Port allocative efficiency and port devolution: a study for the Spanish port authorities (1992-2016)

ABSTRACT

This paper analyses the effect of port devolution on the Spanish port authorities’ allocative efficiency across periods with different regulations during the port devolution process in Spain. To do this, we first obtain a measure of allocative efficiency from a distance system of equations using two different approaches: the error components approach and the parametric approach. Then, we evaluate whether port devolution has had effects on allocative efficiency by applying a quantile regression approach, taking into account control variables related to port characteristics. Our results suggest that allocative inefficiencies occur in the Spanish port system. We also demonstrate that port devolution reforms, both internal and external port characteristics have had a significant effect on the port authorities’ allocative inefficiency.

KEYWORDS:

• Allocative efficiency
• distance function
• quantile regression
• port devolution

Acknowledgement

We thank the editor and the anonymous referees for their useful comments and suggestions.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. Therefore, four periods of regulation can be distinguished. The first period of regulation covers the years from 1992 to 1996 with Law 27/1992 in force. The second begins in 1997 when Law 62/1997 is passed and ends in 2002. The third comprises the years between 2003 and 2009, being in force Law 10/2003. Finally, the fourth begins with the adoption of Law 33/2010 in 2010 and finishes in the last year of the sample.

2. Studies such as Núñez-Sánchez and Coto-Millán (2012), Rodríguez-Álvarez, Fernández-Blanco, and Lovell (2004) and (2007) and, Serebrisky et al. (2016) estimate input-oriented distance function for the port industry.

3. Systematic inefficiencies in the Spanish port sector can arise by several issues: public regulation, lumping investments, difficulties to adjust capital or strategic planning.

4. We consider that this variable might be a valid proxy for the degree of private participation within the port activity given that it presents heterogeneity among the different Spanish port authorities.

5. Parameters related to port authority-specific dummies are not displayed in Table 3. They are available upon request to the authors.

6. Annex 1 shows the kernel density distribution of the relative allocation indices ($z_{ij}$). It can be seen that almost all $z_{k,l}$ and $z_{k,ic}$ indices present values less than one. This finding implies that overuse of capital in relation to labour and intermediate consumption occurs along the port devolution period. However, the distribution of these indices suggests that the level of overcapitalization varies across port authorities. On the other hand, the distribution of the $z_{l,ic}$ indices is partially concentrated around the optimal allocation. Although, throughout the density distribution, it is possible to find observations that overuse labour in relation to intermediate consumptions; observations that efficiently allocate these inputs; or observations that underuse labour in relation to intermediate consumption. These results suggest that a quantile regression (QR) could provide a richer characterization than the OLS regression.

7. Annex 2 shows the values of the relative indices for each percentile considered.