Tree-based approaches for understanding growth patterns in the European regions
Keywords:Regional economic growth, European Union regions, data mining, decision trees, multivariate adaptive regression splines
AbstractThe paper describes an empirical analysis to understand the main drivers of economic growth in the European Union (EU) regions in the past decade. The analysis maintains the traditional factors of growth used in the literature on regional growth - stage of development, population agglomeration,
transport infrastructure, human capital, labour market and research and innovation - and incorporates the institutional quality and two variables which aim to reflect the macroeconomic conditions in which the regions operate. Given the scarcity of reliable and comparable regional data at the EU level, large part of the analysis has been devoted to build reliable and consistent panel data on potential factors of growth. Two non-parametric, decision-tree techniques, randomized Classication and Regression Tree and Multivariate Adaptive Regression Splines, are employed for their ability to address data complexities such as non-linearities and interaction eects, which are generally a challenge for more traditional statistical procedures such as linear regression. Results show that the dependence of growth rates on the factors included in the analysis is clearly non-linear with important factor interactions. This means that growth is determined by the simultaneous presence of multiple stimulus factors rather than the presence of a single area of excellence. Results also conrm the critical importance of the macroeconomic framework together with human capital as major drivers of economic growth of countries and regions. This is overall in line with most of the economic literature, which has persistently underlined the major role of these factors on economic growth but with the novelty that the macroeconomic conditions are here incorporated. Human capital also has an important role, with low-skilled workforce having a higher detrimental eect on growth than high-skilled. Not surprisingly, other important factors are the quality of governance and, in line with the neoclassical growth theory, the stage of development, with less developed economies growing at a faster pace than the others. The evidence given by the model about the impact of other factors on economic growth such as those on the quality of infrastructure or the level of innovation seems to be more limited and inconclusive. The analysis conclusions support the reinforcement of the EU economic governance and the conditionality mechanisms set in the new architecture of the EU regional funds 2014-2020 whose rationale is that the eectiveness of the expenditure is conditional to good institutional quality and sound economic policies.
Acemoglu, D., Johnson, S., Robinson, J., and Thaicharoen, Y. (2003). Institutional causes, macroeconomic symptoms: Volatility, crises, and growth. Journal of Monetary Economics, 50:49-123.
Agresti, A. (1990). Categorical Data Analysis. New York: John Wiley & Sons. Austin, P. C. (2007). A comparison of regression trees, logistic regression, generalized additive models, and multivariate adaptive regression splines for predicting AMI mor- tality. Statistics in medicine, 26:2937-2957.
Barro, R. (1989). Economic growth in a cross section of countries. NBER Working Paper n. 3120.
Breiman, L., Friedman, J., Olshen, R., and Stone, C. (1984). Classi_cation and regression trees. Wadsworth & Brooks.
Charron, C. (2013). From Aland to Ankara: European Quality of Government Index. Working paper series 2013:11, QOG The Quality of Government Institute.
Charron, N., Dijkstra, L., and Lapuente, V. (2014). Regional governance matters: Quality of government within european union member states. Regional Studies, 48(1):68-90.
Cristelli, M., Tacchella, A., and Pietronero, L. (2015). The heterogeneous dynamics of economic complexity. PLOS ONE, 10(2):1-15.
Curtis, P. and Kokotos, P. (2009). A decision tree application in tourism based re- gional economic development. Tourismos: an international multidisciplinary journal of tourism, 4(2):169-178.
Dall'erba, S. and Le Gallo, J. (2008). Regional convergence and the impact of european structural funds 1989-1999: A spatial econometric analysis. Papers in Regional Science, 87(2):219-244.
De Veaux, R., Psichogios, D., and Ungar, L. (1993). A comparison of two nonparametric estimation schemes: MARS and neural networks. Computers in Chemical Engineering, 17(8):819-837.
Deichmann, J., Eshghi, A., Haughton, D., Sayek, S., and Teebagy, N. (2002). Application of multiple adaptive regression splines (MARS) in direct response modeling. Journal of Interactive Marketing, 16(4):15-27.
DG ECFIN (2012). Scoreboard for the surveillance of macroeconomic imbalances. Tech- nical Report 92, European Economy Occasional Papers.
Dijkstra, L. and Poelman, H. (2012). Cities in europe. the new OECD-EC de_nition. Technical Report WP 01/2012, European Commission - DG for Regional and Urban Policy.
Durlauf, S. and Johnson, P. (1995). Multiple regimes and cross-country growth behaviour. Journal of Applied Econometrics, 10(4):365-384.
Friedman, J. (1991). Multivariate adaptive regression splines. Annals of Statistics, 19:1- 141.
Hastie, T., Tibshirani, R., and Friedman, J. (2001). The elements of statistical learning. Data mining, inference and prediction. Springer.
Knack, S. and Keefer, P. (1995). Institutions and economic performance: cross-country tests using alternative institutional measures. Economic and Politics, 7(3):207-227.
Krugman, P. (1998). What's new about the new economic geography? Oxford Review of Economic Policy, 14(2):7-11.
Kwok, C. and Tadesse, S. (2006). National culture and _nancial systems. Journal of International Business Studies, 37:227-247.
Leathwick, J. R., Rowe, D., Richardson, J., Elith, J., and Hastie, T. (2005). Using multivariate adaptive regression splines to predict the distributions of new zealands freshwater diadromous _sh. Freshwater Biology, 50:2034-2052.
Lucas, R. (1988). On the mechanics of economic development. Journal of Monetary Economics, XXII:3-42.
Mankiw, N. G., Romer, D., and Weil, D. (1992). A contribution to the empirics of economic growth. The Quarterly Journal of Economics, pages 407-437.
Mardia, K. and Kent, J.T.and Bibby, J. (1979). Multivariate Analysis. San Diego, U.S.:Academic Press, INC.
Mezrich, J. (1994). When is a tree a hedge? Financial Analysts Journal, pages 75-81.
Mood, A. M., Graybill, F. A., and Boes, D. C. (1974). Introduction to the Theory of Statistics 3rd Edition. Mc Graw Hill.
OECD (2012). Promoting growth in all regions. OECD publishing.
Pescatori, A., Sandri, D., and Simon, J. (2014). Debt and growth: Is there a magic threshold? Working Paper WP/14/34, International Monetary Fund.
Ramajo, J., Marquez, M., Hewings, G., and Salinas, M. (2008). Spatial heterogeneity and interregional spillovers in the European Union: Do cohesion policies encourage convergence across regions? European Economic Review, 52:551-567.
Rodriguez-Pose, A. (2013). Do institutions matter for regional development? Regional Studies, 47(7):1034-1047.
Rodriguez-Pose, A. and Fratesi, U. (2004). Between development and social policies: The impact of European structural funds in objective 1 regions. Regional Studies, 38(1):97-113.
Rodriguez-Pose, A. and Garcilazo, E. (2013). Quality of government and the returns of investment: Examining the impact of cohesion expenditure in european regions. OECD Regional development working papers 2013/12, OECD Publishing.
Rodrik, D., Subramanian, A., and Trebbi, F. (2004). Institutions Rule: The Primacy of Institutions over Geography and Integration in Economic Development. Journal of Economic Growth, 9(2):131-165.
SAS (2014). The HP-SPLIT procedure, on-line material. Technical report, SAS Institute Inc., Cary, NC, USA.
Sobol', I. M. (1993). Sensitivity analysis for non-linear mathematical models. Mathematical Modelling and Computational Experiment, 1:407-414.
Solow, R. (1956). A contribution to the theory of economic growth. Quarterly Journal of Economics, 70(1):65-94.
Stelder, D. (2013). Changes in road infrastructure and accessibility in Europe since 1960. Tender reference nr 2012.ce.16.bat.040, European Commission, DG for Regional and Urban policy.
Sugihara, G., May, R., Ye, H., Hsieh, C., Deyle, E., Fogarty, M., and Munch, S. (2012). Detecting causality in complex ecosystems. Science, 338:496-500.
Weir, N., Fayyad, U. M., and Djorgovski, S. (1995). Automated star/galaxy classification for digitized POSS-II. The Astronomical Journal, 109(6):2401-2414.
How to Cite
REGION is an open journal, and uses the standard Creative Commons license: Copyright We want authors to retain the maximum control over their work consistent with the first goal. For this reason, authors who publish in REGION will release their articles under the Creative Commons Attribution-NonCommercial license. This license allows anyone to copy and distribute the article for non-commercial purposes provided that appropriate attribution is given to REGION and the authors. For details of the rights authors grant users of their work, see the "human-readable summary" of the license, with a link to the full license. (Note that "you" refers to a user, not an author, in the summary.) Upon submission, the authors agree that the following three items are true: 1) The manuscript named above: a) represents valid work and neither it nor any other that I have written with substantially similar content has been published before in any form except as a preprint, b) is not concurrently submitted to another publication, and c) does not infringe anyone’s copyright. The Author(s) holds ERSA, WU, REGION, and the Editors of REGION harmless against all copyright claims. d) I have, or a coauthor has, had sufficient access to the data to verify the manuscript’s scientific integrity. 2) If asked, I will provide or fully cooperate in providing the data on which the manuscript is based so the editors or their assignees can examine it (where possible) 3) For papers with more than one author, I as the submitter have the permission of the coauthors to submit this work, and all authors agree that the corresponding author will be the main correspondent with the editorial office, and review the edited manuscript and proof. If there is only one author, I will be the corresponding author and agree to handle these responsibilities.