Thesis advisor: Claudia Olivetti / Injuries and violence are a major public health issue and represent a threat to individual wellbeing, productivity, and societal development at large. In recent years, the public health approach to reduce violence and injury has become crucial in guiding research and public policy. Governmental and nongovernmental organizations are strengthening data collection and surveillance systems to promote research and inform policy making. Yet the idea that violence and injuries can be prevented through systematic monitoring and research is still a novel one. The three essays that comprise this dissertation make advancements in this direction focusing on the United States. The first two chapters focus on child safety, examining the issues of bullying and firearm violence at school. The third chapter expands on the topic of gun violence examining the impact of firearm legislation on the broader problem of suicide deaths. From a methodological point of view, this dissertation combines economic models with a public health approach employing both structural estimation techniques and a quasi-experimental approach. While quasi-experimental methods are effective in uncovering broad causal relationships between legislative changes and outcome measures, structural estimation methods are essential when interested in recovering deep preference parameters and performing counterfactual policy analysis. As such, this dissertation represents an example of multidisciplinary work combining Economics and Public Health, and highlights the importance of employing diverse methodologies to uncover crucial behavioral patterns and their policy implications. The first essay, titled Is School Bullying Contagious?, uses a nationwide cross-section of students to uncover peer effects in adolescent bullying behavior at school. Victimization at school has been linked to a number of adverse effects for child development and well-being, including depression, higher drop-out risk, and lower earnings during adulthood. While understanding social interactions in bullying behavior is essential to designing effective policies, previous empirical work has overlooked the impact of classmates’ behavior on the individual inclination for bullying. This essay estimates a structural model of bullying with social interactions where the individual bullying effort depends on the average effort among classmates. The model controls for individual and family characteristics, classmates’ characteristics, as well as classroom unobservable factors. The results present strong evidence of peer effects in a large number of bullying behaviors. Considering a median classroom of 20 students with five bullies, the introduction of a new bully would spawn two additional bullies due to peer influences. This suggests that social interactions can be targeted to reduce the prevalence of bullying. In particular, counter-factual policy experiments indicate that schools may achieve sizable reductions in the number of bullies by spreading them out over classrooms. The second essay, titled Gun Laws and School Safety, is joint with Summer Hawkins and Christopher Baum. Motivated by the documented link between school safety and psychological well-being, this essay examines the impact of state-level gun control on adolescent school safety. The analysis uses data on 926,639 adolescents from 45 states in the 1999-2015 Youth Risk Behavior Surveys. Students self-reported on weapon carrying at school, the number of times they had been threatened or injured with a weapon at school, the number of school days missed due to feeling unsafe, and weapon carrying at any location. For each state and year, 133 gun laws were combined into an index of gun control strength. Difference-in-differences logistic regressions were used to evaluate the impact of stricter gun laws on binary measures of school safety. Each regression controlled for individual and state characteristics, as well as year and state fixed effects. An interquartile-range (IQR) increase in the index (i.e. a 15-point increase corresponding to a strengthening of gun control) was associated with a 0.8 percentage point decrease in the probability of weapon threats at school (p=0.038) and a 1.2 percentage point decrease in the probability of missing school due to feeling unsafe (p=0.004). While we did not find a significant impact of gun laws on weapon carrying at school, an IQR increase in the index was associated with a 2-percentage point decrease in the probability of carrying weapons at any location (p=0.002). Our results suggest that the adoption of stricter state gun laws may improve school climate and subjective perceptions of safety. The third essay, joint with Summer Hawkins and Christopher Baum, is titled Gun Laws and Firearm Suicides. Between 2005 and 2015, suicide rates have been steadily increasing in the US, with firearm suicides representing over half of all suicides and the primary cause of firearm mortality. As such, firearm suicides represent an urgent policy matter and a prompt policy response is required. Using a 10-year-long panel of the 50 states, we investigated whether stricter gun laws may reduce firearm suicides, possibly by reducing firearm availability. As a reduction in firearm availability may simply result in a substitution towards alternative suicide methods, we further explored whether stricter gun laws are associated with an increase in non-firearm suicides. We analyzed 2005-2015 National Vital Statistics System mortality files from the 50 states, with 212,804 firearm suicides and 206,795 non-firearm suicides. We measured the strength of state-level gun control using an index that combines 133 different laws. We conducted difference-in-differences regression models to assess whether changes in the index were associated with changes in the number of firearm and non-firearm suicides. We found that implementing an additional gun law would result in a decrease in the number of firearm-related suicides by 2 to 4 percentage points. In addition, significant interactions between the gun score and demographic characteristics suggest that the effectiveness of stronger gun laws is the highest among individuals age 20 to 49, but seems to be null among black individuals. Although we found no overall association between a stricter gun law environment and non-firearm suicides, stricter gun laws seem to increase non-firearm suicides among white and black individuals, suggesting that additional policy actions are required to prevent suicides in these groups. Our findings are robust to controlling for demographic characteristics, state time-varying characteristics, state and year fixed effects, as well as state-specific time trends. We also provide graphical evidence that trends in suicide rates were not dependent on the level of strength of gun control, supporting the parallel trend assumption and a causal interpretation of our estimates. / Thesis (PhD) — Boston College, 2018. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Economics.

New and more advanced prenatal tests have steadily been introduced in the Swedish maternity care system in the last 30 years. The combined test, CUB, was introduced step wise in Swedish maternal care from 2008 and onward. The CUB test detects children with chromosomal abnormalities prenatally and is offered at no charge for women in treated counties. This thesis investigate the reform using a difference-in-difference approach to determine the effect of the CUB test on the number of late abortions performed. My theoretical framework suggest that the introduction of CUB should increase the number of abortions of children with chromosomal aberrations. As supported by theory I find a positive effect of CUB on late abortions for my main group of interest, women 35-39 years old. These women were the ones most effected by CUB. The positive effect of 0.47 percentage units is statistically significant at the 10% level. It corresponds to a 3.6-7.1% decrease in the number of babies born with chromosomal aberrations.

Três modelos teóricos de negociação entre seguradores e provedores privados são desenvolvidos e seus resultados simulados e analisados. Os modelos procuram captar o que seria o encontro entre um segurador e um provedor vinculados entre si por um contrato de parceria e sentados à mesa para negociarem os preços do seguro-saúde e do bem/serviço médico. Na estrutura dos modelos está presente o problema de agente-principal característico da relação entre os dois atores, dadas as assimetrias informacionais inerentes ao campo da saúde. Tanto o segurador quanto o provedor estão restritos por considerações junto a seus consumidores. O processo de negociação em si baseia-se no modelo de Stackelberg para oligopólios, o que produz dois cenários: num primeiro cenário o líder da negociação é o provedor e num segundo cenário o líder da negociação é o segurador. / Three theoretical models about negotiation between private insurers and providers are developed and their results are simulated and analyzed. The models try to catch what would be a meeting between an insurer and a provider tied each other by an association contract bargaining the health insurance price and the medical good/service price. The main problem is the agent-principal relationship in an environment with asymmetric information. Both agents are restricted by consumer behavior. The basis for the bargain process is the Stackelberg model for oligopolies, which gives two scenes: one with insurer leadership and other with provider leadership.

Na literatura de federalismo fiscal, além da discussão relativa ao efeito flypaper, surge para despesas condicionais evidência de que o gasto na área escolhida pode aumentar menos do que a quantidade da transferência destinada àquela área, sendo o restante dos recursos destinados a outros bens e serviços públicos - o efeito fungibilidade. Pretende-se avaliar o impacto das transferências do Sistema Único de Saúde testando além do efeito flypaper, o efeito fungibilidade para despesas municipais de saúde (totais, com pessoal, com medicamentos, com serviços de terceiros e com investimentos). Utilizaram-se modelos clássicos de dados em painel e regressões quantílicas com efeitos fixos. Obteve-se evidência de que, para o total, ocorre o efeito fungibilidade para transferências do Piso de Atenção Básica fixo, uma vez que para cada R$ 1 de transferências para saúde cerca de R$ 0,20 é aplicado em outras áreas. Além disso, tal efeito parece ser menor em municípios que apresentam gastos mais elevados em saúde. Ademais, o efeito flypaper parece estar mais associado a despesas com pessoal. / In the fiscal federalism literature, beyond the discussion on the flypaper effect arises evidence for conditional grants that spending in the chosen area may increase by less than the amount of the transfer meant to that area, with the remaining funds being used for other goods and public services - the fungibility effect. This work aims to assess the impact of transfers from Unified Health System besides the flypaper effect, testing also the fungibility effect for different kinds of municipal health expenditures (total expenditures, expenditures with employees, expenditures with drugs, expenditures with outsourced services and investment expenditures). Classical models of panel data and quantile regressions with fixed effects were used. There is evidence that, for total expenditures, fungibility effect occurs for Fixed Basic Atention National Policy grants, since for every R$ 1 of health transfers to approximately R$ 0,20 is applied in other areas. Moreover, this effect appears to be lower in cities that have higher health care expenditures. Furthermore, the flypaper effect seems to be more associated with expenditures with expenditures.

Este estudo constrói um modelo microeconômico estático baseado nas teorias do consumidor e da firma para sintetizar o funcionamento do mercado de saúde. O modelo envolve os três principais agentes no setor privado: o consumidor, o provedor e o segurador. O seu objetivo é representar simultaneamente os três problemas tradicionais do mercado de saúde: a seleção adversa, o perigo moral e a demanda induzida. Desses três problemas, o perigo moral é o que recebe a atenção mais completa. Não obstante, ao invés da usual distorção do preço da assistência médica observado pelo consumidor segurado, a explicação do modelo para o problema do perigo moral está no custo da cooperação entre os consumidores segurados de um fundo de seguro-saúde. Por sua vez, a seleção adversa aparece no modelo como um viés entre a porcentagem observada de doentes na população e a porcentagem esperada de doentes no fundo, na medida em que a porcentagem no fundo permanece acima da porcentagem na população antes de igualá-la. Por último, a demanda induzida surge como uma discriminação do preço da assistência médica entre os pacientes segurados e não segurados de um provedor. A aplicação do modelo é demonstrada por meio da simulação de um mercado de saúde com dados relativamente parcimoniosos sobre uma população, uma doença e a cura para essa doença. Dentre os resultados, destaca-se o efeito de uma melhor distribuição de renda em reduzir o preço do seguro-saúde e diminuir a distância entre pacientes segurados e não segurados. / This study constructs a static microeconomic model based on consumer and firm theories in order to synthesize the functioning of the health market. The model deals with the three main agents in the private sector: the consumer, the provider and the insurer. Its objective is to simultaneously represent the three traditional problems of the health market: adverse selection, moral hazard and induced demand. Of these three problems, moral hazard is the one which has received the most complete attention. However, instead of the usual distortion of the price of health care observed by the insured consumer, the model?s explanation for the moral hazard problem is the cooperation cost among the insured consumers of a health insurance fund. In turn, the adverse selection appears in the model as a bias between the observed percentage of sick persons in the population and the expected percentage of sick persons in the health insurance fund, since the fund?s percentage stays above that of the population?s percentage rather than equal to it. Finally, the induced demand comes out as a discrimination of the health care price between insured and uninsured patients of a medical provider. The model?s application is demonstrated through a simulation of a health market with relatively parsimonious data on a population, an illness and the cure for this illness. One can detach among the results the effect of a better income distribution in reducing the health insurance price and diminishing the distance between insured and uninsured patients.

Este estudo constrói um modelo microeconômico estático baseado nas teorias do consumidor e da firma para sintetizar o funcionamento do mercado de saúde. O modelo envolve os três principais agentes no setor privado: o consumidor, o provedor e o segurador. O seu objetivo é representar simultaneamente os três problemas tradicionais do mercado de saúde: a seleção adversa, o perigo moral e a demanda induzida. Desses três problemas, o perigo moral é o que recebe a atenção mais completa. Não obstante, ao invés da usual distorção do preço da assistência médica observado pelo consumidor segurado, a explicação do modelo para o problema do perigo moral está no custo da cooperação entre os consumidores segurados de um fundo de seguro-saúde. Por sua vez, a seleção adversa aparece no modelo como um viés entre a porcentagem observada de doentes na população e a porcentagem esperada de doentes no fundo, na medida em que a porcentagem no fundo permanece acima da porcentagem na população antes de igualá-la. Por último, a demanda induzida surge como uma discriminação do preço da assistência médica entre os pacientes segurados e não segurados de um provedor. A aplicação do modelo é demonstrada por meio da simulação de um mercado de saúde com dados relativamente parcimoniosos sobre uma população, uma doença e a cura para essa doença. Dentre os resultados, destaca-se o efeito de uma melhor distribuição de renda em reduzir o preço do seguro-saúde e diminuir a distância entre pacientes segurados e não segurados. / This study constructs a static microeconomic model based on consumer and firm theories in order to synthesize the functioning of the health market. The model deals with the three main agents in the private sector: the consumer, the provider and the insurer. Its objective is to simultaneously represent the three traditional problems of the health market: adverse selection, moral hazard and induced demand. Of these three problems, moral hazard is the one which has received the most complete attention. However, instead of the usual distortion of the price of health care observed by the insured consumer, the model?s explanation for the moral hazard problem is the cooperation cost among the insured consumers of a health insurance fund. In turn, the adverse selection appears in the model as a bias between the observed percentage of sick persons in the population and the expected percentage of sick persons in the health insurance fund, since the fund?s percentage stays above that of the population?s percentage rather than equal to it. Finally, the induced demand comes out as a discrimination of the health care price between insured and uninsured patients of a medical provider. The model?s application is demonstrated through a simulation of a health market with relatively parsimonious data on a population, an illness and the cure for this illness. One can detach among the results the effect of a better income distribution in reducing the health insurance price and diminishing the distance between insured and uninsured patients.

This dissertation focuses on how changes in SNAP (Supplemental Nutrition Assistance Program, formerly called food stamp) policies affect the consumption choices for recipient households and how macroeconomic environment affects individuals’ behaviors. In the first chapter, I examine the consumer response to SNAP benefit change. In the second chapter, I examine the investor attention under different stock market valuations. In the third chapter, I examine the impact of rising housing prices on individuals’ marriage entry. In the first chapter, we examine how SNAP benefits were spent. The American Recovery and Reinvestment Act of 2009 (commonly known as the Stimulus Act) temporarily increased the benefit level for SNAP recipients and released qualification requirements. We use a consumer panel with detailed transaction records to analyze the impact of SNAP benefits changes on participant household’s consumption choices. We find that the marginal propensity to spend (MPS) on food out of SNAP benefits increase is 0.45. The MPS out of SNAP benefit decrease is -0.85. SNAP participant households are more sensitive toward the benefits decrease than increase. We also study how SNAP benefits spent are in response to benefit changes. We find that with more SNAP benefits recipient households consume much more tobacco and with less benefit they consume much less alcohol. Results are robust to various placebo tests. In the second chapter, I examine the impact of stock market valuation on investor attention. Investor attention affects stock return variance and risk premia. Using Google Trend data, I come up with a new proxy for measuring investor attention. This chapter investigates investor attention effect on the merger announcement and how market valuations and days of the week affect investor attention. With the Google search index, this study finds that investor attention has positive correlations on market reactions following the merger announcement. In the third chapter, I estimate the impact of increasing on first marriage age in China. The first marriage age plays a very important role in the population economy, especially for a demographic dividend. The marriage market is affected by income, education, wealth inequality, consumption, etc. In China, the first marriage age has fallen since the late 1990s. This chapter focuses on how the price of the apartments change people decisions on marriage and ascribes that part of the delay to first marriage age is due to increase in housing price over the same period. In China, social norms require men to own an apartment before they get married. Based on the empirical evidence of population dynamics, the chapter suggests that a 1% increase in housing price will result in 0.016 years delay in the age of first marriage in China. When the housing price increases too much, it will have an even larger effect on marriage decisions because the high housing price becomes unaffordable for young couples. The result is robust with a wide range of model specifications.

This thesis examines the effect of neighborhood crime rates on childhood obesity in Los Angeles County over a five-year period 2012-2016. Using yearly pooled cross-sectional geocoded data from the University of Southern California (USC) Price Center for Social Innovation Neighborhood Data for Social Change (NDSC) interactive platform, I run multiple ordinary least squares regressions using different measures of crime to determine if neighborhoods with higher crime rates influence the unhealthy percentage of 5th, 7th, and 9thgrade public school students. I hypothesize that crime influences obesity, violent crime has a stronger correlation than property crime, and that greater parks access reduces obesity. My regression results fail to support hypotheses one and two. Hypothesis three is supported by the available data.

Lameness is considered one of the most important health and welfare issues in the dairy industry. Understanding the total cost per case of disease-specific lameness can help producers select better treatment, prevention, and control strategies for their herds. The first objective of our research was to calculate the costs associated with 3 lameness causing hoof diseases: digital dermatitis, sole ulcer, and white line disease. To accomplish this, a survey of hoof health professionals (hoof trimmers and veterinarians) was conducted to identify treatment related expenditures per case. Data from the hoof trimmer responses to the survey and previously published research were incorporated into a farm-level stochastic simulation model to determine the expected costs per case of each disease and the most influential factors associated with disease costs. The cost per case was calculated by disease type, severity (mild or severe), incidence timing (0 to 60 days in milk, 61 to 120 days in milk, 121 to 240 days in milk, or > 240 days in milk), and parity group (primiparous or multiparous). The second objective of our research was to determine the economic value of investing in different lameness prevention strategies. Two prevention strategies were considered: 1) prevention of infectious hoof diseases and 2) prevention of non-infectious hoof diseases. The total expenditures (therapeutics, outside labor, on-farm labor, and prevention costs) and losses (discarded milk, reduced milk production, extended days open, increased risk of culling, increased risk of death, and recurrence losses) associated with each prevention strategy before and after prevention implementation were calculated and compared to find the breakeven investment cost.

Lumbar degenerative disc disease is a remarkably common condition among patients presenting with chronic low back pain and physical disability. When a surgical treatment option is warranted, patients now have the option of undergoing lumbar total disc replacement (TDR), a relatively new procedure that is designed to replace lumbar fusion, the traditional surgical intervention for degenerative disc disease. The lumbar TDR procedure has demonstrated clinical efficacy equivalent to that of lumbar fusion, although concern remains about the longevity, safety, and costs related to the procedure. These issues were addressed in three separate observational studies using administrative claims data. The first study estimated the revision burden and economic revision burden of lumbar TDR. The second study examined the lumbar TDR hybrid procedure, where both a lumbar TDR and lumbar fusion are performed simultaneously. No observational data have been reported on the frequency, cost, and diagnostic indications related to the TDR hybrid procedure. The third study mapped the geographic variation of procedural rates of lumbar TDR. Previous research has found substantial geographic variation in lumbar spine surgery rates and a similar analysis of lumbar TDR variation has yet to be reported. The present series of studies found the revision burden and economic revision burden of lumbar TDR to be similar to data reported for this procedure from the mid-2000s, though the overall occurrence of the procedure appears to have declined. The economic revision burden made this a lower-cost procedure than lumbar fusion, with a tradeoff in terms of revision burden being higher for lumbar TDR. The lumbar TDR hybrid procedure was found to make up approximately 16% of the total number of TDR procedures, involving much higher costs than a single-level TDR procedure. Finally, geographic variation of the procedural rate of lumbar TDR varied dramatically across the U.S., surpassing the variation observed in lumbar fusion surgery. Limitations of the observational data used in these studies are described. Recommendations for future observational research are offered as well. Finally, implications for these studies on practice guidelines and reimbursement policies are provided.