Institutions and Regional Development : A panel data study of EU regions

Abstract

This PhD thesis contributes to the body of knowledge that uses institutions to explain the differences in economic growth and development across regions. While theoretical (e.g., Rodríguez-Pose, 2013, Rodríguez-Pose, 2020) and empirical studies (e.g., Beugelsdijk & Smulders, 2009; Ezcurra & Rodríguez-Pose, 2013; Putnam, Leonardi, & Nanetti, 1993; Rodríguez-Pose & Di Cataldo, 2015) have advanced in this direction, several issues remain unexplored. Overall, we tend to know less about which institutions and their combinations matter for regional economic development. We also know less about how, why and when they matter as well as change over time. This PhD thesis endeavours to address these issues in four papers and uses the level of GDP per capita as a measure of economic growth and development. Doing this enriches our theoretical understanding of how regional institutions affect economic development as well as informs evidencebased and place-sensitive policies. Paper I examines formal institutions focusing on the quality of government and degree of decentralisation, whereas Paper II examines informal institutions focusing on social capital. Paper III and IV examine the interplay between formal and informal institutions. Specifically, Paper III examines the interaction between social trust and the quality of government, whereas Paper IV looks at the structural relationships between the same variables but adds political trust. The four papers employ secondary data measuring institutions and other regional socioeconomic and demographic characteristics across regions in 21 EU countries, covering eight waves from 2002 to 2016. The analyses use mainly fixed effects panel data in the first three papers. The fourth paper uses a structural equation model (SEM) on pooled cross-sectional data. Overall, the findings show a complex and interdependent process between formal and informal institutions as well as specific forms of the same type of institutions, and other contextual factors such as human capital that affect economic growth. Specifically, Paper I shows that the quality of government is a better predictor of economic growth than decentralisation but also mediates the economic returns of the latter. Paper II shows that bridging social capital promotes economic growth, and the opposite happens with bonding social capital. At the same time, there is no significant differences of their effects on each other. It also shows that human capital moderates bonding and bridging social capital, reducing the negative effects of the former and working as a substitute for the latter. Paper III shows that social trust and the quality of government work as substitutes and both matter for economic growth. Paper IV shows that social trust and the quality of government, similar to political trust, have a direct positive association with economic growth. Also, through political trust, they have an indirect and positive relationship with economic growth. These findings have several policy implications for regional economic development. Broadly, they suggest the importance of place-sensitive policies and balance attention to different types and forms of institutions depending on the conditions. Specifically, the findings suggest the following: first, the need to improve the quality of government before implementing decentralisation reforms. Second, bridging social capital and human capital are both effective tools for promoting economic development. Third, there is a flexibility to use either social trust or the quality of the government to promote economic development. Fourth, attention should be paid to both bottom-up processes of social capital and top-down processes involving the quality of government to promote political trust and promote economic development. However, the PhD thesis is limited in the generalisability of its findings, choice of variables, econometric methods and the level of analyses. Thus, expanding the regions beyond the EU will increase the generalisability of the findings. Similarly, future research should broaden the scope of institutions and measures of economic development. The same applies to the use of other econometric methods such as spatial econometrics and longitudinal SEM instead of a pooled cross-sectional analysis. Respectively, these two approaches will allow us to explicitly examine the spatial effects of institutions and how their structural relationships change over time. Furthermore, extending the regional level of analysis to a multi-level to include the individual level, allows us to examine how agency and the structure or institutions shap each other and jointly affect economic development.