Monetary and Fiscal Interactions in the US During the 1940s Cliometrica (2019)
It is generally assumed that the buildup of liquid assets in the U.S. during WWII played a large role in generating postwar economic activity. Contrary to this assumption, I establish that military contract spending during the war slowed down the growth of bank balance sheets at the state level during the period 1940-1955. State level bank balance sheets are 10.8 cents smaller per $1 of total military spending by 1949 and 5.8 cents smaller by 1955. This is primarily driven by slower growth of demand deposits. The adjustment on the asset side is largely through reserves and Treasury holdings. Local lending also grows more slowly after the war, but this decrease is relatively small and temporary. This suggests that the local real economy was largely insulated from the slower growth in deposits by the wartime build up of paper assets. Historical evidence points to the fact that slower growth of deposits is likely driven by a relative decline in demand for deposits by large corporations in war industries.
Appendix A: Subperiod Analysis
Appendix B: Bank Charters
Appendix C: Regression Tables
Appendix D: Federal Reserve Districts
Appendix E: Tests of Historical Endogeniety
Appendix F: Data Sources
Appendix G: Reserves and Interbank Deposits
The Asymmetric Response of the Economy to Tax Changes Before and After 1980 North American Journal of Economics and Finance (Forthcoming)
This paper explores the asymmetry in the response of GDP to tax shocks before and after 1980 as first noted in Romer and Romer (2010). I find that there are two main reasons why output responds more strongly to tax shocks before 1980 than after. First, a greater sensitivity of the effect of tax shocks on output to the state of the economy before 1980 explains about half of the difference between periods. Second, before 1980 the effect of tax shocks on households is indirect and lowers total personal income and nondurable goods consumption. After 1980 tax shocks affect personal tax payments directly, causing disposable income and savings to change. This finding affirms Romer and Romer’s hypothesis that households are more likely to consumption smooth. However, I find that households after 1980 consumption smooth in response to a change in their direct tax burden not, as Romer and Romer posit, because they have greater access to financial services.
Appendix A and B
Income and WWII Contract Spending
Using a simple fixed effects specification a detailed, underutilized, state level panel of personal income from the BEA I find that state level WWII military contract spending has a complex effect on both contemporary and postwar state level personal income. I find that wartime manufacturing income grows faster in states that receive military contract spending. However, after the war interest dividend and rental income grows more slowly in states that received military contracts. Nonmanufacturing earnings also grow more slowly after the war, though this effect is likely driven by region specific effects. I also find that wartime investment contract spending permanently increases the growth of manufacturing earnings and wartime military infrastructure contract spending permanently increases government payrolls.
Employment, Aggregate Demand, and the Reconversion Recession
This paper traces employment in war and nonwar industries during the transition to a peacetime economy after WWII. The relatively modest increase in unemployment from 0.9\% in 1944 to 4.3\% in 1947 is one indicator of the mildness of the postwar recession. While employment data points to a role for construction and retail industries in the softness of the postwar recession, other factors are likely more significant. Government employment, GI Bill educational benefits, and exports all played major roles in keeping unemployment low in the late 1940s. Further, the decline in employment after the war is concentrated in war industries and is permanent. The postwar export boom is concentrated in war industries as well. Empirical specifications looking at the effect of WWII spending on local economies that do not take these historical facts into consideration are likely biased.
Labor and Institutions in the Wake of the World Wars
This paper compares and contrasts the experience of organized labor during the strike waves after both world wars with a focus on strikes in the steel industry. Both wars saw significant increases in union membership. However, the 1919 strike wave ended in decisive defeats for organized labor that led to a decade long lull in union activity and little wage growth. The result was an increase in inequality by 1929 not seen again until 2008. In contrast, the strike wave of 1946/1947 occurred under vastly different institutional circumstances. These changes were across government, labor and business. This post-WWII strike wave helped maintain some of the wartime gains by labor and helped set the stage for the Great Compression, a generation long flattening of measured inequality.
We find that the share of state level income captured by the top 1% of income earners falls by 1.73pp in response WWII contract spending. We also find a postwar “permanent” decline in top 1% income shares of 0.85pp attributable to wartime spending. This is half the average state level decline in inequality during the war and one fifth of the average postwar decline in inequality relative to 1940. We take this as evidence that WWII military spending induced permanent labor market changes in manufacturing that reduced inequality relative to its level in 1940.