A key question in macroeconomics is the government's ability to stimulate economic activity through expansionary fiscal policy. How much economic activity results when the government increases spending by one dollar, and how does the economic and institutional context affect the answer to that question? This dissertation uses a variety of empirical techniques to explore aspects of this question using historical data on U.S. military spending.
In chapter one I use state-level variation in war production spending to measure the fiscal multiplier during World War II, and examine how features of the wartime economy influenced the size of the fiscal multiplier. Chapter two focuses on how the measurement of government spending influences the estimated size of the multiplier. I introduce a new time series measure of aggregate defense spending. In chapter three I return to World War II, but this time examine the effects of wartime military spending on the post-war economy, establishing causal evidence for its role in driving the immediate post-war boom.
In chapter one I use war production spending to quantify the idiosyncratic factors affecting estimates of the fiscal multiplier during World War II. World War II is often viewed as a quintessential example of government spending stimulating the economy, and is interesting both because it was such a significant economic event and because it strongly influences estimates of the multiplier whenever it is included in the sample. Newly digitized war supply contract data allow me to construct state-level panel data on U.S. spending for 1940-45 and examine state-level outcomes. Using state-level variation I estimate a relative multiplier of 0.25 to 0.3, depending on the estimation approach. This implies an aggregate multiplier of roughly 0.3 to 0.4 given wartime economic conditions. I find small employment effects: an additional job-year is associated with $165,000 to $255,000 of spending (in 2015 dollars), also depending on the estimation approach. I also find evidence that the effects of stimulus were systematically larger in states that had lower employment levels pre-war. To explain why the stimulative effects of war spending were so small, I look for guidance from the historical narrative. I show that unique features of the wartime economy significantly reduced the stimulative impact of wartime spending. Conversion from civilian manufacturing to war production reduced the initial stimulus from war production. At least 75 percent of the income generated by war spending went into increased saving and income taxes, implying that the add-on effects from increased consumption were minimal in the short run.
Chapter two focuses on how the measurement of government spending influences the estimated fiscal multiplier. Economists have previously focused on measuring shocks to expectations rather than the measurement of government spending itself. My approach is driven by the observation that government spending is a long and complex process. Specifically, I introduce an alternative measure of government spending, called budget authority, which uses authorizations to measure the government's commitment to spend. Budget authority is established annually as part of the congressional budget process, and is readily available from 1976 onward. I use historical budget publications to construct defense budget authority for 1938 to 1975, extending the available data backwards by several crucial decades.
Using annualized data (for purposes of comparison) to estimate the aggregate fiscal multiplier using shocks to defense spending, budget authority produces similar point estimates to the traditional NIPA measure, but much more precisely estimated. Budget authority is conceptually different from the best-known measure of shocks to anticipated defense spending, Ramey's narrative measure, particularly in how it measures shocks to expectations and how it treats uncertainty. Budget authority implies an aggregate fiscal multiplier of 0.8, while Ramey's narrative measure implies a much smaller fiscal multiplier, around 0.1. Budget authority shows consumption responses to spending more clearly than other available measures, and also picks up strong investment responses over a one-year time horizon. Ramey's narrative measure shows significant investment responses over all time horizons up to three years. While shocks to all three measures predict strong responses in total government spending, it appears that both budget authority and Ramey's measure understate the response of government spending due to timing differences between those measures and NIPA. The definition of spending mostly closely aligned to national accounting is subtly different from the definition that is most relevant for measuring the stimulative effect of government spending. Thus using the NIPA definition of spending creates a downward bias in measuring the fiscal multiplier. A fourth measure of spending, budget outlays, allows me to estimate a lower bound for the magnitude of this bias. When this bias is corrected, budget authority implies an aggregate fiscal multiplier of 1.3 to 1.4, and potentially as large as 1.4 to 1.6.
Chapter three examines the influence of World War II spending in the U.S. on household consumption and savings in the immediate post-war years (1946-1949). Chapter three uses geographic variation in war spending to measure the effects of World War II spending on household consumption and savings behavior after the war ended and rationing was relaxed. I find that compared to households in locations receiving less war spending, similar households in locations which received more war spending were significantly more likely to purchase both cars and houses in the immediate post-war years. These households also had higher liquid asset holdings and, conversely, higher total debt. With the exception of debt, all of these effects were stronger for households headed by an individual age 45-64, which was the age cohort most likely to have worked in war production.
