Incentives in Bus Concession Contracts: A Review of Several Experiences in Latin America

Abstract

This paper reviews the incentive structure of concession contracts in several Latin American transit reforms. It also presents a conceptual analysis of the optimal design of concession contracts. The conceptual discussion and case studies reviewed indicate that payment to operators should be linked to operational variables and that some demand risk should optimally be transferred to operators. Performance standards linked to fines and penalties are not sufficient to guarantee good quality of service, particularly in citywide reforms and where institutional capacity ― in terms of size, experience and expertise of staff ― is lacking and regulatory processes are slow, bureaucratic and cumbersome. This review may be useful for policy-makers designing transit reforms in other countries. The policy lessons of the paper are particularly relevant to cities in the developing world but they are also important for reforms in other countries.

Keywords:

• bus contracts
• payment formulas
• incentives
• transit reform
• Latin America
• performance standards

Acknowledgements

Andrés Gómez-Lobo thanks the Clean Air Institute for having provided funds to develop this research. This paper is partly based on the policy paper prepared for the Clean Air Institute (Briones & Gómez-Lobo, 2013). Julio Briones thanks Across Latitudes and Cultures-BRT.

Notes

1. See Finn and Mulley (2011) and references therein for an overview of developing country experiences. Gwilliam, Zhi, and Finn (2009) provide a review specific to China. EMBARQ (2010) provides a summary and characterization of reforms in Latin America and Asia.

2. For a wider risk analysis of concession contracts, see Marques and Berg (2010) and Cruz and Marques (2012). These papers deal with infrastructure concessions (water and ports, respectively) that have many issues in common with transit concession contracts.

3. Examples include Transantiago, in Santiago, Chile and the SITP reform in Bogotá, Colombia. The authorities in Lima, Peru, are also considering such a reform.

4. The case study approach also highlights some of the pitfalls encountered in the Santiago experience that is useful for policy-makers in its own right.

5. At a deeper level, the issue of incentives and service quality in a contractual relationship is related to the asymmetry of information between operators and the authorities. As mentioned by González-Díaz and Montoto-Sánchez (2001), contract theory offers two solutions to this problem. The first is for the less-informed party to invest in monitoring and control and the second is to design an incentive system to align the agent's and the principle's interests. These two solutions are related to the performance standards approach and the payment mechanism discussed next.

6. In the transport literature, demand risk in contracts is often couched in terms of gross cost or net cost contracts (Hensher & Brewer, 2001). In the regulatory literature, it is related to the ‘power’ (low or high) of a contract. We prefer to use the term ‘demand-risk’ directly in what follows.

7. However, safety concerns linked to faulty maintenance of vehicles is still an issue that would have to be addressed explicitly in the service standards defined in the contract.

8. We thank an anonymous reviewer for raising this point.

9. Minimum performance standards can be set in the contract or some kind of benchmarking can be applied among operators using the indicators related to service obligations. In this last case, care must be taken to consider aspects of performance that are reasonably under the control of the operator, under the control of the authorities or essentially exogenous.

10. A detailed review of this experience can be found in Gómez-Lobo (2012). See also Beltran, Gschwender, and Palma (in press). An obvious question that comes up is how were the incentive properties of the contract maintained under constant renegotiation? Or how could the authorities transfer risk to operators without being charged a higher risk-premium? It must be mentioned that the authorities started all contract renegotiations. Therefore, there is no evidence of opportunistic behaviour on the part of operators forcing renegotiations. Also, the real danger of a general collapse of the system, particularly between 2007 and 2008, meant that the authorities had a strong bargaining position to impose needed contractual changes upon operators.

11. The GPS signals could now be used to see exactly where each bus was operating, the headway between buses and whether a bus completed or not the assigned route.

12. This is a simplified version of the payment formula.

13. In Santiago, a centralized financial authority in charge of the electronic pre-payment card system receives all revenue. Users cannot pay with cash aboard buses. Therefore, no issues arise regarding the monitoring of fare box revenues on buses in order to avoid overcompensating operators with this revenue-sharing scheme.

14. This is a simplified version of the formula in the bidding documents.

15. We have no information on how the weights were originally determined. For a scientific approach to obtaining such weights, see Hensher and Prioni (2002).

16. This is not the exact notation contained in the contract.

17. This parameter is calculated each month as Av_k = 0.00344·Vel²− 0.19827·Vel + 3.76221, where Vel is the average speed of the fleet obtained by dividing for each firm the number of commercial kilometres operated by the number of hours that the fleet was in operation. It decreases until it reaches a minimum of 0.90 at 28.8 kilometres per hour. This adjustment is made in order to take into account higher costs as service speeds decrease in the city.

18. Operators do face some idiosyncratic demand risk if the operational plan (that determines the commercial kilometres and dispatch for trunk services) is adjusted for permanent changes in demand. However, operational costs also change in this case. It must be noted that this payment mechanism is essentially a benchmarking formula. Operators compete for a share of the income base on the weighted (by the velocity and fare variables) relative kilometres offered by each one and indirectly by patronage if the operational plan is related to demand. This is an interesting example of benchmarking or yardstick competition applied to the transit industry but based on demand and not on performance indicators (see also note 9).

19. The last term of the above formula (ARTZ_i) is an adjustment for non-trunk operations and only applies if the concessionaire is in charge of operating zonal terminals.

20. Just as in Bogotá, there is an indirect link between payments and demand since the authority determines the commercial kilometres in the operational plan for each service.