The rule of law and educational choices: evidence from Italian regions

ABSTRACT

The rule of law and educational choices: evidence from Italian regions. Regional Studies. Educational choices about post-secondary fields of study vary across space and time due to psychological, social and economic motivations. A strand of the literature has focused on the role of institutional environment. In this vein, this paper emphasizes the link between regional institutions and students’ preferences by pointing out how effective rule of law decreases resources that rent-seekers may expropriate, thus reducing incentives to choose careers devoted to protect producers from rent-seeking. The econometric investigation on Italian regions supports this view, finding that institutional quality and rule of law in students’ region of origin are relevant to their choice.

摘要

法律规范和教育选择：来自意大利区域的证据。Regional Studies. 高等教育学习领域的教育选择，因心理因素，社会与经济动机，在时间和空间上有所不同。有文献研究聚焦制度环境的角色。本文于此一观点中，指出法律规范如何有效降低寻租者或能利用的资源，因而减少选择致力于保护生产者不受寻租影响的动机，藉此强调区域制度和学生偏好之间的连结。对意大利区域的计量经济探讨，支持此一观点，并发现学生来自的区域中的制度品质和法律规范，与其选择相关。

RÉSUMÉ

L’état de droit et les choix éducatifs: des résultats des régions italiennes. Regional Studies. Les choix éducatifs en ce qui est des domaines d’étude post-secondaires varient à travers l’espace et dans le temps à cause des motivations psychologiques, sociales et économiques. Un volet de la documentation a porté sur le rôle du milieu institutionnel. Dans cette optique, l’article souligne le lien entre les institutions régionales et les préférences des étudiants en indiquant comment le respect de l’état de droit réduit les ressources que peuvent exproprier les chercheurs de rente, diminuant ainsi les incitations à choisir une carrière consacrée à protéger les producteurs de la recherche de rente. L’examen économétrique auprès des régions italiennes soutient ce point de vue, constatant que la qualité des institutions et l’état de droit dans la région d’origine des étudiants ont un rapport avec leur choix.

ZUSAMMENFASSUNG

Rechtsstaatlichkeit und Bildungswahl: Belege aus italienischen Regionen. Regional Studies. Die Bildungswahl bei postsekundären Studiengängen fällt aufgrund psychologischer, sozialer und wirtschaftlicher Motivationen in puncto Raum und Zeit unterschiedlich aus. Im Mittelpunkt von einem Literaturzweig steht die Rolle der institutionellen Umgebung. In diesem Sinn betonen wir in diesem Beitrag den Zusammenhang zwischen regionalen Institutionen und den Präferenzen der Studenten, indem wir herausarbeiten, wie sich durch wirksame Rechtsstaatlichkeit der Umfang der Ressourcen verringert, die sich Rent-Seeker aneignen können, was wiederum den Anreiz zur Wahl von Berufen zum Schutz von Produzenten vor Rent-Seeking mindert. Bei einer ökonometrischen Untersuchung von italienischen Regionen finden sich Belege für diese Perspektive, da sich zeigt, dass die institutionelle Qualität und Rechtsstaatlichkeit in den Herkunftsregionen der Studenten für ihre Wahl relevant sind.

RESUMEN

El Estado de derecho y las opciones educativas: evidencia de las regiones italianas. Regional Studies. Las opciones educativas sobre las áreas de estudios en la educación post-secundaria varían en el espacio y el tiempo debido a motivaciones psicológicas, sociales y económicas. Una línea de la bibliografía se centra en el papel del entorno institucional. En este sentido, en este artículo subrayamos el vínculo entre las instituciones regionales y las preferencias de los estudiantes al señalar cómo un Estado de derecho eficaz disminuye los recursos que podrían expropiar los especuladores, reduciendo así los incentivos de elegir carreras dedicadas a proteger a los productores de los especuladores. En la investigación econométrica en las regiones italianas se apoya esta idea, y se observa que la calidad institucional y el Estado de derecho en las regiones de origen de los estudiantes son relevantes a las opciones que eligen.

KEYWORDS:

• rule of law
• institutional quality index
• educational choices
• Italy

关键词:

• 法律规范
• 制度品质指标
• 教育选择
• 意大利

MOTS-CLÉS:

• état de droit
• indice de la qualité des institutions
• choix éducatifs
• Italie

SCHLÜSSELWÖRTER:

• Rechtsstaatlichkeit
• Index der institutionellen Qualität
• Bildungswahl
• Italien

PALABRAS CLAVES:

• Estado de derecho
• índice de calidad institucional
• opciones educativas
• Italia

JEL:

• D2
• E2
• I21

Notes

1. The MNL model seems a good alternative candidate to carry out the analysis for several reasons: the dependent variable of the regression equation includes multiple non-ordered responses; the dataset contains many strictly individual data for which the utilization of MNL is suggested by many econometricians (Cameron & Trivedi, 2010); and it has been used by other studies on the determinants of the fields of studies (Nguyen & Taylor, 2003; Boudarbat, 2008).

2. Regression parameters  of equation (1) are assumed to be region specific and modelled to allow for random regional components, i.e., , where k and r are indexes for parameters and regions respectively. As a consequence, individual observations:

become:

where, on the right-hand side, the hierarchical structure of the data is represented by a deterministic part (the first addends) and a stochastic part (the last addends between parentheses). Using STATA, the estimation is made by employing the command MELOGIT and syntax: melogit depvar fe_equation [|| re_equation] [|| re_equation  … ] [, options].

3. Provincial rather regional indicators are used because the authors believe that this is the most suitable level to represent the specific economic condition of families and the cultural background behind the choice of a future occupation in a productive or DEFEN career. On the other hand, regional data are employed on the quality of faculties because some provinces do not have universities.

4. For further details on IQI, see Nifo and Vecchione (2014) and https://sites.google.com/site/institutionalqualityindex/.

5. For logit estimations, beside the usual statistics (number of observations, pseudo- and value of the likelihood function), the percentage correctly predicted is reported. Clustered standard errors are shown in parentheses under coefficients’ estimates. Several tests on multiple exclusion of control variables were run by using both Wald and likelihood ratio tests, which always allow one to reject the hypothesis of no effect on students’ choice. For multinomial logit estimations, the validity of the independence of irrelevant alternatives (IIA) assumption was tested by using a Hausman test, which always yielded negative values. According to Hausman and McFadden (1984), a negative value for the Hausman test can be taken as support for the null hypothesis of IIA validity. The Small–Hsiao test also confirmed the validity of IIA.

6. All specifications of equation (1) reported in Tables 1 and 2 have also been estimated by a probit model. For brevity, results are not shown, since in terms of marginal impact of regressors they turn out to be very similar to logit estimations. Probit estimations are available from the authors upon request.

7. An alternative measure of the marginal impact of regressors is given by the partial effect at the average (PEA), i.e.:

routinely supplied by most econometric packages. However, the economic sense of PEAs is disputable in the presence of dummy, log and quadratic continuous variables (Wooldridge, 2009).

8. The measure of the impact of changes in other dimensions of IQI on the probability of choosing TECH or DEFEN is much lower, similar to IQI in the case of REGUL and VOICE, and insignificant for CORR and GOVERN.

9. Variables SCHOTYPE and (in a few cases) ENGEVAL become strongly significant. Jointly considering the coefficients of SCHOTYPE and SCHOGRAD, it seems that high-grade students coming from science and literature high schools (liceo) have a consistent order of preferences in the choice of university majors: first TECH, then DEFEN and, finally, GNRL.