The role of Cohesion Policy for sustaining the resilience of European regional labour markets during different crises

ABSTRACT

Whilst the analysis of the impact of Cohesion Policy funds is well documented, little is known about the short-term consequences of Cohesion Policy during economic crises and its implications for regional resilience. To study these issues, we assemble novel, panel time-series data on Cohesion Policy covering almost four programming periods and different shocks in the European Union. Using heterogeneous coefficients panel models, we find a general positive impact of Cohesion Policy on regional resilience, although with region- and crisis-specific patterns during different shocks. Our results also suggest the presence of regional variation regarding regional labour market resilience over the past three decades.

KEYWORDS:

• regional labour markets
• European Union Cohesion Policy
• resilience
• heterogeneous models
• economic shocks

JEL:

• H12
• O3
• R11
• R12

ACKNOWLEDGEMENTS

The authors are particularly grateful to the editors, the editorial manager and three anonymous referees for helpful suggestions and comments. They thank the participants at the XXXII Villa Mondragone Int. Ec. Seminar (Rome Tor Vergata), Regions in Recovery Festival, AISRe web conference 2021, and 13th RSAI World Congress. The usual disclaimers apply.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Notes

1. There is a vivid debate on the conceptualization of resilience (Bailey & Turok, 2016), which is outside the boundaries of the present work. In this paper, our focus is on the resistance dimension of resilience defined as the impact of a given shock on a region's economy (Martin & Sunley, 2014). In what follows, we use the terms ‘resilience’ and ‘resistance’ interchangeably.

2. Specific factors can explain differences in regional economic resilience across Europe; for a survey, see Giannakis and Bruggeman (2017) and Di Caro and Fratesi (2018). A non-exhaustive list includes: the degree of urbanization (Brakman et al., 2015); the urban–rural hierarchy (Giannakis & Bruggeman, 2020); macroeconomic conditions (Crescenzi et al., 2017); industrial structure (Capello et al., 2015); regional quality of government (Ezcurra & Rios, 2019); and population and labour market structures (Pontarollo & Serpieri, 2020b). However, only a few studies analyse the role of regional policies as drivers of resilience, with a focus on single countries only (Eraydin, 2016; Arbolino et al., 2020).

3. Studies conducted for a single MS, such as Italy (Arbolino & Di Caro, 2021) and Greece (Psycharis et al., 2020), find mixed evidence on the short-term effects of the EU funds on regional economies. For a survey on the effects of the Cohesion Policy, see Bristow and Healy (2014) and Dall’erba and Fang (2017).

4. The effects of Cohesion Policy on regional economies are conditional, among others, on factors such as: industrial structures (Percoco, 2017); territorial capital (Fratesi & Perucca, 2019) and human capital (Becker et al., 2018); institutions and administrative capacity (Arbolino et al., 2020); and type of EU expenditures (Di Cataldo & Monastiriotis, 2020).

5. We exclude DOM and TOM regions in France and the London region in the UK for which information on the EU funds are not available.

6. We use the latest available data release of 1 April 2019 ending in 2015. For more details, see Di Caro and Fratesi (2022).

7. We recognize, however, that regional observations are not able to disentangle all the spatial patterns present at lower level of aggregation, such as local labour systems, provinces and cities, which necessarily need finer units of observations (Gagliardi & Percoco, 2017; Percoco, 2017).

8. The graphs report data with and without Eastern Germany, given different time coverage.

9. To test the robustness of our crisis dummies, we have introduced year effects covering each crisis one at a time in order to check for placebo effects; the results are available from the authors upon request.

10. This share is also confirmed when looking at the two shocks of the Great Recession as a unique crisis.

11. The mapping does not substantially change when using alternative measures, as those presented in Martin (2012) and related works since all the measures are almost perfectly correlated.

12. This group includes, among others, one region in Austria (AT34), the Czech Republic (CZ01) and the UK (UKJ2), five regions in Germany (DE21, DE22, DE30, DE60, DE94), four regions in Spain (ES30, ES62, ES63, ES64), and three regions in Italy (ITC2, ITH1, ITH2).

13. This group includes, among others, one region in each of Belgium (BE32), Poland (PL52) and Slovakia (SK02), two regions in each of Germany (DED4, DEE0) and Portugal (PT15, PT18), and five regions in Italy (ITF1, ITF2, ITF3, ITF6, ITG2).

14. We have selected the lag-length of the ARDL specification (1, 1) in (3) based on common criteria (i.e., Schwartz–Bayesian criterion) in order to ensure that the residuals of the resulting error-correction model are exogenous and serially uncorrelated.

15. In this paper we do not study the long-run equilibrium properties of relation (3), including unit roots, co-integration patterns and speed of error-correction adjustment, given our focus on the short-term impact of crises and Cohesion Policy on employment growth. For an application of this model to the long-term impact of the Cohesion Policy, see Di Caro and Fratesi (2022).

16. A group-specific linear trend can be also included in (4) to check for time-specific effects in the regional regressions. However, we do not have time fixed effects in our final specifications since we detect a limited number of significant group-specific trends at 5% level. Results with time trends are available from the authors upon request.

17. See note 4.

Additional information

Funding

Ugo Fratesi acknowledges partial funding from PRIN project 'A New Assessment of Cohesion Policies' Effectiveness: Macro and Micro Approaches'.