You can click on the projects to learn more about them. A list of publications can be found here.
The alarming increase in social inequality in the US corrodes the national ethos as the “land of opportunity” (Grusky and Hill 2018). Embedded in that imagery is the ideal of a nation where rich and poor can co-mingle at work, at play, and at school. The causes and consequences of such cross-SES relationships are the focus of the proposed research. Integrating network dynamics with models of social stratification provides untapped potential for our theoretical and empirical understanding of social inequality. Building on research examining network ties that bridge social strata, we propose to study the consequences of adolescent friendship dynamics that bridge socioeconomic boundaries for long-term socioeconomic attainment of disadvantaged youth.
Read more here.
Rising inequality has been linked to growing disparities within and between economic strata. Yet, existing approaches to analyzing inequality often disregard within-group inequality and are limited in addressing causal questions about why inequality is changing. This paper introduces a causal approach to examining how treatment variables impact within-group, between-group, and total inequality. The method permits both cross-sectional and longitudinal analyses. With longitudinal analyses, researchers can disentangle compositional changes (level of pre-treatment inequality, distribution of treatment across groups) from behavioral changes (changing treatment effects). Moreover, researchers can analyze changes relative to a timepoint (e.g., 1980) or relative to a counterfactual scenario (e.g., a counterfactual distribution of treatment). I demonstrate the utility of the approach by analyzing the changing effect of motherhood on women’s earnings and its consequences for women’s earnings inequality between 1980 and 2020. The results show that motherhood decreases women’s earnings inequality because it reduces inequality within economic strata.
Read more here.
Most multilevel analyses examine how lower-level units, such as persons, are affected by their embedding in contextual/aggregate units at a higher level (macro-to-micro link). The generalized multiple-membership multilevel model (MMMM) conceptually reverses this multilevel setup. It allows studying how the effects of units at lower levels propagate to a higher level (micro-to-macro link).
Previous studies examining micro-to-macro links either aggregated or disaggregated the data. These approaches obstruct the inherent aggregation problem, cannot separate micro-level and macro-level variance, and ignore dependencies among observations, thus inducing excessive Type-I error.
The MMMM overcomes these problems by explicitly modeling the aggregation from the micro to the macro level by including an aggregation function in the regression model. It is a theoretically and statistically sound solution to the study of micro-to-macro links with regression analysis.
The R package “rmm” provides an interface to fit the MMMM with Bayesian MCMC using JAGS from within R for a variety of outcomes (linear, logit, conditional logit, Cox, Weibull).
Read more here.
Family earnings inequality in the U.S. has for decades been on a steep upward trajectory (McLanahan 2004; Martin 2006; Western, Bloome, and Percheski 2008; Wodtke 2014). Family processes play an important role in shaping this inequality. One stream of work addressing these family processes points to the growing contribution of single-parent families to increases in income inequality (Martin 2006; Western et al. 2008). Another stream focuses on income dynamics among couples, showing that the increasing economic similarity of spouses has contributed to inequality (Schwartz 2010; Gonalons-Pons and Schwartz 2017). In new work, Gonalons-Pons, Schwartz, and Musick (2021) further demonstrate that, in recent decades, changes in couples’ earnings following parenthood have been the key driver of this increase in couples’ economic similarity. Wives have become more likely to remain employed after parenthood, while husbands’ labor supply following parenthood has remained largely unchanged.
In this paper, we show that increases in partnered women’s work attachment following parenthood have had an important impact on earnings inequality in the last thirty-five years. We use a novel approach to demonstrate how parenthood effects on earnings among partnered women play into aggregate earnings inequality.
Read more here.
Ever since my first class on social network analysis, I have been both intimidated and fascinated by network effect models. Studying how these models have developed in the last decade, I became interested in writing an accessible review of the key challenges for identifying peer effects as well as conducting a simulation study to compare the different approaches.
Thanks to Weihua An for providing me with the opportunity to contribute to such a review in the Annual Review of Sociology. An extension of this review in which I examine network effect models in greater detail can be found here.
Read more here.
Although theoretical work on coalition government survival emphasizes the importance of intra-party politics, few studies examine them empirically. Using novel data on extra-parliamentary party features in 19 countries, I analyze the impact of intra-party unity on government survival. I propose a new method to account for the complex multilevel structure of coalition government data resulting from parties being nested within often multiple governments and both parties and governments being nested within countries. The method allows testing theories on the interplay of coalition parties in their conjoint effect on governments. The results indicate that internally divided parties destabilize the constituting coalitions. This effect, however, is contingent upon the power structure among coalition parties and the complexity of the coordination problem. The evidence points to a crucial role of organizational features that empower party members in producing disunity. Consequently, while intra-party democracy serves many purposes, it impairs coalition governance. These results allow me to revisit untested hypotheses. I find evidence for Strøm’s (1990) theory of political parties, Katz and Maier’s (1995) cartel-party hypothesis, but not for Gamson’s law (1961).
Read more here.
Since for high-schoolers many consequential life choices lie still ahead, it is crucial that their choices suit their abilities. Among other things, students’ misperception of their academic ability can lead to educational misinvestment with potentially severe consequences. While previous research has documented gender differences in the self-perception of academic performance, disparities by socioeconomic status (SES) have not been investigated. This paper redresses the gap. Drawing on the German National Educational Panel Study, I examine disparities in the ability to self-evaluate academic performance by SES. Specifically, I rank school grades and academic self-concepts of 15-year-olds within each class to analyze whether their self-evaluation matches their relative performance in class. My findings indicate that students from disadvantaged families self-evaluate their school performance less accurately than students from high-SES families. This result is particularly strong for Math. Low-SES students whose performance lies two standard deviations below the mean performance overestimate their own performance twice as much as high-SES students. I theorize that and subsequently examine whether differential parental feedback on academic ability is the driving mechanism of these results. The mediation analysis confirms that the higher inaccuracy in the self-perception of ability can be fully attributed to differences in parent-teacher contact, parental involvement with homework, and parental educational aspirations. My study illustrates the interplay of parental feedback and children’s self-perception and indicates that low-SES students suffer from higher inaccuracy in the self-evaluation of their school performance due to less and less informative parental feedback.
Read more here.
A high response of the sample units approached is one of the cornerstones of survey research (Groves, 1989) and the growing nonresponse has been a constant worry of survey statisticians all over the world (De Leeuw & De Heer, 2002). Several theories on the reasons for nonresponse and panel attrition have been developed over the years (Stoop, 2005). Survey climate and attitudes towards surveys are key concepts in these theories (Loosveldt and Storms, 2008). De Leeuw and colleagues (2010) proposed a brief nine-item scale to measure a person’s survey attitude. It consists of three sub-constructs: survey enjoyment, survey value, and survey burden. The present paper examines whether this survey attitude scale (SAS) contributes to the explanation and prediction of nonresponse and panel attrition over and above the usual suspects associated with nonresponse and attrition (e.g. age, gender). Using longitudinal negative binomial regression and survival analysis, we evaluate the explanatory and the predictive power of the SAS in presence of an extensive list of covariates.
Read more here.