Effects of region-specific shocks on labor market tightness and matching efficiency: evidence from the 2011 Tohoku Earthquake in Japan


This paper examines whether region-specific shocks alter regional labor market tightness and matching efficiencies. We adopt the Great East Japan Earthquake of March 2011, which caused a tsunami and nuclear disaster, as a region-specific shock. We find that an increase in labor market tightness, namely, labor shortage, occurs in the damaged regions after the disaster. Matching efficiencies in the damaged regions deteriorate, suggesting that the composition of unemployment and vacancies changes, leading to higher search frictions. Such nature has spatial spillover effects because of the widespread increase in demand for reconstruction and out-migration from the damaged regions.

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

Fig. 1
Fig. 2

Source Author’s calculation based on the Report on Employment Service

Fig. 3

Source Author’s calculation based on the Report on Prefectural Accounts, the Population Census, and the Population Estimates

Fig. 4

Source Author’s calculation based on Table 255 of the 2012 Employment Status Survey, provided by the Statistics Bureau, Ministry of Internal Affairs and Communications

Fig. 5
Fig. 6


  1. 1.

    See Angrist and Pischke (2009) for the DID method in detail.

  2. 2.

    To define the tsunami-hit coastal group, we find inspiration from Belasen and Polachek’s (2009) coastal dummy in their GDD models.

  3. 3.

    The evacuation zones consist of three types of areas: the difficult-to-return zone, restricted residence area, and zone in preparation for the lifting the evacuation order. The categorization in this study is as of December 31, 2013. The regional distribution of the evacuation zones is available from the website of the Ministry of Economy, Trade and Industry (http://www.meti.go.jp/earthquake/nuclear/hinan_history.html, accessed on December 12, 2018; in Japanese).

  4. 4.

    The GDD adopts multiple events as natural experiments; therefore, the treatment and control groups are not fixed but change with events. This allows the GDD to ensure that the estimated effects of events hardly suffer from the characteristics of observations, unlike conventional DID (Belasen and Polachek, 2009).

  5. 5.

    The models for labor market tightness, as well as those for matching efficiency, do not contain region fixed effects because the regression analyses omit some coefficients of treatment group dummies owing to multicollinearity.

  6. 6.

    The surrounding prefectures are Aomori, Akita, Yamagata, Ibaraki, Tochigi, Gunma, and Niigata Prefectures.

  7. 7.

    To construct balanced panel data, we arrange the data as follows. First, certain types of PESOs that provide referral serves only for some job seekers (e.g., foreigners, older people, and women) are eliminated from our sample because their wider jurisdiction areas than general PESOs do not allow us to arrange a coherent regional unit. Second, some PESOs merged during the sample period; therefore, we reaggregate PESOs’ records in each period as of 2015, when the last merger occurred. Higashi (2018) handles the same data in the same way.

  8. 8.

    We obtained the data on population in 2010 and 2015 from the Population Census, and those of other years from the Population Estimates. Both statistics are provided by the Ministry of Internal Affairs and Communications.

  9. 9.

    “Appendix 2” provides a list of industries.

  10. 10.

    Table 4 lists the sample PESO jurisdictions regions by the treatment and control groups.

  11. 11.

    Some studies on the Japanese labor market confirm such cycles (e.g., Kano and Ohta, 2005; Higashi, 2018).

  12. 12.

    See the press releases issued by the Ministry of Health, Labour and Welfare, “Higashinihondaishinsai ni tomonau rodokijunkantokusho, HelloWork no kaicho jokyo nitsuite, dai 1–12 hou (The status of opening of the Labor Standards Inspection Office and the HelloWork due to the Great East Japan Earthquake, no. 1–12)” (https://www.mhlw.go.jp/stf/houdou/2r98520000015q3n.html, accessed on December 1, 2018; in Japanese).

  13. 13.

    Figure 6 captures the effects of the PESOs’ shutdown.

  14. 14.

    See Higuchi et al. (2012) and Cabinet Office (2016).

  15. 15.

    See Table 5.

  16. 16.

    See the website of the Japan Atomic Energy Commission (http://www.aec.go.jp/jicst/NC/iinkai/teirei/siryo2011/siryo33/siryo1-2.pdf, accessed on January 17, 2019; in Japanese).

  17. 17.

    See the website of the Tokyo Electric Power Company Holdings, Inc. (http://www.tepco.co.jp/decommission/progress/environment/index-j.html, accessed on January 17, 2019; in Japanese).

  18. 18.

    See Kitamura (2014), Fukushima Minpo (2012), and Fukushima Minyu Shinbun (2018).

  19. 19.

    Table 6 contains the estimation result of elasticities in the matching function.

  20. 20.

    A region’s matching efficiency is evaluated on the basis of job seekers registered with the PESO in the corresponding region, owing to the definition of the dependent variable (i.e., outflows from unemployment) in our data (see Sect. 3). In other words, a region’s matching efficiency indicates how efficiently job seekers in the corresponding region succeed in matching with vacancies, regardless of the locations of the vacancies.

  21. 21.

    Another possibility is that the operations of the PESOs there worsened.

  22. 22.

    Matsumoto (2012) refers to the relationship between employment and the Great East Japan Earthquake in detail.

  23. 23.

    Although we are unable to confirm, we guess that some non-working individuals with low search intensities might move within a lower geographical scope, while others with high search intensities might move further. Consequently, the rate of searching persons cancels out.


  1. Angrist JD, Pischke J-S (2009) Mostly harmless econometrics: an empiricist’s companion. Princeton University Press, Princeton

    Google Scholar 

  2. Barnichon R, Figura A (2015) Labor market heterogeneity and the aggregate matching function. Am Econ J Macroecon 7(4):222–249

    Article  Google Scholar 

  3. Belasen AR, Polachek SW (2009) How disasters affect local labor markets the effects of hurricanes in Florida. J Hum Resour 44(1):251–276

    Google Scholar 

  4. Cabinet Office (2016) Higashinihon-daishinsai no fukko no shinchoku to kongo no kadai (Progress of reconstruction of the Great East Japan Earthquake and further issues). In: Chiiki no keizai 2015: Keizai kojunkan no chiiki keizai eno hakyu to baratsuku keikyokan (Regional economics 2015: The ripple effects of boom on regional economy and variations of business confidence). Tokyo: Cabinet Office, pp 42–53 (Chapter 5) (in Japanese)

  5. Coles MG, Smith E (1996) Cross-section estimation of the matching function: evidence from England and wales. Economica 63(252):589–597

    Article  Google Scholar 

  6. Fahr R, Sunde U (2006) Regional dependencies in job creation: an efficiency analysis for Western Germany. Appl Econ 38(10):1193–1206

    Article  Google Scholar 

  7. Higashi Y (2018) Spatial spillovers in job matching: evidence from the Japanese local labor markets. J Jpn Int Econ 50:1–15

    Article  Google Scholar 

  8. Higuchi Y, Inui T, Hosoi T, Takabe I, Kawakami A (2012) The impact of the Great East Japan earthquake on the labor market: need to resolve the employment mismatch in the disaster-stricken areas. Jpn Labor Rev 9(4):4–21

    Google Scholar 

  9. Ibourk A, Maillard B, Perelman S, Sneessens HR (2004) Aggregate matching efficiency: a stochastic production frontier approach, France 1990–1994. Empirica 31(1):1–25

    Article  Google Scholar 

  10. Japan Institute for Labour Policy and Training (2015) Nyushoku keiro no henka to min-ei shokugyo shokaigyo ni kansuru chosa (Survey of transition of job search method and private employment service). JILPT shiryo series, No. 159 (in Japanese)

  11. Kano S, Ohta M (2005) Estimating a matching function and regional matching efficiencies: Japanese panel data for 1973–1999. Jpn World Econ 17(1):25–41

    Article  Google Scholar 

  12. Kitamura T (2014) Fukushimagenpatsu jiko hitodebusoku kaisho no tameni (To solve labor shortage for the Fukushima nuclear accident). Agora, April 16. http://agora-web.jp/archives/1590936.html. Accessed Jan 17 2019 (in Japanese)

  13. Kodama T, Higuchi Y, Abe M, Matsuura T, Sunada M (2004) Nyushoku keiro ga tenshoku seika ni motarasu koka (Impact of job search method on job turnover). RIETI discussion paper series, no. 04-J-035 (in Japanese)

  14. Matsumoto Y (2012) Higashinihon daishinsai no hisaichi niokeru koyo kanren service to kyujin kyushoku jokyo: Hellowork gyomu wo chushin toshite (Status of employment service, vacancies, and job search in the damaged area of the Great East Japan Earthquake: focusing on employment referral by HelloWork). Jpn J Labour Stud 622:71–78 (in Japanese)

    Google Scholar 

  15. Petrongolo B (2001) Reemployment probabilities and returns to matching. J Labor Econ 19(3):716–741

    Article  Google Scholar 

  16. Petrongolo B, Pissarides CA (2001) Looking into the black box: a survey of the matching function. J Econ Lit 39(2):390–431

    Article  Google Scholar 

  17. Pissarides CA (2000) Equilibrium unemployment theory, 2nd edn. MIT Press, Cambridge

    Google Scholar 

  18. Sasaki M, Kohara M, Machikita T (2013) Measuring search frictions using Japanese microdata. Jpn Econ Rev 64(4):431–451

    Article  Google Scholar 

  19. Fukushima Minpo (2012) Kuni chokkatsu josen de jinin busoku kenen (Concerning labor shortage in decontamination of lands which is directly controlled by the national government). Fukushima Minpo, November 6, 2012. http://www.minpo.jp/pub/topics/jishin2011/2012/11/post_5446.html. Accessed Jan 17 2019 (in Japanese)

  20. Fukushima Minyu Shinbun (2018) Izentoshite “Hitode busoku” tsuduku Fukushima-ken (“Labor shortage” is still continuing in Fukushima Prefecture). Fukushima Minyu Shinbun, February 11, 2018. http://www.minyu-net.com/news/sinsai/serial/0611/FM20180211-243273.php. Accessed Jan 17 2019 (in Japanese)

Download references


I would like to thank two anonymous reviewers, Kazufumi Yugami, Nobuaki Hamaguchi, Shinya Horie, and Tamotsu Nakamura for helpful comments. I would also like to thank Hirokazu Fujii of the Ministry of Health, Labour and Welfare, and the Employment Policy Division of Employment Security Bureau for their providing the data for Public Employment Security Office. Remaining errors are my own. This research was supported by the Grant-in-Aid for Japan Society for the Promotion of Science (JSPS) Fellows (JSPS KAKENHI Grant No. JP18J11560).

Author information



Corresponding author

Correspondence to Yudai Higashi.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


Appendix 1

See Tables 4, 5 and 6.

Table 4 List of sample regions by group
Table 5 Net-migration by region group (persons in thousands)
Table 6 Estimation result of the matching function

Appendix 2: List of industries of which per capita GPPs are controlled in the DID models

Agriculture; Forestry; Fishing; Mining; Manufacturing; Construction; Electricity, gas, water supply, and waste disposal business; Wholesale and retail trade; Finance and insurance; Real estate; Transport and postal activities; Information and communications; Accommodations, eating and drinking services; Professional and science technology, business support service; Public service; Education; Public health, hygiene, and social service; Miscellaneous services.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Higashi, Y. Effects of region-specific shocks on labor market tightness and matching efficiency: evidence from the 2011 Tohoku Earthquake in Japan. Ann Reg Sci 65, 193–219 (2020). https://doi.org/10.1007/s00168-020-00980-w

Download citation

JEL Classification

  • J64
  • R12
  • R23