Recovery Risk and Labor Costs in Public–Private Partnerships: Contractual Choice in the US Water Industry

Abstract

We use an ordered logistic model to empirically examine the factors that explain varying degrees of private involvement in the US water sector through public–private partnerships. Our estimates suggest that a variety of factors help explain greater private participation in this sector. We find that the risk to private participants regarding cost recovery is an important driver of private participation. The relative cost of labour is also a key factor in determining the degree of private involvement in the contract choice. When public wages are high relative to private wages, private participation is viewed as a source of cost savings. We thus find two main drivers of greater private involvement: one encouraging private participation by reducing risk, and another encouraging government to seek out private participation in lowering costs.

Key Words:

• Privatisation
• public–private partnerships
• water
• contracting out

Acknowledgements

This research received financial help from the Spanish Government under Project ECO2009-06946 and ECO2012-38004, and the Regional Government of Catalonia under project SGR2009-1066. Germà Bel also acknowledges support from ICREA-Academia. We are grateful for comments received at the 1st International Barcelona-Lisbon workshop on Public Private Partnerships. We have benefited from useful suggestions from Trevor Brown and two anonymous reviewers.

Notes

1. It is worth noting that other non-empirical papers have provided interesting insights into factors influencing privatisation of water services, such as interest groups and institutional structures of power (Fitch 2007). See also Bel et al. (2007).

2. Bel and Fageda (2008) conduct an empirical analysis where they merge data on solid waste and on water. They find that water is less prone to contracting than solid waste.

3. In a more recent paper, Picazo-Tadeo et al. (2012) use the same database to conduct a detailed study on political and ideological factors.

4. Available from: http://www.taxfoundation.org/research/topic/9.html [Accessed 16 November 2011].

5. Data for this variable were obtained from Table 1 ‘Tax and expenditure limitations currently imposed statewide on local governments’, in Mullins and Wallin (2004). See the Appendix for a more detailed description of the construction of this variable.

6. There are, however, several other variables included in the Public Works Financing database, such as contract duration and private firm identification, which cannot be exploited due to the number of missing values such that the number of observations would be significantly reduced. Indeed, the inclusion of the Contract_Size variable reduces the number of observations from 165 to 108.

7. Variance Inflation Factor (VIF) = 1.16 < 10 (rule of thumb).

8. Geddes and Wagner (2012), however, find a large effect of political party affiliation on the likelihood that a state passes a public–private partnerships enabling law in a particular year.