The relevance of commuting zones for regional spending efficiency

ABSTRACT

We use data envelopment analysis (DEA) efficiency scores to show that clustering municipalities into encompassing regional clusters improves spending efficiency of single stand-alone municipalities. We propose a new geographic aggregation based on municipalities-to-municipalities commuting flows, defined using hierarchical cluster analysis. Our example for Portugal shows that from an output-oriented perspective between 83% and 98% of municipalities would increase their efficiency scores, while from an input-oriented perspective between 86% and 98% of municipalities would also be better off in terms of efficiency. Then using a linear regression model, we find that population increases positively affects the efficient scores (via scale economies). Also, increases in the share of high-educated and poorer residents leads to higher efficiency scores.

KEYWORDS:

• Public spending efficiency
• local government
• data envelopment analysis (DEA)
• commuting zones

JEL CLASSIFICATION:

• C14
• H72
• R50

Acknowledgements

We thank two anonymous referees for very useful suggestions.

Notes

¹ For statistical purposes, the EU redefined the Portuguese territory into NUTS regions. The NUTS system subdivides the country into three levels: NUTS I (Portugal mainland and 2 autonomous regions of Azores and Madeira), NUTS II (7 regions) and NUTS III (30 subregions). These latter classifications were developed for the purpose of delivering structural funds for less favoured regions and subregions.

² Coelli, Rao, and Battese (2002) and Thanassoulis (2001) offer introductions to DEA.

³ This is the equivalent envelopment form, derived by Charnes, Cooper, and Rhodes (1978), using the duality property of the multiplier form of the original programming model.

⁴ These data are available from the Portuguese Statistic Office’s website under the variable names ‘Commuting of the employed or student resident population by place of residence or destination and place of destination or residence’ and ‘Resident population by place of residence (at the date of Census), sex and by main source of livelihood’.

⁵ For the purpose of our analysis, we exclude the two autonomous regions of Azores and Madeira and only consider the mainland region.

⁶ See Hamilton (1983), Hayes, Razzolini, and Ross (1998), De Borger et al. (1994), and De Borger and Kerstens (1996).

Additional information

Funding

UECE is supported by Fundacão para a Ciência e a Tecnologia (Portuguese Foundation for Science and Technology) through the project [PEst-OE/EGE/UI0436/2011]. Financial support was provided by the Fundacão para a Ciência e a Tecnologia (Portuguese Foundation for Science and Technology) through the project [PTDC/IIM-ECO/3513/2012]. The opinions expressed herein are those of the authors and not necessarily those of their employers.