Skip to main content
Log in

Beveridge Curve Shifts across Countries since the Great Recession

  • Published:
IMF Economic Review Aims and scope Submit manuscript

Abstract

The paper documents the shift in the Beveridge curve in the United States since the Great Recession. It argues that a decline in quits, the relatively poor performance of the construction sector, and the extension of unemployment insurance benefits have largely driven this shift. The paper then introduces a method to estimate fitted Beveridge curves for other OECD countries for which data on vacancies and employment by job tenure are available. It shows that Portugal, Spain, and the United Kingdom also experienced rightward shifts in their Beveridge curves. Besides the United States, these are among the countries with the highest house price and construction employment declines in the sample.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Figure 1
Figure 2
Figure 3
Figure 4
Figure 5

Similar content being viewed by others

Notes

  1. Several recent studies have estimated the magnitude of this decline in match efficiency. Among them Borowczyk-Martins, Jolivet, and Postel-Vinay (2013), Barnichon and others (2012), Davis, Faberman, and Haltiwanger (2012), and Sedláček (2012).

  2. Nonlinear regressions are the most common empirical method of studying Beveridge curves. See Nickell and others (2001), Valletta (2005), Bouvet (2013), and Bonthuis, Jarvis, and Vanhala (2013) for example.

  3. Labor turnover, as in hires and separations, includes job-to-job transitions. This is in contrast to Budd, Levine, and Smith (1987) and Barnichon and Figura (2010) that define the Beveridge curve in terms of the flow-steady-state of unemployment in which inflows into and outflows from unemployment are equal.

  4. In Section II, when we present our results for our sample of countries, we show that the Beveridge curve that we construct for the United States does not line up very well with the data when one does not take into account the cyclicality of separation and assumes that α s =0.

  5. One way to interpret this equation is as the log-linearized reduced form equation of a model of frictional unemployment with unemployment, U t , being the only state variable.

  6. This is slightly different from the definition used in JOLTS where v t =V t /(V t +E t ), which would make the vacancy rate conceptually equivalent to the unemployment rate. The problem with this definition is that it assumes that the unit of measurement of the vacancy statistic is defined as a job comparable with those in the employment number. It turns out that in some of our source data this is not the case. Using the vacancy to employment ratio as a measure of the vacancy rate means that cyclical fluctuations in the rate are not dependent on the unit of measurement of vacancies.

  7. The application of OLS for the estimation of matching functions, (4), has been criticized (Borowczyk-Martins, Jolivet, and Postel-Vinay, 2013) because of the potential for endogeneity of the number of vacancies posted with respect to the level of match efficiency, ɛh,t. Of course, endogeneity bias is also a concern for the application of OLS to the reduced form equation (5). Unfortunately, the length of the annual time series, at maximum 17 observations, on which the cross-country results we present are based prevents us from applying econometric methods aimed at correcting for endogeneity.

  8. This fitted curve is based on the estimates of Equations (4) and (5) using pre-2008 data. The exact specification of the equations used in Barnichon and others (2012) is slightly different because they are tailored to the monthly JOLTS and CPS data available in the United States. See Barnichon and others (2012) for details. The functional forms presented in this paper are the ones we subsequently implement in our cross-country analysis using annual OECD data.

  9. This is the peak of the Congressional Budget Office’s natural rate of unemployment estimates in 1978.

  10. There is a growing theoretical literature that provides models in which match efficiency endogenously fluctuates over the business cycle due to the variation in the composition of the pool of unemployed. Sedláček (2012) and Ravenna and Walsh (2013) both provide models in which firms are more picky in selecting workers during times of high unemployment than during times of tightness in the labor market.

  11. The only other country for which we found data on hires and separations is Japan, where annual time series for hires and separations have been published as part of the Survey of Employment Trends done by the Japanese Ministry of Health, Labour, and Welfare since 1952.

  12. Similar to the unemployment inflow and outflow measures in Elsby, Hobijn, and Şahin (2013) which are affected by the duration dependence of unemployment outflow hazards, our estimates here are affected by the tenure dependence of job separation rates. We discuss in the Appendix how this mainly affects the level of the turnover measures and not the cyclical fluctuations that are of interest for the rest of our analysis.

  13. For Australia, we supplement the OECD data with data from the Australian Bureau of Statistics’ Labour Mobility release.

  14. The U.S. data are taken from the Displaced Workers and Job Tenure supplement that is part of the Current Population Survey (CPS). We use the data available every two years since 1996. Before 1996 the job tenure question was phrased in an ambiguous way that makes it hard to interpret it as a continuous job tenure.

  15. This is the case for Spain and Portugal. For Portugal the fluctuations in the vacancy data published by the OECD and Eurostat are very different. We use the Eurostat data here because those give the smallest deviation from the Beveridge curve, which we already argue to be large.

  16. We do not report the intercepts, μ h and μ s , because they depend on the units of measurement of vacancies which differ across countries. This makes comparison of these parameters hard to do.

  17. It is worth noting that the results for the two countries with the lowest estimated α h ’s are not based on annual data on employment by job tenure. Instead, the results for the United States and Japan are based on biannual data and a direct measure of hires respectively. The range 0.12–0.83 is similar to the range of estimates surveyed by Petrongolo and Pissarides (2001, Table 3). The estimate of 0.27 for the United States, is much smaller than the one of 0.58 obtained using the monthly JOLTS data, that are the source of Figure 2, and is also smaller than the estimated matching function elasticity for the United States of 0.72 in Shimer (2005), which uses an estimate of the outflow rate out of unemployment rather than hires.

  18. This finding is in line with the results in Hagedorn and Manovskii (2013) and Daly, Hobijn, and Wiles (2012) which both suggest that job-to-job transitions are crucial for understanding cyclical fluctuations in real wages.

  19. Ignoring the cyclicality of separations for the construction of the fitted JOLTS-based Beveridge curve of Figure 1 yields a very similar curve to the gray dashed line in Figure 4.

  20. Since the fitted curve is a locus constructed based on estimates of Equations (4) and (5) and the average distance for the sample observations does not have to be zero this is not literally a standard error of the residuals. However, in the rest of our analysis we will informally treat it as such.

  21. See Elsby, Hobijn, and Şahin (2013) for a discussion of these adjustment rates and for estimates for Spain and Portugal.

  22. The limited number of observations in our sample and the fact that the distances, D t , are not really regression residuals prevent us from doing a more formal statistical analysis.

  23. This comes with the caveat that in many countries the unemployment rate continued to increase after 2011, the end of our sample.

  24. Other cross-country analyses of these movements in the (u,v)-curve are Budd, Levine, and Smith (1987) and Bouvet (2013) for example.

  25. The replication files to Elsby, Hobijn, and Şahin (2010, Figure 5) include results for a fully demographically adjusted unemployment rate. These results suggest that the adjusted for the 2010 composition of the U.S. labor force the U.S. unemployment rate would have been a bit more than 2 percentage points lower than the published number.

  26. In addition, they find that in countries where home ownership rates increased the Beveridge curve drifted outward. They interpret this as significant cross-country evidence of an effect of labor mobility on the Beveridge curve.

  27. In addition to the change in wage bargaining, Pissarides (2006) also discusses the change in monetary policy in the United Kingdom in 1993. He argues, on page 218, that the outward movement in the British Beveridge curve in the 1980s can be better interpreted as a prolonged counterclockwise loop than an actual shift.

  28. Seasonality is actually also an issue for the share of workers employed longer than a year.

References

  • Aaronson, Daniel, Bhashkar Mazumder, and Shani Schechter, 2010, “What is Behind the Rise in Long-Term Unemployment?,” Federal Reserve Bank of Chicago Economic Perspectives, Vol. 34, No. 2, pp. 28–51.

    Google Scholar 

  • Abraham, Katharine G. and Lawrence F. Katz, 1986, “Cyclical Unemployment: Sectoral Shifts or Aggregate Disturbances?,” Journal of Political Economy, Vol. 94, No. 3, pp. 507–522.

    Article  Google Scholar 

  • Akerlof, George A., Andrew K. Rose, and Janet L. Yellen, 1988, “Job Switching and Job Satisfaction in the U.S. Labor Market,” Brookings Papers on Economic Activity, Vol. 1988, No. 2, pp. 495–582.

    Article  Google Scholar 

  • Barnichon, Regis and Andrew Figura, 2010, “What Drives Movements in the Unemployment Rate? A Decomposition of the Beveridge Curve,” Finance and Economics Discussion Series 2010-48 (Federal Reserve Board of Governors).

  • Barnichon, Regis, Michael W.L. Elsby, Bart Hobijn, and Ayşegül Şahin, 2012, “Which Industries are Shifting the Beveridge Curve?” Monthly Labor Review, June, pp. 25–37.

  • Blanchard, Olivier J. and Justin Wolfers, 2000, “The Role of Shocks and Institutions in the Rise of European Unemployment: The Aggregate Evidence,” Economic Journal, Vol. 110, No. 462, pp. C1–C33.

    Article  Google Scholar 

  • Blanchard, Olivier J. and Lawrence H. Summers, 1988, “Beyond the Natural Rate Hypothesis,” American Economic Review: Papers and Proceedings, Vol. 78, No. 2, pp. 182–187.

    Google Scholar 

  • Bonthuis, Boele, Valerie Jarvis, and Juuso Vanhala, 2013, “What’s Going on Behind the Euro Area Beveridge Curve?” mimeo, ECB.

  • Borowczyk-Martins, Daniel, Grégory Jolivet, and Fabien Postel-Vinay, 2013, “Accounting For Endogeneity in Matching Function Estimation,” Review of Economic Dynamics, Vol. 16, No. 3, pp. 441–451.

    Article  Google Scholar 

  • Bouvet, Florence, 2013, “The Beveridge Curve in Europe: New Evidence Using National and Regional Data,” Applied Economics, Vol. 44, No. 27, pp. 3585–3604.

    Article  Google Scholar 

  • Brügemann, Björn, 2008, “What Elasticity of the Matching Function Is Consistent with U.S. Aggregate Labor Market Data?” mimeo, VU University Amsterdam.

  • Budd, Alan, Paul Levine, and Peter Smith, 1987, “Long-Term Unemployment and the Shifting U-V Curve: A Multi-Country Study,” European Economic Review, Vol. 31, No. 1–2, pp. 296–305.

    Article  Google Scholar 

  • Chow, Gregory C., 1960, “Tests of Equality between Sets of Coefficients in Two Linear Regressions,” Econometrica, Vol. 28, No. 3, pp. 211–222.

    Article  Google Scholar 

  • Costain, James, Juan F. Jimeno, and Carlos Thomas, 2013, “Employment Fluctuations in a Dual Labor Market,” Bank of Spain Working Paper 1013.

  • Daly, Mary C., Bart Hobijn, Ayşegül Şahin, and Robert G. Valletta, 2012, “A Search and Matching Approach to Labor Markets: Did the Natural Rate of Unemployment Rise?,” Journal of Economic Perspectives, Vol. 26 (Summer), pp. 3–26.

    Article  Google Scholar 

  • Daly, Mary C., Bart Hobijn, and Rob Valletta, 2011, “The Recent Evolution of the Natural Rate of Unemployment,” FRBSF Working Paper 2011-05.

  • Daly, Mary C., Bart Hobijn, and Theodore S. Wiles, 2012, “Dissecting Aggregate Real Wage Fluctuations: Individual Wage Growth and the Composition Effect,” FRBSF Working Paper 2011-23.

  • Davis, Steven J., R. Jason Faberman, and John C. Haltiwanger, 2012, “Recruiting Intensity during and after the Great Recession: National and Industry Evidence,” American Economic Review: Papers and Proceedings, Vol. 102, No. 3, pp. 584–588.

    Article  Google Scholar 

  • Diamond, Peter A., 2013, “Cyclical Unemployment, Structural Unemployment,” NBER Working Paper #18761.

  • Dickens, William T., 2009, “A New Method for Estimating Time Variation in the NAIRU,” in Understanding Inflation and the Implications for Monetary Policy: A Phillips Curve Retrospective, ed. by Jeff Fuhrer, Yolanda K. Kodrzycki, Jane Sneddon Little and Giovanni P. Olivei (Cambridge, MA: MIT Press) pp. 205–228.

    Google Scholar 

  • Dickens, William T. and Robert K. Triest, 2012, “Potential Effects of the Great Recession on the U.S. Labor Market,” B.E. Journal of Macroeconomics, Vol. 12, No. 3, pp. 1–41.

    Article  Google Scholar 

  • Elsby, Michael W.L., Bart Hobijn, and Ayşegül Şahin, 2010, “The Labor Market in the Great Recession,” Brookings Papers on Economic Activity, Vol. 41, No. 1, pp. 1–48.

    Article  Google Scholar 

  • Elsby, Michael W.L., Bart Hobijn, and Ayşegül Şahin, 2013, “Unemployment Dynamics in the OECD,” The Review of Economics and Statistics, Vol. 95, No. 2, pp. 530–548.

    Article  Google Scholar 

  • Farber, Henry S. and Robert G. Valletta, 2011, “Extended Unemployment Insurance and Unemployment Duration in The Great Recession: The U.S. Experience,” mimeo, Federal Reserve Bank of San Francisco and Princeton University, June.

  • Fleischman, Charles A. and John M. Roberts, 2011, “From Many Series, One Cycle,” Finance and Economics Discussion Series 2011-46, Board of Governors of the Federal Reserve System.

  • Fujita, Shigeru, 2010, “Effects of the UI Benefit Extensions: Evidence from the Monthly CPS,” Working Paper No. 10-35 (Federal Reserve Bank of Philadelphia, November).

  • Hagedorn, Marcus and Iourii Manovskii, 2013, “Job Selection and Wages over the Business Cycle,” American Economic Review, Vol. 103, No. 2, pp. 771–803.

    Article  Google Scholar 

  • Hall, Robert E., 2005, “Job Loss, Job Finding, and Unemployment in the U.S. Economy over the Past Fifty Years,” in NBER Macroeconomics Annual 2005, ed. by Mark Gertler and Kenneth Rogoff. (Cambridge, MA: MIT Press) pp. 101–37.

    Google Scholar 

  • Hobijn, Bart, 2012, “The Industry-Occupation Mix of U.S. Job Openings and Hires,” FRBSF Working Paper 2012-09.

  • Karahan, Fatih and Serena Rhee, 2013, “Geographical Reallocation and Unemployment during the Great Recession: The Role of the Housing Bust,” Federal Reserve Bank of New York Staff Reports 605.

  • Lazear, Edward P. and James R. Spletzer, 2012a, “Hiring, Churn, and the Business Cycle,” American Economic Review: Papers and Proceedings, Vol. 102, No. 3, pp. 575–79.

    Article  Google Scholar 

  • Lazear, Edward P. and James R. Spletzer, 2012b, “The United States Labor Market: Status Quo or A New Normal?” mimeo, Stanford University.

  • Molloy, Raven, Christopher L. Smith, and Abigail Wozniak, 2011, “Internal Migration in the US,” Journal of Economic Perspectives, Vol. 25, No. 3, pp. 173–196.

    Article  Google Scholar 

  • Mortensen, Dale T., 1994, “The Cyclical Behavior of Job and Worker Flows,” Journal of Economic Dynamics and Control, Vol. 18, No. 6, pp. 1121–1142.

    Article  Google Scholar 

  • Nakajima, Makoto, 2012, “A Quantitative Analysis of Unemployment Benefit Extensions,” Journal of Monetary Economics, Vol. 59, No. 7, pp. 686–702.

    Article  Google Scholar 

  • Nickell, Stephen and Jan van Ours, 2000, “Falling Unemployment: The Dutch and British Cases,” Economic Policy, Vol. 26, No. S1, pp. 136–180.

    Article  Google Scholar 

  • Nickell, Stephen, Luca Nunziata, Wolfgang Ochel, and Glenda Quintini, 2001, “The Beveridge Curve, Unemployment and Wages in the OECD from the 1960s to the 1990s—Preliminary Version,” Centre for Economic Performance Discussion Paper 502 (London: London School of Economics and Political Science).

  • Organization for Economic Cooperation and Development (OECD). 2009, “How Do Industry, Firm and Worker Characteristics Shape Job and Worker Flows?,” in OECD Economic Outlook 2009: Tackling the Jobs Crisis. (Paris: OECD).

  • Organization for Economic Cooperation and Development (OECD). 2012, “How Does Spain Compare?,” in OECD Employment Outlook 2012. (Paris: OECD).

  • Patterson, Christina, Ayşegül Şahin, Giorgio Topa, and Gianluca Violante, 2013, “Mismatch Unemployment in the U.K.,” mimeo.

  • Petrongolo, Barbara and Christopher A. Pissarides, 2001, “Looking into the Black Box: A Survey of the Matching Function,” Journal of Economic Literature, Vol. 39, No. 2, pp. 390–431.

    Article  Google Scholar 

  • Pissarides, Christopher A., 2006, “Unemployment in Britain: A European Success Story,” in Structural Unemployment in Europe: Reasons and Remedies, ed. by Martin Werding, (Cambridge, MA: MIT Press), Chapter 9, pp. 209–236.

    Google Scholar 

  • Pollock, Alex J., 2010, “Testimony on the Comparison of International Housing Finance Systems,” Senate Committee on Banking, Housing, and Urban Affairs, 29 September, 2010.

  • Ravenna, Federico and Carl E. Walsh, 2013, “Labor Market Flows with Skill Heterogeneity in a Monetary Policy Model,” Journal of Money Credit and Banking, forthcoming.

  • Rothstein, Jesse, 2012, “Unemployment Insurance and Job Search in the Great Recession,” Brookings Papers on Economic Activity, Vol. 43, No. 2, pp. 143–196.

    Google Scholar 

  • Şahin, Ayşegül, Joseph Song, Giorgio Topa, and Gianluca Violante, 2013, “Mismatch Unemployment,” Working paper (New York: New York University).

  • Saint-Paul, Gilles, 1995, “The High Unemployment Trap,” Quarterly Journal of Economics, Vol. 110, No. 2, pp. 527–550.

    Article  Google Scholar 

  • Schulhofer-Wohl, Sam, 2012, “Negative Equity Does Not Reduce Homeowners’ Mobility,” FRB Minneapolis Quarterly Review, Vol. 35, (January) pp. 2–14.

    Google Scholar 

  • Sedláček, Petr, 2012, “Match Efficiency and Firms’ Hiring Standards,” Working paper, Journal of Monetary Economics.

  • Shimer, Robert, 1998, “Why Is the U.S. Unemployment Rate so Much Lower?,” NBER Macroeconomics Annual, Vol. 13, (July) pp. 11–16.

    Article  Google Scholar 

  • Shimer, Robert, 2005, “The Cyclical Behavior of Equilibrium Unemployment and Vacancies,” American Economic Review, Vol. 95, No. 1, pp. 25–49.

    Article  Google Scholar 

  • Shimer, Robert, 2012, “Reassessing the Ins and Outs of Unemployment,” Review of Economic Dynamics, Vol. 15, No. 2, pp. 127–148.

    Article  Google Scholar 

  • Sterk, Vincent, 2012, “Home Equity, Mobility, and Macroeconomic Fluctuations,” DNB Working Paper No. 265.

  • Tasci, Murat and John Lindner, 2010, “Has the Beveridge Curve Shifted?” Federal Reserve Bank Cleveland Economic Trends 08-10.

  • Valletta, Robert G., 2005, “Why Has the U.S. Beveridge Curve Shifted Back? New Evidence Using Regional Data,” Federal Reserve Bank of San Francisco Working Paper 2005-25.

  • Valletta, Robert G., 2010, “House Lock and Structural Unemployment,” mimeo, Federal Reserve Bank San Francisco.

  • Valletta, Robert G. and Katherine Kuang, 2010, “Extended Unemployment and UI Benefits,” Federal Reserve Bank of San Francisco Economic Letter 2010-12.

Download references

Authors

Additional information

*Bart Hobijn is a Senior Research Advisor at the Federal Reserve Bank of San Francisco and a part-time Professor at VU Amsterdam. Ayşegül Şahin is an Assistant Vice President at the Federal Reserve Bank of New York. The authors are grateful to Timothy Ni for his excellent research assistance. The views expressed in this paper solely reflect those of the authors and not necessarily those of the Federal Reserve Bank of San Francisco, Federal Reserve Bank of New York, nor those of the Federal Reserve System as a whole. Prepared for IMF Jacques Polak Annual Research Conference—November 8–9, 2012. The views in this article represent those of the authors and do not necessarily reflect those of the Federal Reserve Bank of New York, Federal Reserve Bank of San Francisco, or the Federal Reserve System as a whole.

Appendix

Appendix

Mathematical Derivations

The Beveridge Curve as an Implicit Function

The estimated matching function that fits the vacancy yield, H/V, as a function of the unemployment to vacancy ratio, U/V is given by:

This implies that hires per employee equals

Since the unemployment to employment ratio can be written in terms of the unemployment rate as

and because we define the vacancy rate as v t =V t /E t , the estimated matching function implies that we can write the ratio of hires per employee as

Similarly, since we estimate the cyclicality of separations using the iso-elastic functional form

we can write this in terms of the unemployment and vacancy rates as

Combining Equations (15) and (17), and the turnover steady-state condition (equation (3)) yields that the combinations of the unemployment rate and vacancy rate that are on the Beveridge curve need to satisfy

which is the equation in the main text. This defines the unemployment rate on the Beveridge curve as an implicit function of the vacancy rate.

Construction of Labor Turnover Estimates Using Annual Employment by Job Tenure Data

Our aim is to estimate the hires and separation rates for different countries over time. In this Appendix we show how the hires and separation rates can be estimated using data on employment growth and on the fraction of workers that report to have a job tenure smaller than a year. Throughout, time t is continuous and measured in years. We consider a year that runs from t∈[0,1].

The parameters that we are interested in estimating are the monthly separation rate, , and the the monthly hires rate, h. We assume that these rates are constant over the year. Our method uses data on employment at the beginning of the year in month 0, E0, and at the end of the year at t=1, E1. Moreover, we also have data on the fraction of workers that have been employed at a job with tenure of less than a year at the end of the year. This fraction allows us to calculate the number of workers at the year with a job tenure shorter than a year, which we denote by E1τ<1, where τ denotes job tenure in years.

With the assumed constant hires and separation rates, h and s, the number of employed persons evolves according to

Solving this differential equation results in the solution that

Hence, the growth rate of employment is the difference between the hiring rate, h, and the separation rate, s.

As for the number of workers hired over the period [0,t) who are still employed at time t, that is, E t τ≤t, this stock evolves according to

This is where the assumption that the separation rate is independent of job tenure comes in, because this equation assumes that the overall separation rate, , applies to those with job tenures in the interval .

Since we know the solution for , we can solve the nonhomogenous linear differential equation for to obtain that

This equation does not depend on the separation rate because of the assumption that the same rate applies to those workers with short tenures as well as long tenure.

This equation thus implies that the share of workers with a job tenure of a year or less as a fraction of total employment at time t=1 satisfies

This is a very useful result, because it implies that, even if there are no direct data on the hires rate as in JOLTS for the United States, the hires rate can be inferred from the share of workers with short tenures. In particular, the monthly hires rate can be calculated using

Given this estimate of the hires rate, the separation rate can be inferred from the survivor rate of those employed at the end of last period. That is,

These are the estimates of the annualized hires and separation rates, equivalent to the ones reported in JOLTS.

These flow rates can be transformed into estimates of actually flows of hires and separations. Let be the total number of hires over the period . These total flows are given by:

Similarly, the total number of separations over [0,t), S t , can be calculated as

These two flow measures satisfy the identity that the change in employment is the difference between total hires and total separations. That is,

The above derivations describe how data on the distribution of job tenures can be used to construct hires and separation measures that are conceptually the same as those measured on a monthly basis for the United States as part of JOLTS.

For the empirical implementation of these equations, we assume that E1 is the employment level observed in the year after which h and s are the turnover rates. This is because the job tenure surveys are done in the first half of the year and thus, those with a job tenure smaller than 12 months generally were hired in the previous year.

Similar Cyclicality of Turnover Measures from JOLTS and Employment by Job Tenure

For our calculation, we assume that workers of all job tenures separate at the same rate . We know this is, at best, an approximation since there is ample evidence that the separation rate from a job is declining in the tenure length. However, this tenure-dependence of the separation rate turns out to be of second-order importance for our calculation and ignoring it results in much simpler expressions.

In an extension of the methodology, we allowed for the separation rate of recently hired workers to be different from s. This requires using additional data on employment by job tenure shorter than 6 months which turns out to be noisy across countries and also depends on when in the year the job tenure distribution is measured.Footnote 28 When we applied this generalized method we got more noisy estimates of hires and separations.

What is, in particular, important for our calculation is not the level of hires and separations imputed from the data but the percent fluctuations over the business cycle. These fluctuations were highly correlated across both methodologies. Because of this, we present the result for the simpler methodology that is less subject to noise measures of employment with tenure less than half a year.

A final way to consider whether our method captures cyclical fluctuations in separations and hires reasonably well is to compare the biannual estimates for 1996 through 2010 for the United States with the direct measures from JOLTS. Figure A1 shows that the logs of the JOLTS and job-tenure-data-based estimates of hires and separations comove a lot. The correlations between the hires measures is 0.99 and between the separations series 0.91. Hence, the cyclical movements in job-tenure-based estimates of labor turnover are highly correlated with direct measures of labor turnover from JOLTS.

Figure A1
figure 6

Comparison of Labor-Turnover Measures for the United StatesSources: JOLTS, OECD, and authors’ calculations. Note: JOLTS measures are annual data and job-tenure-data-based measures are biannual. Series are log indices of hires and separations, normalized to equal 0 in 2007.

Taking Into Account Break in U.K. Vacancy Series in 1998

The U.K. vacancy statistics contain a structural break between 1998 and 2001. There are no data for 1999 and 2000 (the OECD data for this year have been interpolated). The units of measurement of vacancies in the United Kingdom seem to have changes between the pre- and postbreak data. To take this break into account we include a dummy, d t , in the estimated matching and separations functions which equals one for the prebreak data and zero for the postbreak period.

We then estimate the matching and separation functions as

and

Combining these two conditions yields two turnover-steady-state loci. One that holds for the prebreak definition of vacancies, that is, d t =1, and one that holds postbreak, that is, d t =0.

Of course, this way of taking into account the break assumes that the only thing that changed in terms of the vacancy measure is its unit of measurement and not its cyclical properties.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Hobijn, B., Şahin, A. Beveridge Curve Shifts across Countries since the Great Recession. IMF Econ Rev 61, 566–600 (2013). https://doi.org/10.1057/imfer.2013.18

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1057/imfer.2013.18

JEL Classifications

Navigation