![Sample our Global Development journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5929/TOC-Subject-Sample-2021-Global-Development.jpg"})
Abstract
Røed M. and Schøne P. Displacement and immigrant workers' responsiveness to regional labour market opportunities: evidence from Norway, Regional Studies. This paper answers two main questions. First, are immigrants more mobile than natives following job displacement? The results show that immigrants are indeed more mobile than natives following job displacement; however, the difference is fully explained by a compositional effect – immigrants have much shorter potential work experience in Norway than natives. Second, are immigrants more responsive than natives to differences in regional labour market opportunities following job displacement? The results show that displaced Western immigrants are more responsive to differences in regional labour market opportunities compared with both non-Western immigrants and natives. The greasing effects are limited to immigrants with low mobility costs.
Røed M. and Schøne P. 解雇与移民工对区域劳动市场机会的回应：挪威的证据，区域研究。本文回答下列两个主要问题。首先，相较于本地人，移民在解雇之后是否更具流动性？研究结果显示，移民在解雇之后的确较本地人更具有流动性，但此一差异可充分地以组成效应解释之—在挪威，移民的可能工作经历远较本地人为短。再者，相较于本地人，解雇后的移民是否更能回应区域劳动市场机会中的差异？研究结果显示，解雇的西方移民同时较非西方移民及本地人而言，更能回应区域劳动市场机会中的差异。润滑效应限于仅须付出少量移动成本的移民。
Røed M. et Schøne P. Le déplacement de l'emploi et la réactivité des travailleurs immigrés aux possibilités sur le marché du travail régional: des preuves provenant de la Norvège, Regional Studies. Ce présent article cherche à répondre à deux questions principales. Primo, les immigrés, sont-ils plus mobiles que ne le sont les autochtones suite à un déplacement de l'emploi? Les résultats laissent voir que les immigrés sont en effet plus mobiles que les autochtones suite à un déplacement de l'emploi; toujours est-il que la différence s'explique à fond par un effet compositionnel – à savoir, l'expérience professionnelle potentielle des immigrés s'avère beaucoup plus courte en Norvège que celle des autochtones. Secundo, les immigrés, répondent-ils mieux que les autochtones aux différences des possibilités sur le marché du travail régional suite à un déplacement de l'emploi? Les résultats laissent voir que les immigrés déplacés en provenance des pays occidentaux répondent mieux aux différences des possibilités sur le marché du travail régional par rapport aux immigrés à la fois non occidentaux et autochtones. Les effets graissage sont limités aux immigrés dont les frais de déplacement sont moins élevés.
Røed M. und Schøne P. Arbeitsplatzverlust und Reaktion von eingewanderten Arbeitnehmern auf Chancen des regionalen Arbeitsmarkts: Belege aus Norwegen, Regional Studies. In diesem Beitrag werden zwei Hauptfragen beantwortet. Erstens: Sind Einwanderer nach einem Arbeitsplatzverlust mobiler als Einheimische? Aus den Ergebnissen geht hervor, dass Einwanderer nach einem Arbeitsplatzverlust in der Tat mobiler sind als Einheimische; allerdings lässt sich der Unterschied vollständig durch einen Zusammensetzungseffekt erklären: Einwanderer verfügen in Norwegen über erheblich kürzere potenzielle Arbeitserfahrung als Einheimische. Zweitens: Reagieren Einwanderer nach einem Arbeitsplatzverlust stärker als Einheimische auf Unterschiede bei den Chancen der regionalen Arbeitsmärkte? Aus den Ergebnissen geht hervor, dass Einwanderer des westlichen Kulturkreises nach einem Arbeitsplatzverlust stärker auf Unterschiede bei den Chancen der regionalen Arbeitsmärkte reagieren als Einwanderer aus nichtwestlichen Kulturkreisen und Einheimische. Die Schmiereffekte beschränken sich auf Einwanderer mit geringen Mobilitätskosten.
Røed M. y Schøne P. Pérdida de empleo y reacción de los trabajadores inmigrantes a las oportunidades del mercado regional laboral: el ejemplo de Noruega, Regional Studies. En este documento, se responde a dos cuestiones principales. Primero: ¿son los inmigrantes más móviles que los autóctonos cuando pierden el trabajo? Los resultados indican que en efecto los inmigrantes son más móviles que los autóctonos tras una pérdida de trabajo; sin embargo, esto se explica exhaustivamente mediante el efecto de composición social: los inmigrantes tienen una posible experiencia laboral mucho más corta en Noruega que los autóctonos. Segundo: ¿responden mejor los inmigrantes a diferencias en las oportunidades del mercado regional laboral que los autóctonos cuando pierden el trabajo? Los resultados indican que los inmigrantes occidentales que pierden el trabajo responden más a las diferencias en las oportunidades del mercado regional laboral que los inmigrantes no occidentales y los autóctonos. Los efectos lubricantes están limitados a los inmigrantes con bajos costes de movilidad.
Acknowledgements
The authors would like to thank their colleagues, the seminar participants at the Institute for Social Research as well as the participants at the Comparative Analysis of Enterprise Data (CAED) conference in London, 2010, for their valuable comments. Financial support from The Research Council of Norway under the project ‘Does Immigration Grease the Wheels of the Regional Labour Markets?’ (Grant Number 178188/S20) is gratefully acknowledged.
Notes
2. This specification of the conditional logit model in the migratory context is more thoroughly discussed by Davies et al. Citation(2001) and O'Keefe (2004). One problem with this approach is that all regions within a country do not constitute an exhaustive set of alternatives for a resident who is considering where to settle down in the next period. This is so because most people may also move abroad. Since the relevant information about labour market variables in the receiving regions of other countries was not available to the authors, individuals who move abroad are excluded from this analysis. If emigrants, compared with the rest of the population, are more or less sensitive to regional labour market conditions when making their locational decisions, this may create a selectivity bias in the estimated marginal effects.
3. It could be argued that network effects do not follow country borders. Instead, immigrants form networks with other immigrants from the same culture. The authors have experimented with broader categories than country. This reduced the precision of the estimate without changing the coefficient of the main variable. Therefore, it was chosen to use the preferred ethnic network variable.
4. Specifically, the adjusted regional ethnic concentration is constructed as follows:
where IMMgj1990 is the share of immigrants from country g in county j in 1990; and Ngt is the number of immigrants from the group in year t (t = 1996, …, 2005).
5. Google Maps was used to calculate the number of kilometres in (car) driving distance between the county capitals.
6. Still, the correlation between the potential work experience measure and actual work experience (the number of years with positive labour market earnings) is very high in the present material; 0.90 for non-Western immigrants and 0.94 for Western immigrants compared with 0.97 for natives.
7. The standardized numbers of years to complete a level of education are shown in parentheses.
8. It could be argued that the most recent wage available for individual workers should be used. Since the definition of displacement is that the plant is operative by the end of year t, and has been closed down or severely downsized by the end of year t + 1, it was measured at the last point in time, which is equal for all workers in the sample.
9. County was controlled for because the authors wanted to control for fixed county characteristics that affect the mobility behaviour of immigrants and natives differently. It may be relatively easier for immigrants to find a new job within a large county with a varied industry structure. In addition, county of residence was used as the measure of mobility rather than county of plant location. Some workers are not employed in the post-period, and they have no plant location. Therefore, county of residence was chosen.
10. Due to errors in registration, some (but very few) firms reappear with the same firm identifier at t + 2, even if they were not registered at the end of t + 1. These firms were left out of the analyses.
11. The distinction between Eastern European immigrants and those from Western countries is set according to standard definitions in Statistics Norway. The border is set by countries such as Lithuania, Estonia, Latvia, Poland, Hungary, Bulgaria, the Czech Republic, Slovakia, Bosnia, Herzegovina and Slovenia. All these countries are defined as Eastern Europe.
12. The marginal effects were calculated by the delta method using the margins command in Stata.
13. The share with missing information on education is much larger for immigrants than for natives. Therefore, it will be a correlation between the dummy variable for unknown education and the immigrant dummy variables. To shed light on this issue, Model 2 was estimated, though excluding the education dummy variables. The immigrant coefficients for domestic mobility do not change much, but the corresponding emigration coefficients are increased for both immigrant groups. Model 2 reveals a strong positive relationship between unknown education and emigration. When this is not controlled for, the coefficients for the two immigrant groups' emigration likelihood are higher.
14. Separate regression analyses for recent (a maximum of five years since immigration) and non-recent immigrants (more than five years since immigration) were also estimated. The results (available from the authors upon request) show that the mobility difference (both domestically and emigration) between immigrants and natives is larger among recent immigrants.
15. The number of observations is fewer in the t + 5 estimations since the last three years of observations must be left out.
16. One note of clarification: the analyses could have been limited to those displaced workers who experienced post-displacement spells of unemployment in the county in which they lived when they were displaced. However, this would omit a large percentage of the immigrant workers, especially non-Western immigrants. One reason why people register as unemployed is that this is a precondition for receiving unemployment benefits. To gain the unemployment benefit, one must have labour earnings above a threshold, which takes time. Thus, many of the newly arrived immigrants have not yet gained this right and may have weaker incentives than natives to register as unemployed.
17. That is, if calculated using an average settlement probability equal to 1/19.
18. Oslo is excluded as a possible destination by it being omitted as one of the potential destination areas. This means that each individual has one less option regarding counties from which to choose. Observations where individuals are registered as residents in Oslo were also omitted.
19. Regressions were also run where a period fixed-effect dummy that takes the value of 1 if the period is the first five years (1996–2000) and 0 if it is the last five years (2001–2005) was added. To get the necessary variation, this variable must be interacted with the county dummies. Estimating Model 3 with this expanded model showed that the unemployment coefficient for Western immigrants was somewhat strengthened, while the unemployment coefficient for non-Western immigrants was unchanged. Results available from the authors upon request.
20. A similar exercise was also run using potential work experience. This shows a similar picture. This is not surprising considering the high correlation between YSM and potential work experience in Norway.