Abstract
In his papers during the lead up to the birth of the European Monetary Union, Obstfeld considered whether the countries forming the EMU were sufficiently similar to survive a single monetary policy—and more importantly, whether they had the capacity to adjust to asymmetric shocks given a single monetary and exchange rate policy. The convention at the time was to take the United States as the baseline for a smoothly functioning currency union. We expand on stylized facts in the literature to illustrate how stratification in local labor market outcomes appears far more persistent in later years than 3 decades ago in the context of what (Obstfeld and Peri in Econ Policy 13(26):205–259, 1998) call non-adjustment in unemployment rates. We then extend the currency union literature by adding an additional consideration: differences in regional cyclical sensitivity. Using measures of cyclicality and Obstfeld–Peri-type non-adjustment, we explore the characteristics of places that can get left behind when local labor markets respond differently to national shocks and discuss implications for policy.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
The primary input into estimation of the state-level variables are the Current Population Survey (CPS) and the Current Employment Statistics (CES). The county-level variables are not merely imputed from state-level variables, but involve a great deal of additional information. For these, the BLS uses not only CPS and CES, but also the American Community Survey and Census population estimates which contain more detailed data, the Quarterly Census of Employment and Wages (based on data collected from business), and county-level administrative data on unemployment claims.
While county-level dispersion looks to be lower post-1990 than pre-1990, it is important to note that there is a structural break in the LAUS data in 1990 which make comparisons of these two eras precarious. Still, the gap between state- and county-level dispersion may be narrowing since 1990 and have a cyclical component. This raises the question of whether state-level policies may be at play.
The R-squared from this simple regression of state unemployment rates in 1996 on unemployment rates in 1986 shows that the unemployment rate in 1986 alone could explain 24 percent of the variation in unemployment rates across states in 1996.
Figures 12 and 13 in Appendix D illustrate further. The deterioration of unemployment rate outcomes for places with large Black populations in some ways should be a surprise, since the ratio of Black to White unemployment rates nationally is somewhat lower(\(\sim\)1.8–2.0) today than it was in the early 1980s(\(\sim\)2.4). At the same time, the trend is aligned with Derenoncourt et al. (2023) evidence that the racial wealth gap has widened in the US since the 1980s.
Mundell (1961) himself says, “In the real world, of course, currencies are mainly an expression of national sovereignty, so that actual currency reorganization would be feasible only if it were accompanied by profound political changes. The concept of an optimum currency area therefore has direct practical applicability only in areas where political organization is in a state of flux (p.661)."
These two components of UGAP are labelled UNRATE and NROU in the Federal Reserve Economic Data database maintained by the Federal Reserve Bank of Saint Louis.
In the "Appendix", we report the raw coefficients and the standard errors clustered at the state level.
In Appendix Table 3, we reproduce the group-level differences in cyclical sensitivity shown by Aaronson et al. (2019). Furthermore, if contrasting the top quintile to bottom seems an inappropriate comparison for the Black White gap which compares a 1% population to a majority population, one could contrast the top quintile (0.5) to the bottom half of counties (\(-0.4\)) and still find a gap on par with the Black White gap.
One might also ask whether counties with higher unemployment rates or higher cyclical sensitivity recover more slowly. Many studies computing half-lives in macroeconomics use quarterly or higher-frequency data. Our data are annual and we have only 29 periods, so we can get only a very rough estimate at best. We compute a half-life using the formula \(H_i=\frac{\ln (0.5)}{\ln \rho }\), where \(\rho\) is the coefficient when regressing the local unemployment rate on its lag, \(\rho\)as in Taylor (2001). A bar chart of average half-life by decile is included in Fig. 15 in Appendix 9 and shows considerable heterogeneity in the half-life across counties, indicating dispersion in the rates of recovery from shocks. The mean is 3.25 years, with a standard deviation of 1.51, a minimum of 0.28 and a maximum of 15.18. We see some slight positive correlation of the half-life with \(\alpha _i\) and \(\beta _i\), but with very little explanatory power. Explanatory power is greater (R-squared = 11.9%) for Factor 1 from the PCA analysis below.
In the "Appendix", we report the raw coefficients and the standard errors clustered at the state level.
Specifically, we combine the \(u_t\) vectors into a \(T \times I\) matrix we call U, with T the number of years in our sample and I the number of counties. We then extract the principal components as described by Ferroni and Canova (2021). We use their toolkit to implement the PCA in MATLAB, computing the eigenvectors and eigenvalues of \(U'U\), then computing the factors as the first 10 eigenvectors multiplied by U.
Share of college graduates is significantly correlated with cyclical sensitivity to UGAP (\(\beta _i\)) in Table 2 once controls and state fixed effects are added, but not on its own, and the impact is somewhat muted (a one standard deviation increase increases \(\beta _i\) by 0.03).
Specifically, correlation between the first factor and UGAP increases from 0.72 1990–2000 to 0.96 2000–2018.
In contrast, places where the China shock industries were moving out 1960–1980 appears somewhat more sensitive to Factor 1, though the correlation is not robust to inclusion of the full set of controls.
Data for black male and white male (series LNS14000007 and LNS14000004) unemployment rates downloaded directly from Labor Force Statistics from the Current Population Survey. BLS data on unemployment rates for overall black and overall white and by educational group downloaded via Federal Reserve Economic Data database (LNS14000003, LNS14000006, LNS14027659, LNS14027660, LNS14027689, LNS14027662), as were the series for the natural rate of unemployment and national unemployment rate (series names NROU and UNRATE).
References
Aaronson, S. R., M. C. Daly, W. L. Wascher, and D. W. Wilcox. 2019. Okun revisited: Who benefits most from a strong economy. Brookings Papers on Economic Activity, 333–371.
Almeida, A., A. Galiano, A.A. Golpe, and J.M. Martín. 2020. Regional unemployment and cyclical sensitivity in Spain. Letters in Spatial and Resource Sciences 13: 187–199.
Amior, M., and A. Manning. 2018. The persistence of local joblessness. American Economic Review 108: 1942–1970.
Autor, D., D. Dorn, and G. Hanson. 2021. On the persistence of the China shock. Brookings Papers on Economic Activity, 381–447.
Autor, D.H., D. Dorn, and G.H. Hanson. 2013. The China syndrome: Local labor market effects of import competition in the United States. American Economic Review 103(6): 2121–2168.
Beyer, R.C., and F. Smets. 2015. Shocks and migration. Economic Policy 30(84): 643–682.
Byers, J.D. 1990. The cyclical sensitivity of regional unemployment: An assessment. Regional Studies 24(5): 447–453.
Chetty, R., and N. Hendren. 2018. The impacts of neighborhoods on intergenerational mobility I: Childhood exposure effects. Quarterly Journal of Economics 133(3): 1107–1162.
Cipolla, C. 1967. Money, prices, and civilization in the Mediterranean world. New York: Gordian Press.
Cox, R.J. 2010. Crime, incarceration, and employment in light of the great recession. Review of Black Political Economy 37(1): 283–294.
Curtis, K.J., and L.J. Kulscar. 2019. Rural demography, volume 22 of handbooks of sociology and social research. Berlin: Springer.
Derenoncourt, E., C. H. Kim, M. Kuhn, and M. Schularick. 2023. Wealth of two nations: The U.S. racial wealth gap, 1860–2020. Working paper.
Diamond, R., and E. Moretti. 2021. Where is standard of living the highest? local prices and the geography of consumption. Manuscript, Stanford University.
Diamond, R. 2016. The determinants and welfare implications of US workers’ diverging location choices by skill: 1980–2000. American Economic Review 106(3): 479–524.
Donaldson, D., and R. Hornbeck. 2016. Railroads and American economic growth: A “market access’’ approach. Quarterly Journal of Economics 131(2): 799–858.
Eriksson, K., K.N. Russ, J.C. Shambaugh, and M. Xu. 2021. Trade shocks and the shifting landscape of U.S. manufacturing. Journal of International Money and Finance 111: 102407.
Fajgelbaum, P., and C. Gaubert. 2020. Optimal spatial policies, geography, and sorting. Quarterly Journal of Economics 135(2): 959–1036.
Ferroni, F. and F. Canova. 2021. A Hitchhiker’s guide to empirical macro models.
Freeman, R.B. 1972. Labor economics. New Jersey: Prentice-Hall.
Freeman, D. 2011. On (not) closing the gaps: The evolution of national and regional unemployment rates by race and ethnicity. The Review of Black Political Economy 39: 1–18.
Hardy, B. L., T. D. Logan, and J. Parmon. 2018. The historical role of race and policy for regional inequality. The Hamilton Project Framing Pa.
Hardy, B., R. Samudra, and J. Davis. 2019. Cash assistance in America: The role of race, politics, and poverty. The Review of Black Political Economy 46: 306–324.
Holzer, H., S. Raphael, and M. Stoll. 2006. Employers in the boom: How did the hiring of less-skilled workers change during the 1990s? The Review of Economics and Statistics 88: 283–299.
Howland, Marie. 1984. Age of capital and regional business cycles. Growth and Change 15 (2): 29–37.
Hoynes, H. 1999. The employment, earnings, and income of less skilled workers over the business cycle. NBER working paper.
Hyclak, T., and J. Stewart. 1995. Racial differences in the unemployment response to structural changes in local labor markets. The Review of Black Political Economy 23: 29–42.
Kocherlakota, N. and T. Krueger. 1998. Why do different countries use different currencies? IMF working paper no. 98-7.
Li, N. 2022. Innovation clusters and local brain drain. Manuscript, University of California, Davis.
Liu, J. 2021. Some cities are paying people up to \$16,000 to move there. CNBC.com.
Mill, J. 1894. Principles of political economy, vol. II. New York: John W. Parker.
Modestino, A.S., D. Shoag, and J. Ballance. 2016. Downskilling: Changes in employer skill requirements over the business cycle. Labour Economics 41: 333–347.
Mundell, R.A. 1961. A theory of optimum currency areas. American Economic Review 51: 657–665.
Obstfeld, M. 1997. Europe’s gamble. Brookings Papers on Economic Activity 2: 241–317.
Obstfeld, M. 1998. EMU: ready or not. Princeton: International Economics Section, Departement of Economics, Princeton University, Princeton Essays in International Economics.
Obstfeld, M., and G. Peri. 1998. Regional nonadjustment and fiscal policy. Economic Policy 13(26): 205–259.
Okun, A.M. 1973. Upward mobility in a high-pressure economy. Brookings Papers on Economic Activity 4: 207–262.
Rappoport, Jordan. 2009. The increasing importance of quality of life. Journal of Economic Geography 9 (6): 779–804.
Republicans, J. E. C. 2019. Losing our minds: Brain drain across the United States. SCP report 2-19.
Rossi-Hansberg, E., P.-D. Sarte, and F. Schwartzman. 2019. Cognitive hubs and spatial redistribution. NBER working paper 26267.
Taylor, Alan M. 2001. Potential pitfalls for the purchasing-power-parity puzzle? Sampling and specification biases in mean-reversion tests of the law of one price. Econometrica 69 (2): 473–498.
von Lockette, N.D., and W. Spriggs. 2016. Wage dynamics and racial and ethnic occupational segregation among less-educated men in metropolitan labor markets. Review of Black Political Economy 43: 35–56.
Willingham, C. Z. and O. Ajilore. 2021. Effective and ethical rural development. In A. Dumont and D. P. Davis (Eds.), Investing in rural prosperity. St. Louis, Missouri, USA: Federal Reserve Bank of St. Louis.
Wolfers, J. 2019. Comment on: Okun revisited: Who benefits most from a strong economy. Brookings Papers on Economic Activity, 385–404.
Acknowledgements
First draft: July 2022. We thank Yongyeon Oh and Manho Kang for research assistance. We thank Òscar Jordà, Takuya Ura, and USITC; as well as David Autor and Gordon Hanson for comments on early thoughts. Shambaugh’s contribution to this paper preceded his time at the US Treasury Department. This paper does not reflect the views of the US Treasury or US Government. The views expressed herein are solely those of the authors and do not necessarily reflect the views of the Federal Reserve Bank of San Francisco, or the Federal Reserve System. All errors are ours.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
On behalf of all authors, the corresponding author reports no conflicts of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
A. Revisiting Aaronson et al. (2019)
In this literature on cyclical sensitivity broken down by worker characteristics, the basic methodological setup is a regression of the gap in the unemployment rate between a group g at time t and some other group used as a benchmark on a measure of the business cycle:
For example, a commonly estimated gap is the Black–White unemployment gap and which is then regressed on a measure of the business cycle (UGAP, typically the gap between the national unemployment rate and the CBO measure of the long run unemployment rate). A focus on gaps helps remove concerns around long-run trends and stationarity that have played a major role in some prior examinations of regional cyclical sensitivity discussed above.
1.1 A.1 Data
To provide context for our county-level analysis, we recreate the (national) Aaronson et al. (2019) regressions using Eq. (4) and data from the US Bureau of Labor Statistics,Footnote 15 here with the sample period 1976Q1-2019Q4 for black–white gaps, and 1992Q1-2019Q4 for education gaps due to a shorter series for measures of education.
For the male Black and male white unemployment rate, Labor Force Statistics from the Current Population Survey are also available at https://www.bls.gov/cps/data.htm. Click on link to the Data Finder for Labor Force Statistics and enter the series identifiers into the search box: LNS14000007 and LNS14000004. Adjust the sample to begin at 1972.
For other unemployment rates, one can use the same portal, or we used the FRED aggregator portal for convenience. Series identifiers for Black and white overall unemployment rate gaps are LNS14000006 and LNS14000006. Series identifiers for Less than High School Diploma (“<HS”), High School Graduates No College (“HS”), Some College or Associate Degree (“Some college”), and Bachelor’s Degree (“BA”) are, respectively, LNS14027659, LNS14027660, LNS14027689, and LNS14027662.
In all regressions, the national unemployment rate (quarterly, seasonally adjusted) and long-term natural rate (quarterly) are from FRED, with series identifiers UNRATE and NROU. Since the long-term natural rate already is smoothed, it is not seasonally adjusted.
1.2 A.2 Cyclicality by Demographic Group
Table 3 shows that the coefficient \(\beta _g\) in Eq. (4) for the gap between unemployment rates of Black versus White men is 0.9. For the overall Black–White gap, it is 0.7. In such a formulation, a \(\beta _g\) of zero means the two groups’ unemployment rates move up and down together when the economy overall has a rising or falling unemployment rate. That is, there is no change in the gap between groups as the national unemployment rises or falls. A \(\beta\)-coefficient of nearly 1 for the Black–White unemployment rate gap means that the Black unemployment rate rises a full point more than the White rate whenever the economy overall has an unemployment rate that rises 1 point above the CBO long run rate. The constant (\(\alpha\)) in the regression shows the base level of unemployment for the groups. The \(\beta _g\)-coefficient for the gap between workers with a less than a high school degree versus those with a college degree generate a \(\beta\)-coefficient of 1.0 while those with a high school degree have a \(\beta\) of 0.6, closing the gap somewhat.
The constant (\(\alpha _g\)) in the Black–White regressions is 6, indicating a Black–White unemployment gap of 6 percentage points on average when the economy is overall at the CBO long run rate. It is somewhat higher than the constant for the gap between workers without a high school degree versus those with a Bachelor’s degree, though the differences are not statistically distinguishable.
B. Summary Statistics for US County-Level Data
See Table 4.
C. Data Detail
1.1 C.1 Figures 1 and 2: State Unemployment Rates
Data downloaded from the Federal Reserve Economic Data (FRED) aggregator portal at https://fred.stlouisfed.org. Series identifiers are 4 letters, with the first two letters the state’s two-letter postal abbreviation and the second two letters “UR” for unemployment rate. Figures constructed in Excel.
1.2 C.2 County-Level Unemployment Rates and County Characteristics
County-level unemployment rates for 1970 and 1980 are from the US Bureau of the Census County Data Books in digital format through ICPSR 2896, DS76 and DS78, available at https://www.icpsr.umich.edu/web/ICPSR/studies/2896/datadocumentation. County characteristics are all set to 1970 values and also from ICPSR 2896 DS76 (1972 County Data Book). Unemployment rates for 1990 onward are from the US Bureau of Labor Statistics Local Area Unemployment Statistics (LAUS). Download County Data Tables by year at https://www.bls.gov/lau/#tables).
D. Supplemental Figures
See Figs. 11, 12, 13, 14, 15, 16, 17, 18, 19 and 20.
Labor market outcomes in counties, by education levels of population in 1970. Source: Education levels by county in 1970 and unemployment rates in 1970 and 1980 from US Bureau of the Census County Data Books, via University of Michigan ICPSR 2896; unemployment rates by county 1990–2016 from US Bureau of the Census LAUS
Labor market outcomes in counties, by racial composition of population in 1970. Notes: Fraction of Black residents by county in 1970 and unemployment rates in 1970 and 1980 from US Bureau of the Census County Data Books, via University of Michigan ICPSR 2896; unemployment rates by county 1990–2016 from US Bureau of the Census LAUS. Note that a large share of counties have very small Black populations, so being in the lowest quintile of percentage Black population is not very different from the second quintile
Labor market outcomes in counties, by fraction of employment in manufacturing industries in 1970. Notes: Fraction of workers in manufacturing by county in 1970 and unemployment rates in 1970 and 1980 from US Bureau of the Census County Data Books, via University of Michigan ICPSR 2896; unemployment rates by county 1990–2016 from US Bureau of the Census LAUS
Percentage of variance accounted for by each factor. Notes: First 10 factors extracted from the panel of county unemployment rates as described in Section 5.2
Heterogeneity in estimated half-life of adjustment in county-level unemployment
Half-life of adjustment in county-level unemployment and \(\alpha _i\)
Half-life of adjustment in county-level unemployment and \(\beta _i\)
Half-life of adjustment in county-level unemployment and \(\lambda _1i\)
Half-life of adjustment in county-level unemployment and \(\lambda _2i\)
Factor 2 loadings and \(\alpha _i\)
E Supplemental Tables
See Tables 5, 6, 7, 8, 9, 10, 11 and 12.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Russ, K.N., Shambaugh, J.C. & Singh, S.R. Currency Areas, Labor Markets, and Regional Cyclical Sensitivity. IMF Econ Rev 72, 152–195 (2024). https://doi.org/10.1057/s41308-023-00223-w
Published:
Issue Date:
DOI: https://doi.org/10.1057/s41308-023-00223-w