Immigrants' Decision to Stay in the Canary Islands: A Latent Class Approach

Abstract

León C. J. and Hernández Alemán A. Immigrants’ decision to stay in the Canary Islands: a latent class approach, Regional Studies. This paper presents evidence on the socioeconomic and psychological factors that explain international immigrants' decisions to stay in the Canary Islands. The data are modelled using a latent class binary approach that allows two types of migrant to be considered according to their level of probability of staying at the destination. The results show that psychological factors (negative emotions, expectations, social integration) are more important for those migrants with a low probability of staying. In addition, the circumstances at the destination (Spanish language, years of stay, number of migrations, remittances) are significant only for those migrants with a high probability of staying. These results have implications for immigration policies in both the origin and host regions.

León C. J. and Hernández Alemán A. 移民留在加纳利群岛的决定：潜在类别方法，区域研究。本文呈现解释跨国移民决定留在加纳利群岛的社会经济与心理因素。本研究运用潜在类别二分法，对数据进行模式化，该方法得以依据两种移民类型留在目的地的可能性进行考量。研究结果显示，心理因素（负面情绪、期待、社会整合）对于居留可能性较小的移民而言更为重要。此外，目的地的境况（西班牙语、停留年期、移民数量、汇款）只对具有较高居留可能性的移民而言是显着的。这些研究结果，同时对原生区域和东道区域的移民政策具有意涵。

León C. J. et Hernández Alemán A. La décision des immigrés de rester aux Canaries: une analyse de structure latente, Regional Studies. Cet article présente des résultats sur les facteurs socioéconomiques et pyschologiques qui expliquent la décision des immigrés internationaux de rester aux Canaries. On modélise les données à partir d'une analyse de structure binaire latente qui permet d’étudier deux catégories d'immigrés selon la probabilité qu'ils vont rester dans la zone d'accueil. Les résultats laissent voir que les facteurs pyschologiques (tels les émotions négatives, les attentes, l'intégration sociale) s'avèrent plus importants pour les immigrés dont la probabilité de rester est peu élevée. En plus, les circonstances dans la zone d'accueil (la langue espagnole, la durée du séjour, le nombre d'immigrés, le transfert de fonds) ne sont significatives que pour les immigrés dont la probabilité de rester est élevée. Ces résultats ont des conséquences pour les politiques d'immigration à la fois dans la zone d'origine et la zone d'accueil.

León C. J. und Hernández Alemán A. Entscheidungen von Einwanderern zum Verbleib auf den Kanarischen Inseln: ein latenter Klassenansatz, Regional Studies. In diesem Beitrag werden Belege für die sozioökonomischen und psychologischen Faktoren vorgestellt, die für die Entscheidungen von internationalen Einwanderern zum Verbleib auf den Kanarischen Inseln eine Rolle spielen. Die Daten werden mithilfe eines latenten binären Klassenansatzes modelliert, der die Berücksichtigung zweier Arten von Migranten je nach dem Grad der Wahrscheinlichkeit ihres Verbleibs am Zielort ermöglicht. Aus den Ergebnissen geht hervor, dass psychologische Faktoren (negative Gefühle, Erwartungen, soziale Integration) für Migranten mit geringer Wahrscheinlichkeit eines Verbleibs wichtiger sind. Darüber hinaus sind die Umstände am Zielort (spanische Sprache, Jahre des Aufenthalts, Anzahl der Migrationen, Überweisungen) nur für Migranten mit hoher Wahrscheinlichkeit eines Verbleibs signifikant. Die Ergebnisse haben Auswirkungen auf die Einwanderungspolitik in den Ursprungs- und Gastregionen.

León C. J. y Hernández Alemán A. Las decisiones de los immigrates de permancer en las Islas Canarias: un enfoque de clases latentes, Regional Studies. Este artículo presenta evidencias sobre el papel de los factores socioeconómicos y psicológicos que explican las decisiones de los inmigrantes internacionales de permanecer en las Islas Canarias. Los datos se han modelado utilizando un enfoque de clases binarias latentes que permite la consideración de dos tipos de migrantes de acuerdo al nivel de probabilidad de permanecer en el destino. Los resultados muestran que los factores psicológicos (emociones negativas, expectativas, integración social) son más importantes para los migrantes con una baja probabilidad de permanecer. Además, las circunstancias en el destino (lengua española, años de estancia, número de migraciones, remesas) son significativas solo para los migrantes con una alta probabilidad de permanecer. Estos resultados tienen implicaciones para las políticas de inmigración tanto en las regiones de origen como de destino.

• Decision to stay
• Emotions
• Expectations
• Latent class model
• Migrants
• Social integration

• 居留的决定
• 情绪
• 期待
• 潜在类别模型
• 移民
• 社会整合

• Décision de rester
• Émotions
• Attentes
• Analyse de structure latente
• Migrants
• Intégration sociale

• Entscheidung zum Verbleib
• Gefühle
• Erwartungen
• Latentes Klassenmodell
• Migranten
• Soziale Integration

• Decisión de permanecer
• Emociones
• Expectativas
• Modelo de clases latentes
• Migrantes
• Integración social

JEL classifications:

• J15
• J61
• F22

Acknowledgements

The authors are thankful for the useful comments made by two anonymous referees. The usual disclaimer applies.

Notes

1. A large array of literature evaluates the economic impacts of migrants on both the host and origin regions (Massey et al., 1994; Nekby, 2006; Dustmann, 2003; Dustmann and Kirchkamp, 2002).

2. According to data published by the Organisation for Economic Co-operation and Development (OECD) for 2005, ‘inflows of foreign population into selected OECD countries’, 46% of these flows are permanent and 54% are temporary. The intention to stay does not always become effective because of the interplay of diverse factors, such as administrative, economic, legal or social factors.

3. From official statistics, 52% of foreign immigrants are males (Instituto Nacional de Estadística (INE), 2005).

4. From official statistics, 67.2% of African immigrants are males (INE, 2005).

5. From official statistics, the mean age of migrants is 31 years (INE, 2005).

6. The regularity conditions of the likelihood ratio (LR) statistics fail for latent class models. For instance, when comparing one latent class model against another, the mixing proportion under the null hypothesis is on the boundary of the parameter space, and therefore the LR statistics does not asymptotically follow a chi-squared distribution.

7. BIC = −2Λ + k ln(n), where Λ is the log-likelihood; k is the number of parameters in the model; and n is the simple size. AIC = −2Λ + 2k; and CAIC = −2Λ + k (ln(n) + 1).

8. As was also modelled by Jasso and Rosenzweig (1988), Steiner and Velling (1994) and Vadean and Piracha (2010).

9. A factor analysis also enabled a latent variable of satisfaction to be built up based on all aspects of the life of the immigrant. This variable also turned out to be moderately significant in explaining the decision to stay, with the same interpretation as the single variable of the satisfaction with social integration.

10. Emotions were rated using a 1–7 Likert scale, from the lowest to the highest intensity. The categories for negative emotions were feeling bad, sad, mad, indignant, stunned, worried, and I don't like anything. The categories for positive emotions were feeling happy, euphoric, happy and well.

11. The country of origin of the migrants was not significant in explaining the decision to stay in any of the classes identified with the model.

12. For a similar result, see Dustmann (1993), Borjas (1989) and Steiner and Velling (1994).

13. The experience that migratory movements provide to the family has a positive effect on planning a new move (Bharati, 1997).

14. Formal statistical test comparisons of differences in the average levels of specific emotions between both classes of migrants are available from the authors upon request.

Additional information

Funding

Financial support is acknowledged from Project ECO2011-30365 of the Spanish Ministry of Science and Technology.