Regional Sorting of Human Capital: The Choice of Location among Young Adults in Sweden

Abstract

Berck P., Tano, S. and Westerlund O. Regional sorting of human capital: the choice of location among young adults in Sweden, Regional Studies. Migration rates are highest among young adults, especially students, and their location choices affect the regional distribution of human capital, growth and local public sector budgets. Using Swedish register data on young adults, the choice of whether to enrol in education and the choice of location are estimated jointly. The results indicate a systematic selection into investment in further education based on school grades and associated preferences for locations with higher per capita tax bases. For students, the estimates indicate lower preferences for locations with higher shares of older people. The importance of family networks for the choice of location is confirmed.

Berck P., Tano, S. and Westerlund O. 人力资本的区域筛选：瑞典年轻成人的地点选择，区域研究。年轻成人具有最高的迁徙率，特别是对学生而言，而他们的地点选择，影响着人力资本的区域分佈、成长和地方公共部门的预算。本文运用瑞典年轻成人的登记数据，联合评估他们是否决定入学，及其地点的选择。研究结果显示出根据学校成绩和对较高人均税基地点的相关偏好，对未来的教育投资进行的系统性选择。研究评估显示出，学生较不偏好老年人口比例较高的地点。本研究并証实了家庭网络之于地点选择的重要性。

Berck P., Tano, S. et Westerlund O. Le triage régional du capital humain: le choix d'emplacement des jeunes adultes en Suède, Regional Studies. Les taux de migration sont les plus élevés parmi les jeunes adultes, notamment les étudiants, et leurs choix d'emplacement influent sur la distribution régionale du capital humain, de la croissance et des budgets du secteur public local. Employant des données de registre suédoises auprès des jeunes adultes, on estime simultanément le choix d'accéder à l’éducation et le choix d'emplacement. Les résultats indiquent un choix systématique en faveur de l'investissement dans la formation continue fondé sur les notes obtenues pendant la scolarité et les préférences y associées pour des emplacements où les assiettes de l'impôt par tête sont plus élevées. Pour les étudiants, les estimations indiquent des préférences moins élevées pour des emplacements où la part des personnes âgées est plus élevée. L'importance des réseaux familiaux pour le choix d'emplacement est confirmée.

Berck P., Tano, S. und Westerlund O. Regionale Anordnung von Humankapital: Standortwahl von jungen Erwachsenen in Schweden, Regional Studies. Die Migrationsraten fallen unter jungen Erwachsenen und insbesondere unter Studenten am höchsten aus; die Standortwahl dieser Personen wirkt sich auf die regionale Verteilung von Humankapital und Wachstum sowie auf die Etats des kommunalen Sektors aus. Anhand von schwedischen Meldedaten für junge Erwachsene werden die Entscheidung zur Immatrikulation und die Standortwahl gemeinsam geschätzt. Die Ergebnisse lassen auf eine systematische Auswahl der Investitionen in Weiterbildung schließen, die von der Höhe des Schulabschlusses und den damit verbundenen Präferenzen für Orte mit höheren steuerlichen Bemessungsgrundlagen pro Einwohner abhängt. Für Studenten weisen die Schätzungen auf niedrigere Präferenzen für Orte mit einem höheren Seniorenanteil hin. Die Wichtigkeit von familiären Netzwerken für die Standortwahl bestätigt sich.

Berck P., Tano, S. y Westerlund O. Clasificación regional del capital humano: elección del lugar por parte de jóvenes adultos en Suecia, Regional Studies. Las tasas de migración son las más altas entre adultos jóvenes, especialmente estudiantes, y los lugares que eligen afectan a la distribución regional de capital humano y crecimiento, así como a los presupuestos locales del sector público. A partir de los datos de registros suecos sobre adultos jóvenes, se estiman conjuntamente la elección de matricularse en instituciones de educación superior y la elección del lugar. Los resultados indican una selección sistemática de la inversión en educación superior basándose en los grados de escolarización y las preferencias asociadas por lugares con bases fiscales más altas per cápita. Los cálculos indican que los estudiantes muestran menos preferencia por lugares con un porcentaje más alto de personas mayores. Se confirma la importancia de las redes familiares a la hora de elegir el lugar.

• Location choice
• Human capital
• Agglomeration
• Local public sector

• 地点选择
• 人类资本
• 聚集
• 地方公共部门

• Choix d'emplacement
• Capital humain
• Agglomération
• Secteur public local

• Standortwahl
• Humankapital
• Agglomeration
• Kommunaler Sektor

• Elección del lugar
• Capital humano
• Aglomeración
• Sector público local

JEL classifications:

• J24
• J61
• R23

Acknowledgments

The authors are grateful for excellent proofreading and editing by Cyndi Berck; and for the helpful comments made by four anonymous referees.

Notes

1. This is, for example, an apparent pattern in some parts of the United States, Canada and in the European Union, e.g. within inland areas of the northern regions of Finland, Norway and Sweden.

2. Migration may contribute to convergence in regional per capita incomes (e.g. Barro and Sala-i-Martin, 2004; on Swedish data, e.g. Aronsson et al., 2001; Østbye and Westerlund, 2007), but income and population growth are concentrated in cities and metropolitan areas within regional units.

3. The urban economics literature offers numerous studies, typically involving the choice of residence within and around urban centres.

4. Another major reason for the choice of these cohorts is that the present paper will be followed by a second study pertaining to the same individuals at mid-age – when child-rearing and graduation from tertiary education are of importance for their choice of location. The reason for not choosing even younger cohorts is partially governed by the availability of information on school grades.

5. This is a general problem, for example, in the literature on the choice of travel mode or travel route; theoretical alternatives chosen by zero or by very few individuals are excluded or collapsed into an outside option consisting of many seldom chosen alternatives.

6. For further description on the Linnaeus database, see Bonita et al. (2011).

7. Nomenclature of Units for Territorial Statistics (NUTS) is a classification of administrative boundaries made by Eurostat. NUTS-2 incorporates counties or a group of counties with a population between 28 000 and 3 million.

8. The student benefit is split into two parts: a loan and a grant. The grant is generally claimed by all students enrolled at university, no matter the income level, making it a strong indicator of university studies.

9. Studies on public expenditures and location choice show mixed results. Friedman (1981) finds that people are attracted by high local public expenditures, while Quigley (1985) finds a slightly negative effect. On Swedish data, see Westerlund and Wyzan (1995).

10. Dahlberg et al. (2012) show the importance of the public service when it comes to making residential location choices. Municipalities with lower unemployment and larger population size attracted migrants. They found no significant result on housing prices since the signalling effect is twofold: it reflects an attractive region, but at the same time the living expenses will be high.

11. Regions’ attractiveness in young people's eyes also depends on various types of amenities (e.g. Greenwood and Hunt, 1989; Ferguson et al., 2007). These are presumably partially captured by the tax base and share of elderly and other regional attributes in the model. Regional differences in natural amenities are relatively low. Experiments with indicators for coast/shoreline were not successful due to problems with multicollinearity.

12. One reason for the choice of using 9th grade GPA is that it is unlikely that a student's preferences measured at age 19 could influence GPA at age 15.

13. The coefficients of the lower model can be estimated either by dividing by the scale parameter or by not dividing by the scale parameter. Heiss (2002) finds that if there is no coefficient that is common across the nests, dividing by λ_k is not necessary and the model is still consistent with utility maximization.

14. This approach is referred to as the Random Utility Model 2 in, for example, Silberhorn et al. (2008) and Hensher and Greene (2002).

15. This is not an unusual finding in migration studies including all individuals of working age. Individual experience of unemployment is often more important for the decision to relocate than regional employment/unemployment rates.

16. For example, by dividing the university towns Umeå and Luleå into two different locations or merging locations with small choice probabilities. Splitting the two university towns in the north increased the standard errors and affected the results for the non-students.

Additional information

Funding

This work was supported by the Swedish Research Council [Linnaeus grant number 2006-21576-36119-66]; and by the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (FORMAS) [grant number 251-2007-2079]. Sofia Tano thanks the Graduate School in Population Dynamics and Public Policy, Umeå University, for financial support.