Search theory has proven useful for describing and modeling many different Economic interactions. More broadly, the frictions it can account for are key to understanding many of the outcomes in observational data. This dissertation explores how such frictions affect the choices of economic agents, and what this implies for measurement and aggregate behavior. In doing so I am interested in the essential connection between theory and empirics: models help the researcher think about the interpretation of data, and data offers a ruler with which to assess the performance of models.

The first chapter explores how worker perceptions about job finding affect where unemployed searchers choose to apply for jobs and how this impacts the behavior of key labor market variables. Motivated by the observed prevalence of optimistic bias in searcher expectations about job finding, I develop a model of directed search where workers are uncertain about the matching technology, but can learn about it with experience searching for employment. I find that misperceptions dampen the volatility of labor market variables. For example, the standard deviation of the unemployment rate decreases by 10% when accounting for this uncertainty, while its correlation with labor productivity decreases by 12%. I show that optimistically biased job finding expectations increase wages by 0.3%, but also increase the average unemployment spell length by 1.5 weeks and the unemployment rate by 0.6pp.

In the second chapter, joint with Christine Braun and Peter Rupert, we study how the presence of on-the-job leisure, that is, non-work at work, drives a wedge between measured hours of work and actual hours of work. If actual hours of work are lower than measured hours, productivity and wages are actually higher than those calculated by the Bureau of Labor Statistics, for example. Technological innovations, while making an hours of work more valuable, may also make it easier to engage in on-the-job leisure. We document the extent of on-the-job leisure and embed it into a model of technological change with imperfect monitoring to examine its effect on productivity and wages. Using the American Time Use Survey we show that for those workers who engage in OJL spend about 50 minutes per day doing so. We use the model to create a time series of actual hours of work and calculate actual output per hour.

In the third chapter, joint with Daniel Cullen, we ask, "How does the sharing economy affect traditional lodging markets?" The advent of platforms such as Airbnb in 2008 has introduced a new channel of market interaction between those with space and those who seek it. This allows for transactions of lodging services that might otherwise be underutilized. This paper develops a framework to help think about how peer-to-peer transactions interact with traditional rental markets, and what this means for property managers and tenants. Specifically, we examine how the introduction of sharing platforms (e.g. Airbnb) affect the listing decisions of vacant property managers and the lodging choices of dwelling seekers. The model features landlords who choose where to list vacant properties and renters who search for lodging. Renters can be either short or long-term, referencing how long they wish to occupy the property. Sharing platforms give landlords the option of accessing these short-term renters who would otherwise occupy hotels, affecting traditional, long-term renters. We find that Airbnbs decrease hotel prices by about $24 while they increase average rents by $39 per month.

This dissertation consists of three chapters. In the first chapter I analyze the impact of unemployment insurance on personal bankruptcy filings. In response to the COVID-19 pandemic, the US federal government expanded unemployment benefits and extended coverage to previously ineligible workers through Pandemic Unemployment Assistance (PUA) and Federal Pandemic Unemployment Compensation (FPUC). I study the impact of pandemic era unemployment benefit terminations on personal bankruptcy filings in states that ended these federal programs early. Using nonlinear difference-in-differences, I find that Chapter 13 filings increased between 14-15% on average. These findings are consistent with previous work that documented an increase in employment and a decrease in household financial security associated with the termination of pandemic unemployment insurance programs.

In the second chapter, which is joint work with Molly Schwarz, we investigate the extent to which the provision of Medicaid to previously ineligible, low-income childless adults affects their household composition. Using a staggered adoption difference-in-differences design on an urban sample of individuals with less than a high school degree, we find that 26 to 39 year olds experience a significant 4.2% decline in the number of individuals living in the household, which is due to living with fewer extended family members. At the same time, 26 to 39 year olds experience a relatively smaller decline in the number of rooms (1.8%), leading to a 3.1% reduction in the level of household crowding, as measured by persons per bedroom. These reductions in household crowding are strongest for Hispanic individuals and those living in areas with above-median housing costs. In comparison, there are no significant impacts on household composition for 40 to 64 year olds as a consequence of the policy.

In the third chapter I explore the estimation and interpretation of the coefficient on a treatment variable in two-way fixed effects regressions with group-by-time fixed effects. Using the decomposition from de Chaisemartin and D'Haultfoeuille (2020) I show that while this design can be successful in leveraging within group variation to estimate the parameter of interest, in cases with heterogeneous treatment effects it leads to the well-documented weighting issues that arise in the canonical two-way fixed effects regression when there is variation in treatment timing. As opposed to the canonical two-way fixed effects regression, however, when group-by-time fixed effects are included weighting issues can arise when there is variation in treatment timing within groups, across groups, or both. I also show that in this setting it is possible to include groups in which all units share the same treatment sequence or that include only subgroups with constant treatment, and that observations in these groups do not contribute to the estimation of the parameter of interest. In this case, under heterogeneous treatment effects the average treatment effect on the treated cannot be obtained, but rather in certain cases the average treatment effect on the treated for groups that have within group variation in treatment sequences is obtained under the assumption that parallel trends hold in every group. While this is typically still a parameter of interest, including groups in which all units share the same treatment sequence or that include only subgroups with constant treatment incorrectly reduces the size of the standard errors and therefore biases inference.

This dissertation consists of three works which consider how frictional labor market models align with data, as well as how labor market data should be combined and used in applications of theoretical models. What unifies these essays is the underlying goal to further the consolidation of theoretical labor market models with empirical observations of worker and firm behavior in order to better understand and influence policy.

The first essay asks the question: How do changes in the minimum wage affect criminal activity? I answer this question by describing a frictional world in which a worker's criminal actions are linked to his labor market outcomes. The model is calibrated to match labor market outcomes and crime decisions of workers from the National Longitudinal Survey of Youth 1997, and shows that the relationship between the aggregate crime rate and the minimum wage is U-shaped. The results from the calibrated model as well as empirical evidence from county level crime data and state level minimum wage changes from 1995 to 2014 suggest that the crime minimizing minimum to median wage ratio for 16-19 year olds is 0.91. However, the welfare maximizing minimum to median wage ratio is 0.87, not equal to the crime minimizing value.

The second essay, joint with Ben Griffy, Bryan Engelhardt and Peter Rupert, asks the question: Is the arrival rate of a job independent of the wage that it pays? We answer this question by testing how, and to what extent, unemployment insurance changes the hazard rate of leaving unemployment across the wage distribution using a Mixed Proportional Hazard Competing Risk Model and data from the 1997 National Longitudinal Survey of Youth. Controlling for worker characteristics we reject that job arrival rates are independent of the wages offered. We apply the results to several prominent job-search models and interpret how our findings are key to determining the efficacy of unemployment insurance.

Finally, the third essay, joint with Finn Kydland and Peter Rupert argues that not all hours are created equally. In this paper we present a method for adjusting aggregate hours to account for changes in the quality of hours worked. Average human capital has rapidly increased since 1980 as better educated cohorts enter the workforce and the baby boomers continue to work and gather experience. We construct an aggregate labor input series from 1979 to adjust for changes in the experience and education levels of the workforce using the Current Population Survey's Outgoing Rotation Groups. We show that a decrease in labor productivity beginning in 2004, the "productivity slowdown," is understated by 23 percentage points when using aggregate hours instead of labor input to calculate productivity, and that 80% of the average quarterly growth rate of labor productivity can be attributed to increases in education and experience since 2004.

The first chapter studies the role of search frictions and preference shocks and how they lead to sorting heterogeneous agents in the labor market. I develop a stochastic sorting framework where the equilibrium exhibits sorting across types. This feature helps to understand empirical labor market trends, such as mismatch and wage inequality. In particular, I find that differences in productivity, entry cost, market noise, and unemployment benefits have contributed to changes in observed sorting patterns and income inequality in the U.S. labor market.

In the second chapter, I examine the role of various skill levels in a frictional labor market with heterogeneous workers and firms. The economy consists of workers with different skill levels and firms with diverse skill requirements for job vacancies. By assumption, workers qualify for jobs with lower skill requirements, though productivity is lower due to mismatch. In the model, workers may or may not choose to match firms with lower job skill requirements. This has implications on equilibrium unemployment, wage inequality, and the optimal mix of job openings.

The third chapter studies the decision-making behavior of agents who are simultaneously searching in dual markets, namely the labor market and the marriage market. When considering the marriage decision in the single agent labor search model, I find that individual searchers have similar reservation wage behavior that affects wage dynamics and marriage outcomes under risk-neutral preferences.

This reports studies what contributes to bank failure during the Great Recession and after the

introduction of the Dodd-Frank Act using survival analysis. I show that measures of CAMELS

ratings such as capital adequacy and solvency play large roles. I found that while a high

efficiency ratio increases risk of failure for financial institutions, banks can mitigate this by

focusing the burden of non-interest expenses on salary. While small banks (defined as having

below one billion dollars in assets) had an overall lower failure risk, they also had higher relative

risks of failure if they were inefficient than risks for large banks. However, small banks are

shown to reap the rewards of a high salary ratio more than large banks. Salary ratio influenced

post-Dodd-Frank bank failure more than it did pre-Dodd-Frank.

This dissertation explores how households allocate their financial resources across expenditure categories and over time. The interaction between income changes and consumption is important for the well-being of households. This interaction exhibits great heterogeneity across goods as well as across wealth. Notably, when the households are liquidity-constrained, their budget allocation decision is altered substantially. This dissertation analyzes the interaction between household income and intertemporal budget allocation with an emphasis on liquidity constraints.

In the first chapter, I explore how the liquidity constraints affect the household intertemporal allocation of consumption differentially across wealth and across expenditure categories. This chapter comprises two sub-chapters for the theoretical and empirical evaluation of the effect of binding liquidity constraints on intertemporal budget allocation. Further, the chapter compares the effect of liquidity constraints on healthcare expenditure with the effect on non-health consumption in particular on food consumption. I extend a standard incomplete markets model with a health capital in the felicity function. Theoretically, I show that households reduce their healthcare expenditure due to the binding liquidity constraints in the current period, whereas expenditure declines in the next period due to the expected binding constraints one period ahead. I use the extended model to test the incidence of binding liquidity constraints with a linearized Euler equation. Empirically, I show that the test of liquidity constraints for healthcare expenditure reveals different implications than a standard test of liquidity constraints for nondurable consumption. In particular, current binding constraints and expected binding constraints lead to the opposite direction of bias when the liquidity constraints are omitted. The resulting overall bias depends on which constraint has a stronger effect. Moreover, the correlation between income and healthcare expenditure varies significantly between asset poor and rich families, more than the elasticity of non-health consumption among wealth quintiles. Altogether, my findings show that the effects of liquidity constraints are heterogeneous across households and across expenditure categories.

In the second chapter, I estimate the income elasticity of consumption for various expenditures. Estimating income elasticity of consumption is found to be a challenging task. The causal impact of income changes on expenditure is hard to measure due to the endogeneity of the treatment variable income. I use a shift-share instrumental variable design `a la Bartik [1991] to mitigate the endogeneity concerns by exploiting variation due to local labor market exposure to aggregate shocks. I estimate the income elasticity of consumption that results from the changes in national employment growth in industries weighted with regional employment share of the industry. I find an average elasticity of total household consumption in the ranges between 0.4 to 0.53 depending on the construction of the instrument. Food consumption elasticity ranges between 0.11 to 0.2 though is not significantly estimated. Of particular interest for income elasticity estimates is the household out-of-pocket healthcare expenditure which has an elasticity of around 3.14 to 3.59. This finding adds to the discussion of health spending being a luxury good with an elasticity above one which is found in aggregate cross-country or time-series estimates. I find elasticities above one using household-level micro consumption and regional employment growth data, whereas micro studies usually conclude health expenditure elasticities around zero.

In the third chapter, I provide a detailed analysis of household wealth and portfolio allocation. This chapter presents several stylized facts on how households allocate wealth among asset classes, how portfolio allocation changes over the lifecycle and over the business cycle, and how portfolio and income are related. The chapter combines various survey data to show that household income and portfolio allocation are highly correlated especially for middle-income households. Asset accumulation has an inverted-V shape over the lifecycle whereas debt is front-loaded in working ages. Income and consumption follow a hump-shaped over the lifecycle. Old households hold assets in liquid forms. As for the business cycle, the Great Recession has devastating effects on the welfare of households such that both networth and consumption declined, and poverty rates increased. The effect is more severe for non-white, low educated, and female-headed households.

The Mortensen-Pissarides model with unemployment benefits and taxes has been able to account for the variation in unemployment rates across countries but does not explain why geographical mobility is very low in some countries (on average, three times lower in Europe than in the U.S.). We build a model in which both unemployment and mobility rates are endogenous. Our findings indicate that an increase in unemployment benefits and in taxes does not generate a strong decline in mobility and accounts for only half to two-thirds of the difference in unemployment from the U.S. to Europe. We find that with higher commuting costs the effect of housing frictions plays a large role and can generate a substantial decline in mobility. We show that such frictions can account for the differences in unemployment and mobility between the U.S. and Europe.

This dissertation examines how unfavorable macroeconomic conditions — climate change and credit constraints — influence economic activities and its implication for social welfare. In expecting climate change, people adapt their behavior to reduce its harm. Despite its importance, the optimal interplay between individual adaptation and climate policies is understudied. The first chapter studies optimal taxation in a general equilibrium model in which households compete against final goods producers for carbon-intensive intermediate goods. I theoretically show that an increase in market demand for carbon-intensive goods increases energy producers' marginal profit, which leads to more energy production and higher carbon emissions. In a calibrated model with heat-related mortality and cooling loads as an example, I find that the mortality social costs of carbon in 2020 are underestimated by 7% if such feedback is not considered.

Many U.S. municipalities finance public projects by issuing bonds secured by taxes. But emerging climate risks can impair the ability to pay debts due to the tax base loss from property damages. Rating agencies have been considered trustworthy arbiters of creditworthiness in financial markets. The second chapter studies how much climate risks are factored into credit ratings using wildfires and school district finances in California. Using historic wildfires as a source of salience shocks, I find that school districts with 0.08% of the tax base at risk of burning over the next 30 years face 3.22% lower credit ratings. This finding implies that the integration of climate risks in credit ratings can encourage policymakers to adapt to climate change to avoid higher borrowing costs.

The third chapter, joint with Kieran Walsh, uses the U.S. Census Bureau Household Pulse Survey to study heterogeneity in the spending response to stimulus checks. We find that while the fraction of households who use the payment for spending declines in pre-COVID incomes for the 2020 payments, this pattern changes into a U-shaped one in 2021. This is because, during crises, liquidity constraints are binding for poorer households due to a tightening of lending standards, making them anxious to consume. On the other hand, in a normalized economy, many poorer households borrow to avoid the inconvenience of not meeting unexpected expenditure needs such as car repairs or medical bills. When faced with a random splash of cash, these households use it for saving as they had previously dissaved to meet their unexpected expenditure needs. The macro state-dependence of spending propensity distribution is crucial to understanding the propagation of fiscal and monetary shocks.