Locational Choices and the Costs of Distance: Empirical Evidence for Dutch Graduates

Abstract

This study identifies and analyzes the effects of university/college graduates’ personal, household and employment characteristics as well as the relative attributes of their study and work locations on their college-to-work distances. The results illustrate that graduates are drawn to prospering regions with ample job opportunities, supposedly in order to advance their careers. Furthermore, we model and estimate the effect of college-to-work distance on college-to-residence and commuting distances. We find that graduates tend to choose their places of residence so as to balance their commuting distances and the distances to their previous places of study.

Résumé

cette étude identifie et analyse les effets des caractéristiques au niveau personnel, domestique et de l’emploi de diplômés universitaires, ainsi que les attributs relatifs à leur lieu d’étude et de travail sur les distances faculté—emploi. Les résultats indiquent que les étudiants sont attirés vers des zones prospères, offrant de nombreux débouchés professionnels, probablement pour l’avancement de leur carrière. En outre, nous modélisons et nous évaluons l’effet de la distance faculté—travail sur la distance faculté—domicile et le rayon de déplacement jusqu’au lieu de travail. Nous en concluons que les diplômés ont tendance à choisir leur lieu de domicile de façon à équilibrer leur distance de déplacement et la distance qui les sépare du lieu de leurs études

Resumen

Este estudio determina y analiza los efectos de las características personales, familiares y laborales de los egresados de universidades e instituciones de educación superior, al igual que los atributos relativos a sus lugares de estudio y trabajo en relación con las distancias entre el centro de formación y el trabajo. Los resultados ilustran que los egresados son absorbidos por las regiones prósperas que tienen amplias oportunidades de trabajo, probablemente para avanzar en sus carreras. Además, elaboramos el modelo y estimamos el efecto de la distancia entre el centro de formación y el trabajo en relación con las distancias entre el centro de formación y la residencia y de desplazamiento diario. Concluimos que los egresados tienden a elegir sus lugares de residencia de manera que haya un equilibrio entre las distancias de desplazamiento diarias y las distancias para llegar a sus lugares de estudio anteriores

摘要

本研究旨在发现和分析大学毕业生的个人、住所和就业特点及其学校和工作地点的有关特性对其学校到工作地点的距离的影响。研究结果表明, 为了追求更美 好的前程, 毕业生倾向于去工作机会较多的发达地区。我们还建立了一个模型, 并评估了学校到工作地点的距离对学校到住所的距离和通勤距离的影响。我们发现毕 业生在选择住所时, 更倾向于选择通勤距离和到其毕业院校的距离都比较适当的地方

Keywords:

• Distance
• commuting
• college-to-work
• college-to-residence

JEL Classification:

• R23
• R41

Notes

1. Evidently, a person is likely to have additional locations of interest, e.g. the original hometown he or she might want to return to after graduation. Regrettably, the data at hand did not provide any information regarding a student's original place of residence before attending university/college. Yet, it should be noted that out of the 93.1% of the Dutch students having received a student grant (in 2007/2008), 43.5% were still living with their parents (KENCES Kenniscentrum Studentenhuisvesting, 2010). Taking this into consideration, and also acknowledging that our sample consists of both college and university students, we believe that it is safe to assume that the majority of students in our sample studies and lives in or nearby their original hometown. In this respect, the Netherlands are different from e.g. the USA, where it is common to move across the country in order to attend college, and where the population is much more geographically mobile.

2. We assume travel cost to be a quadratic function of distance, as the travel time between two locations will be subject to higher variance. Counterbalancing efforts (e.g. leaving earlier to be at work on time) thus generate a convex loss function (Juster & Stafford, 1991).

3. Theoretically, the dependent variables would also have to be right-censored, since the maximum distance between two Dutch municipalities is 314.48 km. Yet, as values of this magnitude are exceptionally rare, we abstained from right-censoring.

4. In 2007/2008, 206,800 (40.3%) of the 512,800 Dutch students studied at a university, and 306,000 (59.7%) attended a college. Both universities and colleges are spread across the country as much as population is, and there are no major dissimilarities between the different institutions regarding status or quality.

5. Many Dutch students work part-time while studying, often even having more than one ‘small’ job, all of which may be unrelated to one's field of study. By only selecting graduates who work in the same job at least three days per week, we expect to limit the dataset to those with degree-related post-graduation jobs.

6. More than 95% of the firms are single-site firms. Since the exact location of a person's workplace cannot be determined for those working in firms with subsidiaries in various locations, the latter had to be excluded from the analysis.

7. Again, by imposing this restriction, we expect to exclude pre-graduation student jobs.

8. In the dataset at hand, a person's place of residence as well as all other personal and household characteristics are only established once a year (last Friday in September).

9. In 2004, the Netherlands consisted of 483 municipalities. Since information on employment growth and accessibility was not available for all municipalities, eight municipalities had to be excluded from the analysis. Graduates who were working and/or living in one of these municipalities were consequently excluded as well.

10. The values for sectoral specialization depend on the sector (two-digit NACE level) the graduate is employed in.

11. With the available data, it is only possible to determine whether or not an individual is living in a registered partnership. While in most cases, registered partners are indeed married, the variable also captures those who registered their partnership without getting married.