Research

ORCID ID

Publications..

“A Decomposition of the Relationship Between Internet Access and Earnings”
The Journal of Applied Business and Economics, 2023 Vol 25 (6), pg 220-238 https://doi.org/10.33423/jabe.v25i6.6583

Gains in internet access and quality have increased output and revenue for companies, but these gains are not uniformly distributed across regions, industries, or worker types. I examine county-level wage and income measures on time-dynamic broadband uptake rates, mobile internet access, and local education levels. Datasets from the Census and FCC are combined to construct a rich dataset of all counties in the continental US from 2010-2019. Interaction terms between internet access, duration of access, and educational levels are included to capture the heterogeneous relationship between internet access and wages across across different levels of education. Decompositions of these regressions indicate that only 25% of the differences in wages and income are driven by heterogeneous returns to internet access and education across metro densities, with the majority of the difference determined by the levels of these variables themselves.

Threats to ACA Temporarily Increase uptake of Long-Acting Contraception
Applied Economic Letters , 2022, 1-9, https://doi.org/10.1080/13504851.2022.2058695
Long-Acting Reversible Contraception methods (LARCs) have a lower uptake rate than oral contraception, despite being more reliable and potentially cheaper in the long run. Historically, high insertion costs have stunted their uptake. The Affordable Care Act eradicated these costs by classifying contraception as preventative care that insurers must offer at no charge as of late 2012. Additionally, the ACA required that all employer-provided insurance plans have contraceptive coverage. After four years of coverage, the 2016 national election represented a credible threat to the ACA and contraceptive coverage and access. Utilizing the National Survey of Family Growth, I estimate how LARC uptake changed during and after 2016 using data from 2011 to 2019. I find that use of LARCs spiked in 2016 and 2017 for Medicaid users and 2017 and 2019 for women with private insurance.

Gender, Coauthorship, and Academic Outcomes in Economics With: Andrew Hussey & Wendy Stock
Economic Inquiry - Vol 60 (2), 2022, 465-484 https://doi.org/10.1111/ecin.13047
Given the growing importance of coauthorship in academics, we consider potential long-term determinants of coauthorship behavior as well as how coauthorship relates to career outcomes. In particular, we supplement a unique dataset of recent economics Ph.D. graduates with early career measures of their productivity, including number and quality of publications, coauthorship outcomes, citation rates, and tenure status. We examine relationships between student, Ph.D. program, and advisor characteristics and coauthorship outcomes. We also investigate relationships between coauthorship (including the gender composition of coauthors), and publication and tenure outcomes, and examine whether these relationships differ among males and females. Significant differences across gender are found in the formation and effects of coauthorship.

How Broadband and Cell Phone Access have Impacted Marriage and Divorce in the US
Review of Economics of the Household, Volume 18 (2), 2020, p 431-459 https://doi.org/10.1007/s11150-019-09464-z
In the last decade, the use of social media, chat rooms, and dating applications have vastly expanded. Consequently, search costs have lowered in the marriage market, as participants (both single and married) can now search for and gather information more easily on potential partners. By exploiting county level variation in broadband access and cell phone carriers I identify the impact of internet access on marriage and divorce. Search costs are predicted to be inversely related to population density; therefore, interactions between urban density and internet access are included to test for heterogeneous effects of internet expansion on marital demographics.

Fixed effect results indicate that increases in internet access, either through broadband or cell phone access, have decreased marriage and increased divorce in rural counties. However, in counties in large metropolitan statistical areas internet access has had the exact opposite effect. Expansion in household broadband accounts and cell phone carriers are correlated with increases in the married population and decreases in the divorced or separated populations in large metropolitan counties. Lastly, state level measurements on age at marriage and percent of the ever married population in their first, second or third marriage are incorporated to help place the initial findings in context. Results support that the rise in the married population is both from initial uptake in marriage increasing amongst the youth in urban areas, and by increased re-matching of divorcees in the secondary marriage market.

"The Procyclicality of Divorce and Urban Density"
Economics Bulletin, Volume 37, (1), January 2017, p.A37
This study measures the response of marriage to fluctuations in local labor market conditions more precisely than previous studies by using a newly-digitized data set of all divorces and annulments granted in US counties from 1970 to 1988. The observations are aggregated to the census-defined Labor Market Areas level, which allows for testing of heterogeneous treatment effects depending on the urban density levels in the LMA. Results show a strong, robust, and pro-cyclical relationship between employment rates and divorce rates. Importantly the disaggregated testing shows that the observed trend is driven by rural regions, which is significant in understanding regional trends and demographic patterns.

"The Environmental Dimensions of Emigration from Rural Mexico" with: Lori Hunter and Fernando Riosmena.

International Migration Review Volume 47, (4), December 2013, pp. 874-909(36)
Initial funding provided by The Center for Environment and Population & The Institute of Behavioral Sciences CU Population Program.

In many developing countries, natural resource dependency is a day‐to‐day reality for many rural households. As such, changes in weather and climate patterns hold tremendous potential to impact livelihoods. When livelihood options are constrained due to shifts in environmental conditions, migration becomes a significant, adaptive livelihood strategy reducing household vulnerability. In this project, we offer a preliminary answer to the question: Is emigration from rural Mexico to the U.S. associated with recent patterns of precipitation, net of other socio‐economic factors shaping migration patterns? Using data from the Mexican Migration Project (MMP), we model U.S. emigration from rural Mexican communities as related to community, household and environmental factors. We find that households subjected to drought conditions are far more likely to send a migrant as compared to those subjected to wet conditions. The results have important implications for policy and programmatic response to current migration pressures emphasizing diversification of rural Mexican livelihoods in the face of contemporary climate change.

Working Papers Under Review

Mentor Prestige, Job Market Signals, and Career Outcomes in Economics with Wendy Stock, and Andrew Hussey

Graduate school mentors can have significant impacts on PhD students’ careers. They can provide professional, emotional, and financial support to their students, and recommendations from prestigious mentors can be a positive signal in the job market. Using a unique dataset that links economics PhD graduates’ early and later career outcomes to their mentor and PhD program characteristics, we examine whether mentor prestige, publications while in graduate school, and other factors are related to graduates’ career outcomes and whether these relationships differ by PhD program rank. After controlling for other characteristics, mentor prestige has relatively little direct value for predicting graduates’ early or 10-year post-PhD academic career outcomes, particularly for graduates of top-tier PhD programs. For graduates of lower-tier programs, having a more prestigious dissertation advisor is positively associated with landing a tenure-track job, particularly for women and non-US citizens. Publishing with a mentor during graduate school is also positively related to initially placing in academia and being in academia 10 years post-PhD for graduates of lower-tier but not top-tier programs. Publishing with a mentor while in graduate school may also assist in getting an initial tenure track job and predict further research collaboration. However, coauthorship with mentors after graduation is negatively associated with tenure for graduates of lower-tier programs. Under review at Economics of Education Review

Internet Diffusion over Time and US Presidential Election Outcomes with Shatanjaya Dasgupta
Internet access has drastically changed how, when, and where people gather information. Politicians, Community activists, and political influencers have actively utilized internet outlets and social media to reach voters. In this paper, we study the relationship between internet diffusion in the aughts and US presidential election outcomes. Focusing on the 2012 and 2016 elections, we examine how the saturation rate of internet use and the length of time since saturation affect election outcomes. Utilizing FCC and Dave Leip’s Atlas of U.S. Presidential Elections, we examine how increased hardwired Internet adoption and mobile Internet access affect voter turnout and partisan vote share at the county level. American Community Survey data is used to construct a robust set of controls. Flexible estimation models that allow for the impact of the internet to vary from time of adoption show that the correlation between internet adoption and outcomes changed throughout the decade, and the duration of internet exposure increased the rate of correlation. Under review at Telematics and Informatics

Working Papers

Seeking Contraception: A Case Study of Charity Applicants in the South with Deborah M. Mullen

Tennessee was one of the few states that did not expand Medicaid after the passage of the Affordable Care Act and remains in the bottom quartile of insurance coverage at the state level. Across the state, a charitable organization has funded long-acting reversible contraception (LARCs) from 2010 to 2022. Women seek assistance directly through the agency or with a clinic that partners with the charity. Clients are provided family planning counseling, transportation when needed, and a free LARC if desired after their counseling appointment. Using proprietary data on charitable clinic partnerships that cover 74 of the 95 counties in TN, we examine the demographics of an underserved health community. Reviewing thousands of applications for LARC funding provides insight into patterns across the state’s rural and urban female populations. Future work will explore how policy and Title X changes affect this demand, and how LARCs received per county affect fertility rates and educational attainment as measured by the National Vital Health Statistics and American Community Survey.

The Impact of Internet on Local Fertility Rates
The internet has been shown to increase marriage rates in urban areas, particularly amongst those under 35 years of age. As more individuals match into unions there may be spillover effects of the internet on fertility rates. Additionally, the internet may affect fertility decisions by changing women’s wages, or by facilitating telecommuting and more flexible work schedules. Thereby, changing the opportunity cost of pregnancy and child rearing. Lastly, internet access can provide medical information and resources for effective family planning. In this paper, county-level fertility rates from the American Community Survey are regressed on broadband access rates and cell phone carrier data as measured by the Federal Communications Commission. Interaction terms between internet access and population density are included to allow for heterogeneous impacts of the internet on local matching and fertility. Further, regressions are run on women of different age ranges and education levels to see if impacts vary across these variables.

How Demographic and Urban Density Patterns affect the Pro-Cyclicality of Marriage and Divorce
This study measures the marital response to fluctuations in local labor market conditions by using a newly-digitized data set of all marriages, divorces, and annulments granted in US counties from 1970 to 1988. Market fluctuations are analyzed at the local Labor Market Area (LMA) level to allow for testing of heterogeneous treatment effects depending on the urban density levels in the LMA. Results show a strong, robust, and pro-cyclical relationship between all measures of economic well being and marriage and a weaker but still pro-cyclical relationship with divorce rates. Importantly the disaggregated results show that the pro-cyclical divorce trend is driven by rural regions with many urban areas showing an a-cyclical or counter-cyclical pattern. This trend is observed for both employment and income per capita. Marriage however is shown to be pro-cyclical in urban and rural areas alike, with urban areas being slightly more pro-cyclical to changes in income per capita.

When demographic controls are added, the percent of the population with a bachelors degree, female labor force participation, percent of the population that lived in different state five years prior, and the percent of the population that is black are all statistically significant variables in determining marriage and divorce but they explain only a small portion of the urban coefficient. Despite these controls, the coefficient on the urban interaction term remains strong in most regressions. Approximately 60% of the magnitude of the interaction coefficients are accounted for by age differences between urban and rural areas.

The Effects of Educational Attainment on Female Family Formation , with: Catherine Massey

Using the Fair Labor Standards Act of 1938, as a source of random variation in educational attainment, this paper aims to assess the relationship between years of education and female family formation. Under the new law, an individual under the age of 16, could not work during school hours. Using the 1980s IPUMs data, I construct a non-treated sample of women born between the years of 1917-1921, and those who would turn 16 after the law took effect, those born between 1923-1927. A two-stage least squares approach is used with education being estimated with an instrumental variable. In the second stage I examine how increases in education affected multiple marital outcomes: age of first marriage, number of children, and spousal characteristics. Preliminary research finds an additional year of schooling is associated with a decrease in age of first marriage and number of children born, and an increase in both the income and education attainment of husbands.

"The Effects of Permanent Declines in Low-Skilled Male Earning Potential on Divorce Rates"
This paper contributes to the literature by analyzing how job destruction and not transitory shocks to the business cycle affects marital stability. The industries of steel production, manufacturing and coal mining are characterized by relatively well paid, low-skilled workers. During the 70s and 80s, the US experienced a fall in steel production and manufacturing as well as both the boom and bust of coal mining. The loss of these high-paying, low- skilled and predominantly male filled jobs represents a permanent wage shock to the county and state residents where these industries were concentrated. In this paper I evaluate how the loss of high- paying low-skilled male jobs has affected the divorce rate. Using variation in the concentration of these industries across counties, I am able to identify the effects of a permanent decline in low-skilled male earning potential on divorce rates.

Divorce Spillover Effects: The Effects of Marriage Market Participation on Future Divorce Rates
An increase in the pool of singles can be destabilizing to existing marriages by decreasing search costs and increasing the probability of remarriage, particularly when the existing pool of singles is relatively limited. Additionally, increases in the stock of divorcees may lower the stigma of divorce and decrease the aversion to remaining single after separating. Using an innovative approach and a new data-set, this paper conducts an empirical examination of how previous fluctuations in marriage market participants correlate with current divorce rates. A difference-in-difference methodology is used to determine if state border-counties are more affected than interior counties by the divorce rate of a neighboring state’s border region. Results indicate that when a neighboring state's border-population exceeds one's own border-population, or is more urban, that the neighboring state's previous divorce rate is positively and significantly correlated with own state’s current border divorce rates. This finding is important for understanding the development of marriage markets over time and provides additional context for understanding the impacts of divorce liberalization in the US.
Currently under review at Cliometrica