Cities are more vulnerable to political and economic dislocation than to physical destruction (Glaeser 2021). For example, mass layoffs shrink the local labour force by inducing some of the displaced workers to migrate (Foote et al. 2015), which may have permanent long-run effects on the city’s growth and socioeconomic composition. How the latter changes in the wake of negative economic shocks – and what city-level characteristics favour urban resilience – are not well known. In a recent study (Behrens et al. 2021), we shed some light on these questions.
The effects of plant closures and mass layoffs on workers and local labour markets
Research on job displacement has shown that workers who lose their jobs due to big-plant closures or mass layoffs suffer from long-lasting income losses, longer unemployment durations, and other adverse outcomes such as reduced fertility, higher mortality, and lower income for their kids when they become adults (e.g. Jacobson and LaLonde 1993, del Bono and Winter-Ebner 2008, Oreopoulos et al. 2008).
The effects of mass layoffs and big-plant closures on local economies are still debated. Some studies show significant negative spillovers on other local firms, so that the number of locally available jobs decreases by more than the number of displaced jobs (Gathmann et al. 2020). Other studies find that part of the job losses generated by big-plant closures is compensated by new or incumbent local firms (Jofre-Monseny et al. 2018).
Job displacement triggers outmigration, but not all workers react in the same way to local labour demand shocks. Several studies show that high-skilled workers and immigrants are generally more responsive to local shocks (e.g. Albouy et al. 2019). Beyond different mobility costs, the inelastic housing supply, the existence of social transfers, and the immigration selection criteria can explain this heterogeneous response of workers to local labour-demand shocks (e.g. Notowidigdo 2020).
Big plant closures, mass layoffs, and the age structure of cities
In Behrens et al. (2021), we evaluate the impact of closures and substantial downsizing of big manufacturing plants on the growth and demographic composition of Canadian cities. Around 33% of Canadian manufacturing jobs have disappeared between 2003 and 2017 due to the closure or massive downsizing of establishments with 50+ employees, many of them not being replaced. However, this job-loss rate is quite heterogeneous across Canadian provinces and across cities within provinces, as shown in Figure 1.
Figure 1 Relative job loss rates due to big-plant closures in Canadian urban areas, 2003–2017
Note: Distribution of manufacturing job-loss rates due to large (50+) plant closures in Canadian urban areas. Canadian urban areas’ job-loss rates are measured relative to the Canadian average. A value of 1 on the map means that the urban area’s job-loss rate is the same as the Canadian mean. Cyan contours outline cities with population of at least 300,000 inhabitants.
We can see in Figure 2 that, between 2003 and 2017, Canadian cities experienced very different demographic evolutions too. For example, the population of Campbellton in New Brunswick shrank the most (-18.2% from an initial population of 16,980 in 2001), while the population of Wood Buffalo in Alberta grew the fastest (+72.4% from an initial population of 42,475 in 2001). We thus compare the population growth in cities that were severely hit by big-plant closures and mass layoffs to the demographic evolution of cities where the manufacturing job loss rate is lower.
Figure 2 Relative population growth rates in Canadian urban areas, 2001–2016
Note: Growth rates are measured relative to the Canadian average. A value of 1 on the map means that the urban area’s growth rate is the same as the Canadian mean. Cyan contours outline cities with population of at least 300,000 inhabitants.
To ensure that our analysis is not biased by confounding factors, we account for the initial size and composition of cities, as well as for various amenities such as average temperatures, distance to the coast, and distance to the closest big city. Also, to ensure that we capture the impact of big-plant closures on demographic change and not the reverse (since firms could also follow workers and shut down or downsize in shrinking cities), we use the sectoral employment growth rate in the US and the initial composition of manufacturing activity in Canadian cities to build a shift-share (Bartik) instrument.
We find that plant closures lead to lower subsequent population growth, especially among working-aged (20–59) and very young residents (0–19). Cities severely hurt by mass layoffs become ageing cities since working-age residents (and their kids, if any) are more likely to leave in search of job opportunities elsewhere.
The most likely groups to leave cities affected by negative labour-demand shocks are singles and people with a migration background. This is easy to understand: the latter have already moved before in their life, while the former have lower migration costs as they have no joint location decisions to manage.
We further show that the closure and massive downsizing of big manufacturing plants negatively affect the employment growth of several other sectors in the local economy, especially in construction, cultural services, and finance, insurance and real estate (FIRE). These negative spillovers might partly explain why negative employment shocks in the manufacturing sector have such a significant depressing effect on the demographic dynamics of cities.
Cultural and public services as factors of city resilience
When exposed to similar job-loss rates, not all cities experience the same demographic decline. More specifically, we identify two factors that favour city resilience. First, cities whose initial share of locals employed in education and in health and social services is the highest suffer less demographic decline following big-manufacturing-plant closures and mass layoffs. This mitigating effect is especially significant for migrants, who thus seem to value more (or benefit more from) these services in case of adverse local labour-demand shocks.
We also find that cities whose initial share of residents employed in arts and entertainment and recreational activities is the highest withstand better the adverse effects of massive job displacements. The effect is especially concentrated among the working-age residents and those with at least a bachelor’s degree, since they are bigger consumers of cultural services.
Studies on factors that favour the resilience of local economies are relatively scarce (Behrens et al. 2018). We show that public and cultural services do favour city resilience in case of adverse labour market shocks. At a time when COVID-19 puts these services seriously under pressure, our findings are a reminder of how important they are for our economies.
Albouy, D, A Chernoff, C Lutz and C Warman (2019), “Local labor markets in Canada and the US”, Journal of Labor Economics 37(S2): 533–94.
Behrens, K, B Boualam and J Martin (2018), “Plants’ resilience and clusters: Evidence from the Canadian textile and clothing industry”, VoxEU.org, 3 January.
Behrens, K, M Drabo and F Mayneris (2021), “Cultural and public services as factors of city resilience? Evidence from big plant closures and downsizing”, CEPR Discussion Paper 16723.
del Bono, E, and R Winter-Ebmer (2008), “The effect of job displacement on women’s fertility decisions”, VoxEU.org, 25 February.
Foote, A, M Grosz and A H Stevens (2015), “How workers exit the labour market after local economic downturns”, VoxEU.org, 17 November.
Gathmann, C, I Helm and U Schönberg (2020), “Spillover effects of mass layoffs”, Journal of the European Economic Association 18(1): 427–68.
Glaeser, E L (2021), “Urban resilience”, NBER Working Paper 29261.
Jacobson, L S, and R J LaLonde (1993), “Earnings losses of displaced workers”, American Economic Review 83(4): 685.
Jofre-Monseny, J, M Sánchez-Vidal and E Viladecans-Marsal (2018), “Big plant closures and local employment”, Journal of Economic Geography 18(1): 163–86.
Notowidigdo, M J (2020), “The incidence of local labor demand shocks”, Journal of Labor Economics 38(3): 687–725.
Oreopoulos, P, M Page and A H Stevens (2008), “The intergenerational effects of worker displacement”, Journal of Labor Economics 26(3): 455–83.