Abstract
Hysteresis implies that shocks to unemployment have permanent effects. In this paper, we undertake a robust examination of the hysteresis hypothesis for 14 OECD countries by utilizing both linear and nonlinear unit root tests. After performing several linear unit root tests, we employ unit root tests with structural breaks and a Fourier test that allows for unknown breaks and nonlinear functional forms. To further examine the robustness of our findings, we additionally utilize tests that consider non-normal errors. Testing is undertaken using quarterly data for the time period 1983Q1–2013Q3. Overall, we find support for unit root hysteresis in 4 of the 14 countries, the traditional natural rate hypothesis in 3 countries, and the structuralist hypothesis (a natural rate with structural breaks) in 7 countries. In sum, we find that shocks to unemployment have permanent effects in 11 of the 14 countries, while the permanent effects are most often infrequent.
Similar content being viewed by others
Notes
The term “hysteresis” comes from an ancient Greek word roughly interpreted as “lagging.” The modern usage originates in 1881 in the physics literature on electromagnetic fields and is described as “effects which remain after the initial causes are removed” (Cross 1995, p. 6). Phelps (1972, p. xxiii) is perhaps the first to offer a description of unemployment hysteresis in the economics literature when he notes that “The natural unemployment rate at any future date will depend upon the course of history in the interim. Such a property is sometimes called hysteresis.” However, as noted by Amable et al. (1995), among others, a precise definition of unemployment hysteresis is unavailable and different definitions remain in the literature.
Røed (1997) provides a summary of factors that can lead to unemployment hysteresis.
We thank an Associate Editor for noting this.
At the time of writing, we are aware of three other papers that also test for hysteresis using data that include the recent US financial crisis and global recession (e.g., Cheng et al. 2012; Garcia-Cintado et al. 2015; Gali 2015). However, our testing methodologies and data differ from each of these papers in several ways. We discuss more on these papers in Sect. 2.
Given this outcome, the results in some earlier papers that test for hysteresis using an ADF-type endogenous break unit root test should be interpreted with caution since a rejection of the null may be spurious. See Lee and Strazicich (2001, 2003) and Perron (2006) for additional discussion on these issues.
First-differenced lagged terms \(\Delta ({\tilde{{S}}} )_{{t-i}},\; i = 1,\ldots ,p\) are included in (1) as necessary to correct for serial correlation. At each combination of break points, \(\lambda = (\lambda _{1}\), \(\lambda _{2})^ \prime \), where \(\lambda _{j} = {T}_{{Bj}}/{T},\; j=1, 2\), in the time interval [.1T, .9T] to eliminate end points, we begin with a maximum of \(p = 8\) lagged terms and examine the t statistic on the last term to see whether it is significantly different from zero at the 10% level (in an asymptotic distribution). If not significant, the term is dropped and the model re-estimated using \(p = 7\) terms, etc., until either the maximum lagged term is found or \(p = 0\). This type of general-to-specific procedure has been found to perform well as compared to other similar procedures (e.g., Ng and Perron 1995).
For example, the distribution of \({\tilde{\tau }}^{*}\) with one break (two breaks) is the same as that of the untransformed test using \(\lambda ={1}/{2}~ (\lambda _{1}={1}/{3} \hbox { and } \lambda _{2}={2}/{3})\). After transforming the data, the distribution of \(\tilde{\tau }^*\) will depend on the number of trend breaks, but not on their locations (e.g., Meng et al. 2016).
We use the critical values provided in Meng et al. (2016).
The frequency k is selected by using the data-driven grid search methodology proposed in Enders and Lee (2012), which involves estimating Eq. (6) for a range of values of k and choosing the value \((\hat{k})\) that minimizes the sum of squared residuals. See Enders and Lee (2012) for more detailed discussion.
In each case, we begin by applying the two-break LM unit root test to see whether each break is significant at the 10% level using the t-statistic in an asymptotic normal distribution. If only one or no breaks are significant, we then re-estimate the model using the one-break test. If no break is significant in the one-break test, no results are reported in column 2 of Table 3.
References
Amable B, Henry J, Lordon F, Topol R (1995) Hysteresis revisited: a methodological approach. In: Cross R (ed) The natural rate of unemployment: reflections on 25 years of the hypothesis. Cambridge University Press, Cambridge, pp 153–180
Arestis P, Mariscal IB-F (1999) Unit roots and structural breaks in OECD unemployment. Econ Lett 65:149–156
Becker R, Enders W, Lee J (2004) A general test for time dependencies in parameters. J Appl Econ 19:899–906
Becker R, Enders W, Lee J (2006) A stationarity test in the presence of an unknown number of smooth breaks. J Time Ser Anal 27(3):381–409
Blanchard OJ, Summers LH (1986) Hysteresis and the European unemployment problem, NBER macroeconomics annual, vol 1. MIT Press, Cambridge, pp 15–90
Blanchard OJ, Summers LH (1987) Hysteresis in unemployment. Eur Econ Rev 31:288–295
Camarero M, Tamarit C (2004) Hysteresis vs. natural rate of unemployment: new evidence for OECD countries. Econ Lett 84:413–417
Camarero M, Carrión-i-Silvestre JL, Tamarit C (2006) Testing for hysteresis in unemployment in OECD countries: new evidence using stationarity panel tests with breaks. Oxf Bull Econ Stat 68:167–182
Camarero M, Carrión-i-Silvestre JL, Tamarit C (2008) Unemployment hysteresis in transition countries: Evidence using stationarity panel tests with breaks. Rev Dev Econ 12(3):620–635
Chang T (2011) Hysteresis in unemployment for 17 OECD countries: stationarity test with a Fourier function. Econ Model 28:2208–2214
Cheng KM, Durmaz N, Kim H, Stern ML (2012) Hysteresis vs. natural rate of US unemployment. Econ Model 29:428–434
Cross R (1995) Hysteresis revisited: a methodological approach. In: Cross R (ed) The natural rate of unemployment: reflections on 25 years of the hypothesis. Cambridge University Press, Cambridge, pp 181–200
Elliott G, Rothenberg TJ, Stock JH (1996) Efficient tests for an autoregressive unit root. Econometrica 64(4):813–836
Enders W, Lee J (2012) A unit root test using a Fourier series to approximate smooth breaks. Oxf Bull Econ Stat 74(4):574–599
Friedman M (1968) The role of monetary policy. Am Econ Rev 58:1–17
Gali Jordi (2015) Hysteresis and the European unemployment problem revisited. Working Paper 21430, National Bureau of Economic Research, Cambridge
Gallant RA (1981) On the bias in flexible functional forms and an essentially unbiased form: the Fourier flexible form. J Econ 15(2):211–245
Garcia-Cintado A, Romero-Avila D, Usabiaga C (2015) Can the hysteresis hypothesis in Spanish regional unemployment be beaten? New evidence from unit root tests with breaks. Econ Model 47:244–252
Im KS, Lee J, Tieslau M (2005) Panel LM unit root tests with level shifts. Oxf Bull Econ Stat 67:393–419
Im KS, Schmidt P (2008) More efficient estimation under non-normality when higher moments do not depend on the regressors, using residual augmented least squares. J Econ 144:219–233
Jaeger A, Parkinson M (1994) Some evidence on hysteresis in unemployment rates. Eur Econ Rev 38:329–342
Lee CC, Chang CP (2008) Unemployment hysteresis in OECD countries: centurial time series evidence with structural breaks. Econ Model 25:312–325
Lee JD, Lee CC, Chang CP (2009) Hysteresis in unemployment revisited: evidence from panel LM unit root tests with heterogeneous structural breaks. Bull Econ Res 61(4):324–325
Lee J, Strazicich MC (2001) Break point estimation and spurious rejections with endogenous unit root tests. Oxf Bull Econ Stat 63(5):535–558
Lee J, Strazicich MC (2003) Minimum Lagrange multiplier unit root test with two structural breaks. Rev Econ Stat 8(4):1082–1089
Lee J, Strazicich MC (2013) Minimum LM unit root test with one structural break. Econ Bull 33(4):2483–2492
Leon-Ledesma MA, McAdam P (2004) Unemployment hysteresis and transition. Scott J Polit Econ 51(3):377–401
Meng M (2013) More powerful fourier unit root tests with non-normal errors. Working Paper, The University of Alabama
Meng M, Im KS, Lee J, Tieslau M (2014) More powerful LM unit root tests with non-normal errors. In: Sickles R, Horrace WC (eds) Festschrift in honor of Peter Schmidt. Springer, Heidelberg, pp 343–357
Meng M, Lee J, Payne JE (2016) RALS-LM unit root test with trend breaks and non-normal errors: application to the prebisch-singer hypothesis. Stud nonlinear dyn econom. doi:10.1515/snde-2016-005
Mitchell WF (1993) Testing for unit roots and persistence in OECD unemployment. Appl Econ 25:1489–1501
Neudorfer P, Pichelmann K, Wagner M (1990) Hysteresis. NAIRU and long term unemployment in Austria. Empir Econ 15:217–229
Ng S, Perron P (1995) Unit root tests in ARMA models with data-dependent methods for the selection of the truncation lag. J Am Stat Assoc 90(429):269–281
OECD (Organization of Economic Co-operation and Development) (2014) http://stats.oecd.org/
Papell DH, Murray CJ, Ghiblawi H (2000) The structure of unemployment. Rev Econ Stat 82:309–315
Perron P (1989) The great crash, the oil price shock and the unit root hypothesis. Econometrica 57:1361–1401
Perron P (1997) Further evidence on breaking trend functions in macroeconomic variables. J Econ 80:355–385
Perron P (2006) Dealing with structural breaks. In: Mills TC, Patterson K (eds) New Palgrave handbook of econometrics, vol 1. MacMillan, London, pp 278–352
Perron P, Vogelsang T (1992) Nonstationarity and level shifts with an application to purchasing power parity. J Bus Econ Stat 10:301–320
Pissarides CA (1992) Loss of skill during unemployment and the persistence of employment shocks. Q J Econ 107(4):1371–1391
Phelps E (1968) Money wage dynamics and labor market equilibrium. J Polit Econ 76:678–711
Phelps E (1972) Inflation policy and unemployment theory: the cost-benefit approach to monetary planning. W. W, Norton and Company Inc, New York
Phelps E (1994) Structural slumps: the modern equilibrium theory of unemployment, interest, and assets. Harvard University Press, Cambridge
Røed K (1996) Unemployment hysteresis–macro evidence from 16 OECD countries. Empir Econ 21:589–600
Røed K (1997) Hysteresis in unemployment. J Econ Surv 11:389–418
Schmidt P, Phillips PCB (1992) LM tests for a unit root in the presence of deterministic trends. Oxf Bull Econ Stat 54:257–287
Author information
Authors and Affiliations
Corresponding author
Additional information
We thank an Associate Editor and two anonymous referees for helpful comments.
Rights and permissions
About this article
Cite this article
Meng, M., Strazicich, M.C. & Lee, J. Hysteresis in unemployment? Evidence from linear and nonlinear unit root tests and tests with non-normal errors. Empir Econ 53, 1399–1414 (2017). https://doi.org/10.1007/s00181-016-1196-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00181-016-1196-z