Research

Cross-Ethnic Diversity and Innovative Output

In many settings, diversity is increasingly being motivated by a “business case” in terms of economic benefits. I study the effect of diversity in scientific collaborations, through cross-ethnic coauthorships, on productivity and the flow of ideas. Using data on publications and their authors, I identify scientific topics that are disproportionately heavily studied by the most commonly represented ethnicities, and show that gaining coauthors of a different ethnicity induces a scientist to subsequently study topics associated with that ethnicity. In a causal analysis, I focus on the impact of Chinese coauthorships, leveraging the rapid increase in Chinese student and high-skilled worker admissions into the United States in the 21st century as a source of exogeneous variation in ethnic collaboration. IV estimates imply that having one Chinese coauthor induces a scientist's research to become significantly more aligned with that of Chinese scientists, increasing Chinese topic usage by 9% and closing up to 25% of the difference between the average non-Chinese and Chinese author. This impact on research direction is persistent over time. The quantity of output is also impacted positively, but this is accompanied by insignificant or negative responses in the average quality of output. The findings suggest that collaboration can induce meaningful change in research direction and make scientists more productive, producing a greater quantity of publication output but not necessarily greater quality.

Aging, Selection, and Scientific Productivity (with Matthew B. Ross, Huifeng Yu, Joseph Staudt, Julia I. Lane, Jason Owen-Smith, Gerald Marschke, and Bruce Weinberg) (under review)

Despite widespread concern about the aging of the scientific workforce, empirical evidence on the effect of age on scientific productivity is decidedly mixed. We show that a failure to account for an almost Darwinian selection process helps to reconcile the conflicting evidence. The combination of four large scale datasets on biomedical researchers provides new and unique insights on scientific age and productivity, and suggests that current allocations across scientific careers are broadly aligned with productivity, peaking earlier than previously believed and closer to when productivity peaks. They further suggest that long-career researchers maintain a high quantity of research by expanding and professionalizing their research teams.

Kinship Structure, Scientist Homophily, and Research Outcomes (with Seung-hun Chung) (under review)

Historical kinship structure is one of the most important determinants of the cultural aspect of psychology. Applying recent insights into the interplay of culture, psychology and the outcomes they shape, we investigate the relationship between the historical kinship structure of ethnic groups and the research output of scientists who are associated with those groups. Using data on biomedical researchers funded by the National Institutes of Health (NIH), and using the names of researchers as a proxy for association with various ethnic groups, we find that researchers whose ethnicity is related to a higher kinship density are more likely to coauthor with researchers from their own ethnicity. We also show that such researchers tend to display slightly lower levels of research productivity. As a potential explanation for this relationship, we discuss possible psychological differences arising from kinship structure.

Is Homeownership Related to Improved Health and Child Well-being Outcomes? (with Erkan Erdem, Orest Pazuniak, and Konul Riegel) (under review)

We explore whether homeownership is related to improved social outcomes, specifically, better health and child well-being. We measure child well-being in terms of enrollment in gifted classes and the lower prevalence of both developmental delays and behavioral differences that introduce classroom limitations. Using data from the Census Survey of Income and Program Participation and regression analyses, we explore the short and medium-run association of homeownership on the outcomes of interest. Controlling for socioeconomic covariates, we find that homeownership is associated with reporting better health and in certain situations increases the likelihood of children’s well-being. Although not causal in nature, our findings provide strong support for the positive effects of homeownership for children’s wellbeing. Lastly, we examine demographics and find that the homeowner-renter differential in reporting good health is greater among female than male adults but find little evidence of demographic differences along other dimensions.

A Model for a Sustainable Biomedical Research Workforce (with Bruce Weinberg)