Regulation, Public Service Provision and Contracting 1 Stéphane Saussier Sorbonne Business School Saussier@univ-paris1.fr http://www.webssa.net Class 2 Incomplete Contracts and the Proper Scope of Government
Goal of Today s Session: Key Ideas 2 Main question: what is the proper scope of government? Recap: Without any transaction costs few insight on which institutional framework for public services Main consequence of transaction costs: contracts are incomplete Incomplete contract theory is not a translation of transaction cost theory but gives alternative answers
Make or buy? Transaction Cost Economics! Starts from the 70 s Williamson O.E. U. of California Berkeley 1932 - Starting point: let s take Ronald Coase seriously and identify the source of transaction costs - cost of using the market Antitrust issues at the beginning A theory of the firm, contracts and vertical integration A theory that applied to both private and public contracts First case study 1976 CATV case An empirical success story
Transaction Cost Economics Williamson, Oliver E, 1973. "Markets and Hierarchies: Some Elementary Considerations," American Economic Review, vol. 63(2), pages 316-25, May. Williamson O.E. U. of California Berkeley 1932-1975 1985
Transaction Cost Economics: basics A set of realistic behavioral assumptions Bounded rationality Opportunism Williamson O.E. U. of California Berkeley 1932 - Coupled with characteristics of the transactions Complexity and/or uncertainty Small number relationships / asset specificity Generating transaction costs
Transaction Cost Economics: basics Bounded Rationality Complexity / Uncertainty Incomplete contracting Williamson O.E. U. of California Berkeley 1932 - Possible Opportunism Small Numbers Relationship Need for Credible Commitments
What Are Transaction Costs? The Costs of contracting. Ink costs Ex ante Costs Negociation costs Search costs Contract costs Maladaptation costs Ex post Costs Renegociation costs Monitoring costs Breach costs
Asset specificity level is endogenous Small number relationships and asset specificity Uncertainty Lock-in effect and long-term relationships Organizational surplus Rooms for opportunism
"It's always been a danger, but it looms like a shadow over everything we've built here. But things have developed that will ensure security. I've just made a deal that will keep the Empire out of here forever." The Hold-Up Issue Lando: «But, this was not our initial agreement!» Darth Vader: «I changed the terms of the agreement. Consider yourselves happy that I did not change them more.»
One example! «But it is difficult to be clear and comprehensive in defining service targets. Governments tend to be nervous about providing only general performance obligations, fearing that the concessionaire will do less than they deem necessary. An example in the United Kingdom helps show why. In the competition for a build-operate transfer (BOT) contract for a prison, granted under the country s private finance initiative, it turned out that the winning company s bid was based on a plan to house several prisoners in each cell. The government had wanted single occupancy, but had forgotten to specify this in the tender documents.» M. Klein, World Bank, Bidding for Concessions The Impact of Contract Design, 1998 http://siteresources.worldbank.org/extfinancialsector/resources/ 282884-1303327122200/158klein.pdf
What Implications? Never choose to integrate only for production cost considerations: market always do better! Only transaction cost considerations should drive the decision If no credible commitment is possible then a more complex and costly governance structure is needed, namely vertical integration or regulation
Main proposition 12 Franchise-bidding solution works in very specific situations BUY MAKE In complex or unpredictable environments, transaction costs are likely to be minimized by removing exchange decisions from the market context and substituting the continuous oversight that regulatory mechanisms entail (Crocker-Masten 1996).
The cost of integration / regulation 13 Coming back to Coase 1937: Inside the firm, «If a workman moves from department Y to department X, he does not go because of a change in relative prices, but because he is ordered to do so» (Coase 37, The Nature of the Firm) Vertical integration is characterized by fiat and forbearance contract i.e. non negotiated authority. Cost is that inevitably, incentives fail i.e. bureaucratic costs Vertical integration is the solution of last resort Each governance structure is flawed Depending on transaction s characteristics some are less inefficient than others.
Franchise-bidding and transaction costs Choice of the Private Operator Contract renewal t Service specification Contract execution Competition for the field Competition for the field
Franchise-bidding s failures (1) The difficulties to put firms in «competition for the field» Price criteria does not always resume what is expected from the private operator Price criteria might be complex (price vector instead of one single price) Aggressive bids and then quality shading (see the Nasa example!) Winners curse Collusion: ex: case of urban transport in France (2005) penalty 12M Corruption: anti-corruption law in France in 1993 Few bidders
Potential (imperfect) solutions Bids formulated in terms of a constant revenue stream Least-Present-Value-of-Revenue Auctions and Highway Franchising, (with R. Fischer and A. Galetovic), J. of Political Economy,109 (5), October 2001, 993 1020. $ High Traffic Construction Discounted Revenues Expected Traffic A Low Traffic Maximum length T 1 T 0 Years
Franchise-bidding s failures (2) Difficulties to enforce (incomplete) contractual agreements Disconnection between price and costs over time Penalties are difficult to apply (See the CATV Case) Non-verifiable dimensions of the contract Opportunistic behaviors might arise Efforts to evade or renegotiate the contract (Guasch 2004) Underinvestment Lower level of quality than promised (but not enforceable) Absence of responsiveness to consumer s needs Connected to the demand risk, transferred or not Concession vs. PFI
PPP s failures (3) Long term contracts and refranchising are problematic Specific Investments Long term contracts Lack of bidding parity at franchise renewal («Fundamental transformation») Water sector and urban transport in France: 90% of renewed contracts with the same firm! The solution of franchise bidding is not a free lunch. A need for regulation? For direct public management? Less incentives and more bureaucratic costs!
What is driving vertical integration? Proposition 1 : The more the transaction needs specific assets to be realized, the more probably externalization strategy will be difficult. Proposition 2 : The more the transaction is uncertain, the more externalization will be difficult, especially if the transaction necessitates specific investments.
An Empirical Success Story? Source: Boerner & Macher 2008
An Empirical Success Story? 21 Imprecise propositions? Several kinds of asset specificity A whole range of proposition on other aspect of contracting such as Duration Optimal level of completeness Price provisions See Masten & Crocker 1996. Too many ingredients in the theory: what are the main driving forces? Need for formalization?
To know more 22 Levin J. & S. Tadelis Contracting for Government Services: Theory and Evidence from U.S. Cities Journal of Industrial Economics, 2010, 58(3): 507-541 An empirical analysis of privatization by U.S. cities + 1 000 US cities / + 60 local services Main empirical findings: Services for which it is harder to write and administer performance contracts are less likely to be privatized. Complexity Hold-up sensibility The relationship is greater for larger and more urban cities. One explanation is that these cities have the resources to use in-house provision and perhaps also a more readily available pool of external providers.
23 Levin J. & S. Tadelis Contracting for Government Services: Theory and Evidence from U.S. Cities Journal of Industrial Economics, 2010, 58(3):507-541
The incomplete contract theory O. Hart Harvard U. Pathbreaking article: Grossman S.J., and O.D Hart [1986], The Costs and Benefits of Ownership : a Theory of Vertical Integration, Journal of Political Economy, 94, 691-719. One announced goal: To propose a good formalization of the transaction cost theory One main strategy: The assumption of bounded rationality is not useful in order to build an organization theory neither the assumption of asymmetric information
Assumptions 25 A set of assumptions Rationality Symmetric information Risky environment Incomplete contracting Contractual incompleteness comes from asymmetric information or bounded rationality of third parties (not formalized) Some specific investments are observable, but not verifiable Mainly human asset specificity A theoretical framework that is criticized, mainly because of internal inconsistency: Tirole 1999 Econometrica and Journal of Legal Studies 2000.
Main conclusions 26 Contractual incompleteness opens room for hold-up behaviors. Agents anticipate this. This anticipation distort ex ante investments Property rights distribution permits to influence such distortion By changing outside options There is no way to reach the first best
However Vertical integration issues: Integration does not change the level of control or negotiation power per se Non contractible investments are still non contractible! Integration gives to the party that owns the assets a higher outside option. Incentives to invest are thus influenced by property rights «It would be too easy to construct a theory of the firm with the assumption that incentives are decreasing in the firm compared to the market» (Hart 1995) Who integrates is crucial: Party with the more crucial specific investments should integrate the other but : if investments made by parties are equally important then what is optimal is the non-integration situation
TCT vs. ICT: Some quotations "The need, plainly, was to develop formal models of incomplete contract. That got under way with the publication of the path-breaking paper by Sanford Grossman and Oliver Hart [1986]" (WILLIAMSON [1993], page 42). "It is noteworthy that Grossman and Hart (and related papers of this kind) work from transaction cost economics premises - albeit with terminological differences. Thus Grossman and Hart employ the terms noncontractibility and nonverifiability rather than bounded rationality. And they refer to "relationship-specific investments" rather than asset specificity. Unanticipated state realizations, and the need to adapt thereto, are what pose contractual strains in their model. So uncertainty makes an appearance" (WILLIAMSON [1990], page 16). Versus If some of the central and critical ideas in transaction-cost economics as formulated by Williamson have been captured, others, equally important, have been missed. In particular, mathematics-based theory still lacks the language needed to capture essential ideas of bounded rationality, which are central to Williamson's concepts of transaction costs and contractual forms David M. KREPS [1996], page 2. Williamson 1999, 2000, 2002, Whinston 2003
Hart-Shleifer-Vishny 1997 Quarterly Journal of Economics 29 The incomplete contract theory is now one of the leading theoretical framework to analyze make or buy issues Surprisingly, this is not yet the case when this question is applied to the public services cases First step : Hart-Shleifer-Vishny 97 - with an application to the prisons in US. Now this is a canonical paper on this issue
Proper Scope of Government 30 Stylized facts: Private solution (contracting out) would be efficient at reducing costs Public solution (public agency) would be efficient at providing quality If the politician can sign a complete contract with employees or contractors Same outcomes in each case Incentive viewpoint: motivating the contractor or the employee present the same problem even in the presence of moral hazard or adverse selection See Class 1 To understand the costs and benefits of contracting out we need to consider a situation where contracts are incomplete
The Framework 31 G: Politician M: manager of the prison G and M are able to write a long term contract specifying some aspects of the good or services to be provided and the price Long term agreement because of needed relationship-specific investments Basic contract, for a basic public provision at price P 0 - P 0 is the price paid to the private manager or the wage paid to the public manager for a basic public provision
The Framework 32 Many not anticipated contingencies can call for some modifications of the basic good Need for innovation e.g. increased security The modified good yields a benefit B to society Cost C to the manager to produce The manager can manipulate B and C through two types of innovation efforts (e, i) e : efforts devoted to cost innovation (reduction) i : efforts devoted to quality innovation
The Framework 33 With β(i) the quality increase net of costs from the quality innovation M s overall costs:
The Framework Timing of the game e and i are observable but non contractible G s benefits and M s costs are observable but not transferable Revenue and cost-sharing arrangement are infeasible
First best as a benchmark 35 Let s suppose that all investments are contractible A complete contract is then signed in order to maximize global surplus We know that e* and i* must be such that:
CASE 1 - A private manager runs the prison 36 The manager owns the assets He invests in i and e and he decides unilaterally to implement or not innovations that reduce costs He renegotiates in order to implement innovations that increase quality Why? Because otherwise, he is not paid for the increase of quality
CASE 2 - A civil servant runs the prison 37 λ Let s be the ease with which the civil servant might be replaced Renegotiation takes place over the fraction of both the cost and quality innovation that G cannot appropriate λ
38 Organizational choices and incentives to invest in e? c - b 0
39 Organizational choices and incentives to invest in i?
Propositions 40 Public ownership leads to underinvestment in i and e Private ownership leads to overinvestment in cost reduction efforts e
What is the optimal organizational choice? 41 Comparison of the generated surplus The private solution will be preferred when Deterioration of quality from cost reduction is small The opportunities for cost reduction is small The public solution is nevertheless efficient when: There exist few opportunities to innovate to increase quality (i.e. technology is mature) and Adverse effect of cost reductions on quality are high There is no ex post competition This is not in the model Prisoners cannot choose their prison
US Prisons: Facts 42 95 % of them are publicly run / 5% are privately run Contracts signed between the State and managers are incomplete Some dimensions are not contractible Use of violence Death, violence stats Capabilities of the guards Average wage compared to public prisons What the model would predict?
US Prisons: Facts 43 Remind: The public solution is nevertheless efficient when: There exist few opportunities to innovate to increase quality (i.e. technology is mature) and Adverse effect of cost reductions on quality are high What the model would predict? When adverse effects are low then private prisons should be chosen Citizens do not care about violence and escapes When adverse effects are high then public prisons should be chosen Citizens do care about violence and escapes Why do we observe at the same time public and private prisons? Prisons are specialized by type of prisoners Private prisons are mainly youth correctional facilities Maximum security prisons are publicly run
Other applications 44 Education (role of ex post competition) Health (Experience good) Defense Williamson, O.E., 1999. Public and private bureaucraties: A Transaction Cost Economics perspective. Journal of Law, Economics and Organization, 15(1), p.306 342. Adds Probity concerns in the analysis Some transactions needs a high level of probity because opportunistic behavior would destroy a large part of the surplus
To sum up 45 Because of contract incompleteness, property rights distribution influences ex ante investments Adverse effect of cost innovation on quality is essential to understand make or buy issues in the provision of public services But No theoretical framework that takes into account particularities of having a public contractor (Wait for Pablo Spiller Class 4) All those theories apply for make or buy generally