Comments

Ernesto Schargrodsky:

In the 1990s, many Latin American countries transferred the operation of several economic activities from public to private provision. There is currently no agreement on the welfare evaluation of this privatization process. Public opinion polls and press articles suggest widespread discontent with privatization, whereas academic research shows improved results under private operation relative to public performance.¹ This divergence of opinions may reflect the fact that the benign impact of privatization was below original expectations.

Engel, Fischer, and Galetovic look at the Latin American experience of highway privatization. The provided evidence suggests that the promised benefits of privatization failed to materialize. In this sector, contract renegotiations and opportunism were pervasive, the diversion of public funds to bail out franchise holders was frequent, and consumers sometimes ended up worse off after privatization. The authors look at one of the sectors in which privatization results were most disappointing and ask themselves, in a thoughtful and unbiased analysis, how to improve franchising design and when highway franchising is superior to public operation.

Two main contributions for the improvement of highway provision stand out from their work. The first regards the appropriate franchise term structure. Most concessions were awarded using fixed-term contracts, which make franchise holders bear the demand risk and thus creates pressure for subsidies and guarantees during bad states. The authors suggest the use of present-value-of-revenue (PVR) auctions, in which all demand risk is borne by the state (which ultimately bears it in practice after inefficient renegotiations). Under these contracts, franchises are automatically extended in low-demand states, so franchisers do not need to ask for bailouts. PVR auctions offer the additional advantage of reducing contractual incompleteness. The contract indicates the noncontingent amount that the state should pay when it prefers to buy back the project.²

The second contribution lies in the analysis of the cost-of-funds argument and the desirability of highway privatization. As they explain, a typical argument in favor of highway franchising is that private firms have access to funds from toll revenue at lower social costs than public sector funds obtained from distortionary taxation. They show the fallacy of this argument: the government can also use the highways for fund collection, reducing distortionary taxes elsewhere. The desirability of highway franchising then depends on the relationship between inefficiencies in subsidization and tax distortions. Although franchises with government subsidies and guarantees are very frequent in practice, the authors show that when government subsidies are required (on grounds other than externalities), full public highway provision should be preferred over privatization.

In addition to these insights, the article opens several roads for empirical research. First, a remaining question is whether highway franchising has improved or diminished social welfare. The evidence provided in this paper focuses the discussion about the success or failure of franchising on supply variables, such as contract characteristics, investment, and frequency of renegotiations. The analysis could usefully be furthered by a study of the impact of privatization on variables measuring users' welfare, such as transportation costs, motor-vehicle insurance costs, or road accidents. The impact on taxpayers could also be quantified. Of course, the relevant counterfactual should be public provision, not a comparison with the first best.

Second, it would be interesting to analyze the main determinants of the likelihood of renegotiation. The authors argue that the two flaws in regulatory design that induced opportunistic renegotiations were the use of fixed-term contracts and the lack of a proper ex ante regulatory structure. The effect of these contractual characteristics could be confirmed through empirical study on a sample of highway franchises.

Finally, it would be productive to examine why privatization has been particularly unsuccessful in this sector.³ The features described in this article are not exclusive of highway franchising. Other privatized activities also suffered from high demand risk, were transferred to private operation under the "privatize now, regulate later" approach, or used fixed terms in concession contracts. Future cross-industry comparisons could shed light on the determinants of the problematic performance of the highway sector.

Juan-Pablo Montero:

After reviewing the highway privatization experience of three Latin...