For her thesis “From Cronkite to Williams: The Effect of News Coverage on the Number of Students Graduating with STEM Majors,” Welch gathered data on the number of science related news stories appearing on the nightly news on ABC, CBS and NBC from Vanderbilt’s Television News Archive. She related the number of individuals graduating with a STEM major in a given year to the number of stories that appeared when they were sophomores in high school. She finds a positive, statistically significant effect, and concludes that having more people, like Neil deGrasse Tyson, bring science into pop culture will directly affect the number of individuals who gravitate towards STEM majors. 

Gullquist’s thesis, “The Effects of Local Government Spending on Unemployment”, exploits the fact that government spending is lower under a mayor-council form where the mayor and city council are elected separately than a council-manager form where the city council has more leeway in policy-making, to identify the effects of local government spending on local unemployment rates. Gullquist collected data on city government form by hand and found that local spending does not directly impact unemployment rates. She concludes that while most local spending goes towards education and there is less discretionary funding than at the state or federal level, local spending still may have an effect years later increased education spending leads to a more productive work force.

Projections for 2014 US healthcare inflation sit at around 6.5 percent, four times greater than that of price inflation. As health insurance costs continue to increase more rapidly than wage rates, firms may respond by adjusting their employee compensation and/or their demand for labor. The hypothesized tradeoff between employer-provided health insurance and wages is uncontroversial, yet puzzlingly difficult to prove. The small pool of economists who have found evidence of a negative relationship have estimated models plagued with weak instruments and omitted variable bias. For one, this research fails to control for individual health status, which results in a bias and overestimation of the effects of health insurance on wages. I build on previous work by adding a variable for self-reported health status. I estimate multiple two-stage least squares models with 2009 cross-sectional data from the Survey of Income and Program Participation. By introducing health status, I find no evidence that a negative relationship exists between employer-provided health insurance and wage rates. More interestingly, for above-average earners, I find a positive relationship between wage rates and employer-provided health insurance; this is coupled with an increase in total hours worked for that group. One possible explanation is that institutional factors, such as a desire to retain the best employees, prevent firms from decreasing wages for this group, despite the ever-increasing cost of health insurance. 

The abstract for the article follows:

Numerous studies have examined the enrollment responses of traditional undergraduate students to the introduction of government-provided tuition subsidies, but far less attention has been devoted to the elasticity of demand for graduate education. This paper examines how the tax code and government education policies affect graduate enrollment and persistence rates along with the ways in which students fund their graduate education. Our empirical methodology is based on exogenous variations in the availability of an income tax exemption for employer-provided tuition assistance for graduate courses. We find that graduate attendance among full-time workers age 24–30 is higher when the tax exemption is available, mostly due to higher persistence in public universities and vocational course work. The use of employer aid for individuals enrolled in full-time and public part-time graduate programs also increases. We present some evidence that universities may adjust tuition to capture part of the incidence.

Do early-life health interventions have spillovers into politics? We test the effect of an exogenous shift to the left tail of the cognitive ability distribution on voting behavior. The rapid adoption of iodized salt nearly eliminated in utero iodine deficiency in the U.S., an impairment that is medically linked to decreased intelligence. Geographic variation in naturally occurring iodine levels before salt iodization allows us to identify the impact of changes in cognitive ability on whether and how people vote. We find that as iodine deficiency levels decrease, turnout for candidates from the Democratic Party increase leading to a greater share of the House of Representatives and more liberal congressmen.