In this paper we re-examine the recent evidence that technology shocks do not produce business cycle patterns in the data. We first extend Gali's (1999) work, which uses long-run restrictions to identify technology shocks, by examining whether the identified shocks can be plausibly interpreted as technology shocks. We do this in three ways. First, we derive additional long-run restrictions and use them as tests of overidentification. Second, we compare the qualitative implications from the model with the impulse responses of variables such as wages and consumption. Third, we test whether some standard "exogenous" variables predict the shock variables. We find that ilshocks, military build-ups, and Romer dates do not predict the sholck labeled "technology." We then show ways in which a standard DGE model can be modified to fit Gali's finding that a positive technology shock leads to lower labor input. Finally, we re-examine the properties of the other key shock to the system.

Understanding the effects and transmission mechanism of fiscal policy is crucial for designing efficient policies and enhancing efficacy assessments. This dissertation leverages recent trends in accurate data measurement and network data usage to enhance our understanding of fiscal policy's transmission mechanisms.

Chapter 1 presents evidence that early impacts of defense news shocks on GDP stem from increased business inventories as contractors prepare for new defense contracts. These do not impact government spending until payment-on-delivery, occurring 3 quarters later. Data on defense procurement obligations reveal that contract awards Granger-cause shocks to government spending identified via Cholesky decomposition, highlighting that Cholesky shocks miss early inventory changes, resulting in underestimated multipliers.

Chapter 2 introduces novel data on defense contracts to study government purchase effects in the US, developing new facts about their transmission mechanism. The data includes a new quarterly series of US military prime contracts starting from 1947. These contracts are shown to be exogenous to output fluctuations and provide robust measures for timing shocks, eliminating the need for narrative analysis. Findings suggest that positive shocks to defense contracts enhance output, consumption, labor earnings, and productivity, supported by "learning-by-doing" effects in military production.

Chapter 3 investigates fiscal consolidations' effects in the United States and their network propagation. Using a narrative approach, we identify exogenous fiscal adjustments and employ a spatial autoregression (SAR) model to differentiate direct and network effects. Tax-based consolidations exhibit a more pronounced recessionary impact than expenditure-based ones, with about 27% of tax-based effects attributable to network spillovers versus 11% for expenditure-based adjustments. This difference in network effects highlights the stronger propagation of tax increases compared to spending cuts.

Chapter 1 measures gender differences in college major choice and major switching behavior using a nationally representative sample of beginning postsecondary students. I find that switchers overall have worse graduation outcomes and take longer to earn their degrees than non-switchers. GPA appears to be a driver of switching: for students initially enrolled in a given field, the average GPA of female and male switchers is lower than the overall average for women and men, respectively. However, for most fields, the difference between the average GPA of female switchers and the average overall GPA is small. This suggests that women might interpret the signal sent by GPA in different ways than men. I then show that, conditional on GPA and initial STEM enrollment, women are more likely to change majors than men.

Chapter 2 builds a model of group-based beliefs and human capital specialization. Individuals belonging to a particular group choose to work or study in heterogeneous fields. This maps to the scenario where group type corresponds to gender, and the specialization decision is an individual’s college major choice. Agents form initial beliefs about their unknown abilities based on existing group outcomes, and update these beliefs as they study. Therefore, women may form their initial beliefs about their abilities in each field by considering how many women have specialized in that field in the past. These differences in initial beliefs can cause men and women to respond to signals about their ability in different ways, which can ultimately drive gender gaps in college major choice. This model explores a mechanism for why,conditional on GPA and initial STEM enrollment, women are more likely to change major than men. If women form their initial beliefs about their abilities based on existing group outcomes, then women in historically underrepresented fields (like in STEM) might be more likely to switch than men at a given GPA level.

Chapter 3 explores the literature on technological adoption in agricultural climate change adaptation. Specifically, this chapter reviews two recent approaches to studying climate change adaptation in agriculture: panel data methods and spatial general equilibrium models.

This dissertation studies three policy-oriented macroeconomic questions. The first chapter examines whether traditional monetary policy in the U.S. becomes less effective when foreign governments accumulate large amounts of Treasury debt. I estimate a macro-finance model and find foreign official purchases have shifted the entire yield curve down. This suggests the increasing presence of international factors in U.S. financial markets influences the Federal Reserve's interest rate policy. The second chapter asks why low-skilled men are less likely to be employed relative to high-skilled men and why this differential has increased since the 1970s. I build and calibrate a labor-search model and find a demand shift and job separations are the main drivers of employment inequality, while a supply shift had no robust effects, and search frictions actually reduced employment inequality since the 1970s. The third and final chapter studies why wages of newly hired workers are more pro-cyclical than wages of workers who do not switch jobs. We construct a novel measure of occupational mismatch by comparing a newly hired worker's current skill profile to his previous skill profile. Including our measure of occupational mismatch in standard wage regressions can account for half of the new hire wage cyclicality previously documented in the literature.

Chapter one examines the cyclical behavior of low-income versus high-income household price indices and documents two new facts: (1) during recessions prices rise more for products purchased relatively more by low-income households (necessities); (2) the aggregate share of spending devoted to necessities is counter-cyclical. I present a mechanism where adverse macroeconomic shocks cause households to shift expenditure away from luxuries toward necessities, which leads to higher relative prices for necessities. I embed this mechanism into a quantitative model which explains over half of the cyclical variation in necessity prices and shares. The results suggest that low-income households are hit twice by recessions: once by the recession itself and again as their price index increases relative to other households.

Chapter two presents evidence that the high estimated MPCs from the leading household studies result in implausible macroeconomic counterfactuals. Using the 2008 tax rebate as a case study, we calibrate a standard medium-scale New Keynesian model with the estimated micro MPCs to construct counterfactual macroeconomic consumption paths in the absence of a rebate. The counterfactual paths imply that consumption expenditures would have plummeted in spring and summer 2008 and then recovered when Lehman Brothers failed in September 2008. We use narratives and forecasts to argue that these paths are implausible. We go on to show that reasonable modifications of the model result in general equilibrium forces that dampen rather than amplify micro MPCs. We also show that estimators of the average treatment effect yield smaller micro MPC estimates than the standard two-way fixed effects estimators. The combination of smaller micro MPCs and dampening general equilibrium forces implies general equilibrium consumption multipliers that are below 0.2.

In chapter three, I construct novel measures of household-level inflation and show that an increase in a household's personal inflation rate leads to a persistent increase in their price index. Households respond to a personal inflation shock by decreasing nominal consumption, which means their real consumption falls more than one-for-one. I also find a statisically robust relationship between inflation dispersion (the variance of household inflation rates) and the level of absolute aggregate inflation.

This dissertation examines economies that may occasionally enter periods of crisis. I first develop a model of asset pricing in the presence of frictions to financial intermediation. This model generates recurrent financial crises due to its strong non-linear dynamics. Next, I develop methodological tools for analyzing these types of highly non-linear dynamic equilibrium models. I then apply these tools to a theory of housing boom-bust cycles driven by endogenous shifts in lending standards.

Chapter 1 introduces a model of asset pricing in the presence of agency frictions between savers and financial intermediaries. This model can generate asymmetric price movements where asset values suddenly collapse during a financial crisis. During normal times, intermediaries arbitrage away excess returns on assets and traditional asset pricing conditions hold. During a financial crisis, the net worth of intermediaries limits their ability to borrow from savers and they are unable to arbitrage away excess returns. Since their net worth depends on realized asset values, collapsing prices further tighten borrowing constraints leading to a large and sudden collapse in asset values.

In chapter 2, I introduce a local approach to solving highly non-linear models, generalizing perturbation to handle the class of piecewise smooth rational expectations models. First, I formalize the notion of an endogenous regime by introducing a regime-switching equilibrium (RSE) concept. This framework uses non-linear model features to explain macroeconomic regime changes, and makes the distribution of the regime an equilibrium object instead of imposing an external regime-switching structure. Then, I demonstrate how to apply perturbation within a slackened model, approximate the policy functions associated with a given belief about the regime, and solve for the equilibrium regime distribution using backwards induction. This approach (1) accounts for expectational effects due to the probability of regime change; (2) provides a framework for modeling regime-switching from first principles; and (3) connects macroeconomic theory to reduced-form regime-switching econometric models.

Chapter 3 develops a theory of housing boom-bust cycles driven by endogenous shifts in lending standards. The key friction is asymmetric information about default risk, implying that the economy occasionally and endogenously switches between two credit regimes. These regimes differ by whether or not lenders use income verification to screen the marginal homebuyer. A switch from the ``screening'' regime to the ``pooling'' regime leads to rapid home price appreciation, a collapse in down-payment requirements, and a reduction in income documentation --- consistent with the shift in lending standards during the US housing boom. The episode ends in a foreclosure crisis once fundamentals revert and the screening regime returns. The theory predicts patterns for debt accumulation and mortgage spreads consistent with existing micro evidence.

This dissertation is composed of three chapters, all related to the effects of highway spending. By considering the construction of the Interstate Highway System (IHS), Chapter 1 asks how big, if any, are returns to constructing new highways? I find that the biggest threats to identification are endogeneity and anticipation. To overcome endogeneity I note that a state’s initial population and area shares played an important role in determining the assignment of interstate highway funds. To overcome the anticipation I use news to identify the timing of shocks. I combine my solutions in an IV local projection framework, as in Ramey and Zubairy (2018), and estimate a cumulative relative multiplier of 1.8 and a discounted relative multiplier of 2.3 at the 15 year horizon. I then extend my specification to allow for spillover effects, and find evidence of zero spillovers across states.

Chapter 2 asks what is the relation between the relative and the aggregate multiplier from transportation spending? If we take as given a relative multiplier of 1.8 and no across-state spillovers from highway spending, what can we say about the aggregate multiplier? To answer this question, I build a multi-region model that captures two key mechanisms from building highways: (1) bilateral trade costs decrease, which leads to positive spillovers; and (2) TFP shifts up, which leads to a factor-stealing effect and negative spillovers. When only one region is shocked, positive and negative spillovers can cancel each other out. The calibrated model predicts an aggregate multiplier of 2.54.

Finally, Chapter 3 provides evidence of the popular belief that incumbent parties can attract votes by increasing government spending. A simple OLS analysis provides a downward biased estimate of this relationship as incumbents target spending on areas where they are less popular. To overcome this problem I employ an IV methodology that exploits the 1947 plan of the IHS. I find that receiving one extra highway mile causes the share of votes to the gubernatorial incumbent party to increase by 0.65 percentage points.