and university networks from the National Longitudinal Study of Adolescent Health and Facebook that range in size from 110 to 16,278 nodes. The results show that the average DE for NSM on these 162 networks is 1.16, which is very close to the efficiency of a simple random sample (DE = 1) and 98.5 percent lower than the average DE we observed for RDS.

Network Structure and Biased Variance Estimation in Respondent Driven Sampling.

This paper explores bias in the estimation of sampling variance in Respondent Driven Sampling (RDS). Prior methodological work on RDS has focused on its problematic assumptions and the biases and inefficiencies of its estimators of the population mean. Nonetheless, researchers have given only slight attention to the topic of estimating sampling variance in RDS, despite the importance of variance estimation for the construction of confidence intervals and hypothesis tests. In this paper, we show that the estimators of RDS sampling variance rely on a critical assumption that the network is First Order Markov (FOM) with respect to the dependent variable of interest. We demonstrate, through intuitive examples, mathematical generalizations, and computational experiments that current RDS variance estimators will always underestimate the population sampling variance of RDS in empirical networks that do not conform to the FOM assumption. Analysis of 215 observed university and school networks from Facebook and Add Health indicates that the FOM assumption is violated in every empirical network we analyze, and that these violations lead to substantially biased RDS estimators of sampling variance. We propose and test two alternative variance estimators that show some promise for reducing biases, but which also illustrate the limits of estimating sampling variance with only partial information on the underlying population social network.

Social Support Among Latina Immigrant Women: Bridge Persons as Mediators of Cervical Cancer Screening

New Latina immigrants face numerous linguistic, cultural, logistical, and material barriers to cervical cancer screenings. Promotoras (lay health advisors) are a proven strategy to promote utilization of care. Since the mid-1990s, interventions in North Carolina have aimed to connect Latina immigrants to a broader range of bridge persons. This study assessed the effect of bridge persons on utilization of cervical cancer screening by Latina immigrants in North Carolina. Women were recruited in Spanish-language churches in four counties (N = 223). Logistic regression results show that persons known through advocacy organizations appeared to increase probability of recent Pap screening by an average of 10.4 percentage points (p < 0.05).Promotoras remain more effective, increasing probability of screening by 12.9 percentage points (p < 0.05) but few women (14%) knew one. No association was found with other bridge person profiles. Interventions are needed to better engage all bridge persons in linking immigrants to preventive health services.

The interplay of spatial diffusion and marital assimilation of Mexicans in the United States, 1980–2011

ABSTRACT Recent trends suggest a decline in the rate of intermarriage between Mexicans and non-Hispanic whites. In this paper, we argue that interpretations of this trend as a decline in preferences for intermarriage are misleading because of the lack of adequate data that captures both spatial and temporal variation in the level of intergroup contact. Using data from the Decennial Census (1980–2000) and the American Community Survey (2008–2011), we employ a novel methodological approach to disentangle the impact of spatial diffusion, ethnic replenishment, and shifts in preferences for homophily on Mexican ethnic intermarriage patterns across 543 Consistent Public Use Microdata Areas (c-PUMA). Once changes in the demographic composition of c-PUMAs are accounted for, multilevel models for repeated cross-sectional data provide no evidence of a change in the marital preferences of Mexicans over time. Trends in intermarriage rates are predominantly explained by compositional and structural changes.