From fixed to state‐dependent duration in public‐private partnerships

• Author: Annalisa Vinella,Daniel Danau
• Published date: 01 September 2017
• Date: 01 September 2017
• DOI: http://doi.org/10.1111/jems.12204

Received: 7 March 2015 Revised: 29 November2016 Accepted: 15 December 2016

DOI: 10.1111/jems.12204

ORIGINAL ARTICLE

From fixed to state-dependent duration in public-private

partnerships

Daniel Danau1Annalisa Vinella2

1Universitéde Caen Normandie -

Centre de Rechercheen Economie et

Management, Esplanade de la Paix, Caen,

France (Email: daniel.danau@unicaen.fr)

2Universitàdegli Studi di Bari “Aldo Moro”—

Dipartimento di Scienze economiche e metodi

matematici, Largo AbbaziaS. Scolastica, Bar i,

Italy (Email: annalisa.vinella@uniba.it)

Abstract

A government delegates a build-operate-transfer project to a private firm. In the con-

tracting stage, the operating cost is unknown. The firm can increase the likelihood

of facing a low cost, rather than a high cost, by exerting costly effort when building

the infrastructure. Once the infrastructure is in place, the firm learns the tr ue cost and

begins to operate. Under limited commitment, either partner may renege on the con-

tract at any moment thereafter. The noveltywit h respect to incentivetheor y is that the

contractual length is stipulated in the contract in such a way that it depends on the cost

realization. Our main result is that, if the break-up of the partnership is sufficiently

costly to the government and/or adverse selection and moral hazard are sufficiently

severe, then the efficient contract is not robust to renegotiation unless it has a longer

duration when the realized cost is low. This result is at odds with the literature on

flexible-term contracts, which recommends a longer duration when operating con-

ditions are unfavorable, yet, with regard to a different setting, where the demand is

uncertain and the cash-flow is exogenous.

1INTRODUCTION

Public–private partnerships (PPPs) in infrastructure projects include two main phases, namely construction and operation, and it

is well known that incentive problems affect their performance in either phase. When building the infrastructure, the private firm

may be little motivated to exert costlyeffor t (Bennett& Iossa, 2006; Har t, 2003; Iossa & Martimort, 2015; Martimort & Pouyet,

2008). In the operation phase, the firm is likely to observe the operating conditions privately, as agency theory suggests, and to

conceal them vis à vis the government (e.g., Danau & Vinella, 2015; Guasch, Laffont, & Straub, 2006, 2008; Iossa & Martimort,

2015; Laffont, 2003). Moreover, both the firm and the government may have an interest in abjuring the PPP contract (see the

report of Guasch, 2004, and the cases described by Estache & Wren-Lewis, 2009). As a choice variable of the contract designer,

the contractual length represents a powerful tool to address these incentive problems (Danau & Vinella, 2015). This is because

variations in the contractual length permit the contract designer some flexibility in adjusting the per-period compensation, which

can be exploited to solve incentive problemsar ising during operation, without affecting the total compensation, which is instead

used to address moral hazard arising during construction. The degree of flexibility available depends on the severity of moral

hazard. The theory of incentives tells us that the more severe moral hazard is, the more uncertain the total compensation should

We are indebted to an associate editor and two anonymous reviewers for their useful suggestions. We are grateful to Alexander Galetovicfor fruitful discus-

sions. We also acknowledgehelpful comments from Stéphane Saussier, participants in the 2015 Workshop on Economics and Management of Public-Private

Partnerships (Venice) and seminar participants at Nanzan University (Nagoya)and Nihon University (Tokyo). A previous version of the paper was presented

at the 2015 EEA Meeting (Mannheim), the 2014 EARIE Conference (Milan), and the 2014 IO Workshopon Theor y,Empirics and Experiments (Alberobello).

We thank those audiences for their various suggestions. All remaining errors are our own.

J Econ Manage Strat. 2017;26:636–660. © 2017 WileyPeriodicals, Inc. 636wileyonlinelibrary.com/journal/jems

DANAUAND VINELLA 637

be. However, as the firm is exposed to more risk, there is less flexibility in adjusting the per-period compensation through

changes in the contractual length. Hence, it becomes more difficult to incentivize the firm during operation. The choice of a

suitable contractual length is thus related to how important each of the incentive problems is.

Despite this essential link between the duration of PPP contracts and the partners’ incentives during construction and opera-

tion, the choice of the optimal contractual length in PPP projects remains underexplored. Particularly, to the best of our knowl-

edge, the literature on agency relationships has not yet considered the possibility of conditioning the duration of the contract

on the state of nature as a tool to solve incentive problems. The idea of a state-dependent duration is inspired by the studies of

Engel, Fischer, & Galetovic (1997, 2001) although, in their case, the duration is not stipulated in the contract, but rather deter-

mined through the mechanics of the contract. The authors focus on frameworks in which fixed-term contracts are incomplete for

an exogenous reason (the market demand is uncertain) and show that incompleteness is eliminated if the contractual length is

adjusted according to the realized state of nature (the level of demand) in such a way that the firm attains its reservation utility

regardless of the specific state. This requires allowing the firm to operate the activity for a larger number of periods when the

operating conditions are unfavorable. Contracts with this characteristic are referred to as flexible-term contracts. However, one

may wonder whether the benefits of flexible-term contracts also extend to the frameworks we have in mind, in which moral

hazard and adverse selection have bite and contractual frictions are due to the lack of enforcement mechanisms (as in the cases

described by Estache & Wren-Lewis, 2009) rather than to the parties’ inability of writing a complete contract for all future

contingencies.1As we have noted, both moral hazard and adverse selection require the firm to be exposed to some risk. This

involves providing lower compensation to the firm in bad states of nature. Moreover, if a contract with a long duration is used

in bad states of nature, as is the case for flexible-term contracts, then, as time passes, the firm might prefer to cease honoring

its obligations, provided that its residual compensation falls below some alternative opportunity, which could be derived from

another activity or from a new deal with the same partner. One then needs to understand whether a contract with a state-dependent

duration would be useful in the environments we consider and, if so, howexactly it should be structured. We explore these issues

in our paper.

The analysis we develop delivers one main lesson. In situations in which the firm enjoys an informational advantage early in

its relationship with the government and, in addition, either partner may behave opportunistically during the operation phase,

the contract that stipulates an efficient allocation may fail to be renegotiation proof, unless its duration is conditioned on the state

of nature. When that is the case, unlike in flexible-term contracts, the duration should be set longer in favorable states than in

unfavorable states. A contract with this characteristic is more likely to be necessary the more severe moral hazard and adverse

selection are and/or the more costly the break-up of the partnership would be to the government.2

As an additional contribution, our analysis offers a foundation to the understanding of when and how often each of the

partners may want to initiate renegotiation during the execution of the contract, which is essential to draw conclusions on

the optimal contractual design in PPPs. We ascertain that, should the firm initiate renegotiation in some period of operation,

it will take the initiative again in each of the subsequent periods. In so doing, the firm will collect the annuity of the gov-

ernment’s cost saving from the continuation of the partnership, leveraging itself on the damage that it can occasion to the

government by reneging on the contract, rather than on the characteristics of the PPP project. On the other hand, should the gov-

ernment initiate renegotiation, it will do so at most once, as renegotiation is onerous to it. The government will take the initiative

early in the operation phase to appropriate more of the up-front private investment under the new deal that replaces the initial

contract.

This study is related to Danau and Vinella (2015), which serves as the basis for the model. However, although that paper

explores the financial structure of the project, here, we assume that the firm is the only investor. This simplification does not

affect the general insights of our study, but it serves to focust he analysison the use of a state-dependent duration as an incentive

tool in PPP contracts. However,this simplification leads to a complication. In any period in which some par tner could breacht he

contract, renegotiation would be Pareto-improving on the termination of the partnership. Hence, following a contractual breach,

the parties would actually reach a new agreement and continue the relationship. The contract must thus be robust to the possibility

of repeated renegotiation. In Danau and Vinella (2015), the convenience of seeking new deals is excluded by the presence of a

financial institution that can impose high debt payments on the government in the event of a successful renegotiation, to destroy

any surplus to be shared in renegotiation.

Our framework differs fromt he literature on flexible-term contracts in twoessential respects, which pave the way for a different

result. First, the compensation to the firm is endogenous and used as a tool to fine-tune incentives. Second, it is necessary to take

explicit analytical consideration of the renegotiation game in which the partners engage following a contractual breach. This is

not the case in the framework of Engel et al. (1997, 2001), in which renegotiation can only be due to the parties’ inability of

addressing all possible future contingencies in the contract, an issue that is eliminated by making the firm’s payoff independent

of the realized operating conditions.