IV. CRITICISMS OF SEMI-PRIVATIZATION SCHEMES AND COSTANERA NORTE
Direct government intervention may not be the best means of providing infrastructure. Semi-privatization schemes could be a more efficient alternative. In highway construction, technological advances have now made privatization quite feasible, so that market mechanisms like congestion pricing can yield substantial social benefits. While Costanera Norte certainly follows the worldwide trend toward privatization, it does not represent plenary privatization. Semi-privatization schemes such as Costanera Norte only differ from direct government provision in that it is to be financed and built by the private sector, and in that, like all infrastructure concessions, it expands the feasible set of fundable projects (Cobin, 1999a).
Correspondingly, given that privatization is a desirable policy, questions still remain regarding the optimal extent of privatization, along with what political and legal framework provides the optimal basis for it. Specifically, it is not clear that privatizing only part of the infrastructure provision process best solves the problems associated with direct government provision. Moreover, it is not clear that semi-privatization is the most efficient and effective means to provide highways in "the public interest".
Accordingly, Costanera Norte is not without its detractors. Businessmen and concerned citizens post signs of protest along the proposed route. Interviews (23) revealed that nearly all potential concessionaires and the mayors of two wealthy municipalities opposed the project. In examining their concerns, as well as the theoretical work of others, it is possible to identify ten general problems of semi-privatization concession projects: (1) technical flaws that slow production and augment costs, (2) doubtful cost estimates produced by regulators, (3) legal difficulties and political uncertainties, (4) completion time and end-of-term maintenance problems, (5) environmentalist clashes in the political process, (6) social losses due to the strategic use of concession policy by vote-seeking politicians, (7) rent seeking problems, (8) problems with financial guarantees in the bidding process, (9) social losses from knowledge problems in planning and forecasting, and (10) efficiency problems.
Technical Flaws That Slow Production And Augment Costs
Building a highway is not tremendously complicated, but it does involve many planning criteria, including design, soil management, paving methods (with either longer lasting concrete or quieter asphalt -- which will be used in at least part of Costanera Norte), tolling, lighting, drainage, various other technologies, and maintenance (Humphrey, 1996). Providencia mayor Cristián Labbé thinks that Costanera Norte has technical flaws, not the least of which is the problem of parking -- an issue which has also been raised by others (Serra 1996, p.231). Accordingly, many potential concessionaires are too apprehensive about Costanera Norte's specifications to consider participating. Vincente Dominguez Vial, leader of the Chilean-Spanish Concessiones Chilena Limitada, thinks that while the problem of integrating Avenida Kennedy is small, problems with the specification of the inner-city tunnels and trenches required, and especially their ventilation, pose considerable difficulties.
Carl Weber, General Manager of Obras y Desarrollo, also argues that Costanera Norte has been poorly designed. For instance, little consideration has been given to the potential damage to buildings when tunnels and trenches are installed downtown. In addition, he claims that the technology needed to scan and charge tolls at high speeds (100 km/hr.), as the project specifications demand, simply does not exist, debilitating the project and elevating risks for concessionaires. However, many would disagree with Weber's technology assessment (cf. D'Souza 1995; Malone 1990; and Rhode, 1992). Claudio Garín Carrasco, Manager of Urban Concessions for the Ministry of Public Works, is sure that the technology exists, noting Hughes products in use in Los Angeles and Toronto (which can handle speeds up to 150 km/hr.).
Doubtful Cost Estimates Produced By Regulators
People who have analyzed Costanera Norte believe that Chilean planners erred in their cost projections. Labbé argues that the project is not the most cost-effective and environmentally sound alternative, preferring projects such as ring roads around Santiago. Accordingly, Dominguez would prefer instead to bid for the Americo Vespucio ring road concession. Weber, like Dominguez, believes that ring road project should have been the first one started, along with the planned improvements to the North-South route. Las Condes mayor Joaquín Lavín Infante says the bidding and construction process is slow, costly, and plagued with a number of legal difficulties. Maximiliano Ruiz and Fernando Vergara of Infrastructure 2000 believe the plan's design is not very economical and the estimated costs used in the studies are inexact (i.e., they are likely underestimated). Dominguez is very pessimistic about the project, noting his belief that tolls would have to be increased by as much as 60% to cover the costs of construction and a decent profit margin of perhaps 15% (which would in turn reduce quantity demanded).
Legal Difficulties And Political Uncertainties
One of the biggest concerns is the project's uncertainties. Labbé notes that Costanera Norte is and will be embroiled in legal difficulties, mainly resulting from usurping the real property in a recalcitrant sector of Providencia known as Pedro de Valdivia Norte. Residents fear that the project will lower their property values and increase pollution. Accordingly, Dominguez cites many logistical difficulties, not the least of which arise from the complaints of these residents and the condemnations in the Recoleta area. Chilean legislation secures private real property from condemnation without compensation and mandates strict criteria for the taking of property when it does occur.(24) Ruiz and Vergara contend that the rules and regulations for the "game" are not clear, leaving far too many questions about the project unanswered. Not the least of their concerns are the political risks associated with the project, also noting problems with condemnations, delays associated with bidding prequalification and selection, and how toll violators will be fined.
Weber feared that there is a risk of state-led competition, since governments could operate an alternative to Costanera Norte that would change demand and thus lower projected revenues (which is in fact what would happen later with the advent of Costanera Sur). Serra notes, "In highway concessions, traffic, and therefore the franchise-holder's income, can be strongly influenced by government decisions regarding the rest of the road network" (Serra 1996, p.214). Such state-led competition was seen in Chile previously when a state by-pass road was constructed over private tunnel Melón, a couple hours north of Santiago.(25)
Similarly, administration and demand forecasting in Mexican highway projects have been troublesome, including the Cuernavaca-Acapulco and Guadalajara-Tepic routes (Emmons and Brand 1993, p.11). Moreover, Michael Meacher argues that overestimates of traffic growth adversely affect revenue forecasts and could potentially leave the government responsible for any shortfall, meaning that many of the associated risks with concession schemes are ultimately still born by the state (Meacher, 1995). Such was the case in the recent Mexico City-Acapulco highway debacle (Sheridan, 1996).
Completion Time And End-Of-Term Maintenance Problems
Concessionaires always have an incentive to complete their projects quickly in order to receive revenues faster, but this benefit might be offset by costs such as the risk of losing the bidding war to begin with (presuming that completing the project faster will be more costly). The low bid might often be the one which utilizes a longer time to completion, and thus might not best serve highway consumers who prefer present over future consumption. Alternatively, under the Mexican concession system, bids are selected according the following priorities: "(1) shortest construction period; (2) total cost; (3) reliability of the proposed financing; and (4) experience and qualifications of the bidder" (Emmons and Brand 1993, p.8). It appears that the Mexican system eliminates the completion problem while the Chilean system does not.
Additionally, in any highway concession, problems with road maintenance are to be expected during the final years of the concession, when concessionaires have little incentive to make costly repairs. This fact suggests that semi-privatization schemes will tend to produce socially inefficient outcomes during part of the concession period because of institutional incentives.
Environmentalist Clashes In The Political Process
Toll highway concessions have been widely criticized as being impractical, if not deleterious by environmentalist groups. For example, a toll road between Canada and the USA has been denoted as "an environmental, health and safety nightmare" (Ghent, 1996). Some claim that antidotes like sound walls do not work, implying that they are a ruse to get people to believe that they help environmental quality (Smith, 1993). In many cases, radical environmental groups have been successful at stopping toll road expansion in Europe (Boyle, 1995), just as they have hampered construction efforts in Mexico (Emmons and Brand 1993, p.11). Environmental groups in Chile caused a number of costly roadblocks for Costanera Norte. Ruiz and Vergara of Infrastructure 2000 were very concerned about the unknown costs these groups' activities would impose. In 1997, environmental groups caused delays in Costanera Norte by demanding that more studies be conducted, in spite of the then current favorable impact studies.(26) Continuing throughout 1998, the project proved to be a hot issue in Santiago because of environmentalist concerns and other political problems.
Garín said that the Ministry of Public Works took a more pragmatic stance on environmental issues. If there is going to be more infrastructure then there are going to be added social costs. The goal is not to reduce these costs to zero but to minimize them (without spending a fortune) to provide more transportation infrastructure. Chile might have an advantage in employing environmental technologies that help minimize costs, since, as some argue, the government retains many standards-setting functions and operates them well (Bretl, Deluca, and Fredell, 1996). Indeed, scholars have offered practical solutions to smog and congestion problems through combinations of market and governmental regulatory policies (McDonald and Mohammadioun, 1994). But these pragmatic oblations and abstract postulations did little to appease environmentalists in the case of Costanera Norte.
Strategic Uses By Vote-Seeking Politicians
Gómez-Ibañez (1998, pp.106-107) argues that vote seeking leads to the production of over capacity in highways at subsidized prices and thus prevents socially efficient solutions. Likewise, Mohring (1998, p.198) points out that congestion pricing (which is socially beneficial) is "a very hard sell" politically and thus has rarely been used in practice. Thus, vote seeking leads to sub optimal outcomes.
Concession projects are not free of the burdens of the political process and policy debate. Mark Crain and Lisa Oakley argue that there is a "strategic use of infrastructure" which is used to affect future public choices (Crain and Oakley 1995, pp.3, 15). Indeed, the theory of public choice suggests that, under a democracy, it will be nearly impossible to eliminate such tendencies. Correspondingly, Lavín claims there are political interests backing Costanera Norte. In particular, Ricardo Lagos, the head of the Ministry of Public Works when Costanera Norte was announced and the leading presidential candidate in 1999, might have a private and personal interest of wanting to garner wider public support by completing highly visible infrastructure projects. Lavín is running against Lagos, and his bid might be harmed by opposition to the project, although opposition does help him retain local support. Therefore, given the political issues surrounding the project, Weber rightly suggests that aspects of public choice theory should be applied in analyzing it.
Public choice theory suggests that vote seeking politicians will always seek to optimize the number of votes they receive by promoting apparently beneficial policies (such as infrastructure) vigorously while avoiding detailed discussions of costs or taxes which might offend voters. Conformably, the objectives of the Chilean ministries (noted earlier) for changes in urban policy are ambiguous and nebulous, lauding vague benefits without detailing costs. It is also unclear that Chilean planners' desire to subordinate private interests in favor of public ones, as they have suggested, is either realistic or expedient. This sort of rhetoric might sound good to large blocks of voters, but in reality it probably has little practical significance. Public choice theory suggests that people pursue self-interest goals. Private interests dominate, whether in the market process or the political process, rhetoric to the contrary notwithstanding. In other words, even political actors are going to pursue their personal and private interests primarily, and only pursue public interest goals secondarily. Thus, in analyzing the Costanera Norte case, semi-privatization schemes can not solve public choice problems entirely; they can only hope to reduce them somewhat.
Rent Seeking Problems
Because government is still involved in semi-privatization schemes, rent seeking is not eliminated entirely. Some rent seekers simply shift to a new arena, focusing on the political implementation and regulation of toll highway concessions. Without toll highways, the booty available for rent seekers is constrained by the government budget (especially in Chile where deficit spending is prohibited). Concessions can augment the potential rent seeking pie because government still heads the process, even though financing comes from the private sector rather than taxes. Accordingly, significant rents appear to be available. Taking into account what may be observed: (1) there have been attractive, high-priced projects proposed or completed in the United States, Mexico, Chile, and elsewhere, (2) there has been considerable political interference by special interest groups, and (3) there have been considerable "paperwork contests" (Higgins and Tollison 1988, pp.150-151) and other scholarly work on infrastructure topics, indicating both academic and practical value. Scholars have provided conflicting views over the usefulness of guarantees, environmental issues, as well as other conflictive aspects of highway concessions under semi-privatization policy, but at any rate the subject seems to be good business for all side within and without the academic market. Since all action is costly, the levels of activity surrounding infrastructure development indicate the existence of considerable rents. Accordingly, rent seeking will tend to retard efforts to improve social efficiency by means of semi-privatized concession schemes.
On the surface, there are no obvious cases of crass rent seeking (outside the circle of bidders perhaps) with respect to Costanera Norte. But the denouement of the plan, and its political development, tend to make onlookers suspicious that considerable rent seeking is occurring. Correspondingly, Estrategia, one of Chile's two leading financial papers, reported that of the many firms competing for past concessions, three have won 82% of the contracts, exemplifying a trend toward concentration among franchises.(27) Such concentration or monopolization would seem antithetical to promoting the public interest.(28) Of the fifteen firms that were pre-qualified to bid on Costanera Norte, only a few were considering doing so (based on inquiries in the southern spring 1997). At that time, only two firms were committed to participate, and at least four firms would not participate. At the close of 1998, only a French firm, Transroute, entered a bid. This lack of participation indicates that Costanera Norte is not widely recognized as a desirable project in the market.
Complexities and rigidities can be useful to rent seekers looking to carve a niche for themselves, but they also tend to create inefficiencies. For instance, legislation mandating a certain technology must be utilized, which is currently provided by only one or even a few firms, would tend to concentrate benefits narrowly while dispersing social costs over the mass of ill-informed voters.
Problems With The Bidding Process And Project Development
Likewise, the political process in Chile has complicated the concession bidding process and hampered efficiency. Chilean policy grants relative benefits to larger firms who have large fixed investments in organizational and institutional costs that facilitate complex bidding. It is very costly to prepare a good bid, which could cost as much as $300,000 or more. Unless a firm is satisfied that the terms of the project will generate a reasonable net rate of return (in the range of 9% to 14%), including reasonable estimates for various risks, it will not be willing to spend scarce resources to prepare a bid. And so it was with Costanera Norte.
The bidding process for Chilean concessions has not passed without scholarly criticism. Eduardo Engel, Ronald Fischer, and Alexander Galetovic argue that information asymmetries and inherent risks associated with highway construction will lead the current bidding system to falter. They suggest that a scheme which permits flexibility in the length of the concession and in toll rates (both of which are presently fixed by the regulator) would reduce bankruptcy problems, costly renegotiations or contract modifications, risks of forecasting error, and maintenance problems (Engel, Fischer, and Galetovic 1996, pp.6, 20-22). They offer a more simple system to reduce risks and facilitate smooth bidding based on a formula to find the net present value of cash flows. Moreover, they point out defects in the Costanera Norte plan: including toll fixing, inadequate regulation, and the difficulties for the Ministry to estimate construction costs in order to provide financial guarantees (Engel, Fischer, and Galetovic 1997a, pp.182-186, 207-209).
Jean Tirole applauds their work, but notes that the proposal does not provide a reasonable discount rate, method to estimate tolls, and does not improve the problem of forecasting highway use or demand. He also says there will be regulatory and practical difficulties, including poor incentives to maintain highway quality and support commercial services along the route, suggesting that regulators should use financial awards to improve incentives and set minimum quality standards (Tirole 1997, pp. 201, 210-214). Also citing problems in forecasting demand, Michael Klein of the World Bank is more critical of the proposal, which would weaken regulatory control, arguing that concessions should be closely regulated like natural monopolies to provide a fair rate of return and equitable distribution of risks (Klein 1997, pp.220-223). In reply, Engel, Fischer, and Galetovic concede most of Tirole's points, suggesting that giving regulators a range of possible tolls would be helpful, especially if coupled with a congestion pricing scheme. They are more critical of Klein, denying that highway concessions will be mortally susceptible to interest group pressures or that low demand will always make projects unattractive to the private sector. The key is flexibility (Engel, Fischer, and Galetovic 1997b, pp.218-221). The rigidities that exist may imply systemic, accidental inefficiencies and/or intentional regulatory capture designed to benefit rent seekers and favor brokers.
Ives Montélimard, leader of Consorcio Internacional Autopistas Chilenas (Transroute), is in favor of Costanera Norte, and virtually stood alone supporting it when it finally failed. However, he foresaw problems that would arise because the Ministry of Finance was not going to provide the victorious firm with sufficient financial guarantees. Many scholars view state financial and contractual guarantees are crucial to the success of any infrastructure concession project (Lange and Howson, 1996), although some do not agree (see Gutiérrez1996, pp.146-147). These guarantees lower the risk of future revenues to the franchise operator and make projects more feasible (Serra 1996, p.229). Accordingly, firms bidding on the Costanera Norte concession have been very concerned about the guarantees, as well as assurances of consistent traffic flows and how users will be charged. In August 1998, regulators thought that they had found a satisfactory means to assuage the firms.(29) However, at the end of the year, seven of the eight remaining firms dropped out of the bidding competition, leaving only Transroute to submit a lame bid. (Transroute did not meet all the bid specifications, committing less than half the funds required to make its bid valid.) This poor result was a "hard blow" to the Ministry of Public Works, embroiled in the prolonged and costly debacle, which in early January 1999 consoled itself in knowing that at least the project "will not die".(30) However, the bidding process for the emblematic initiative was cancelled on February 15th, leaving uncertainties about when or if it would be resuscitated.(31)
The Ministry of Public Works announced in January 1999 that it was considering revamping the Costanera Norte package by combining it with the Americo Vespucio ring road and the North-South system projects into one "mega-project", in hope of attracting more international investors. Montélimard felt that whether his firm got the project simply became a political decision, adding that his firm would prefer to do all three projects. He considered the selection process to have been entirely unjust insofar as his firm was affected. If the Ministry had re-bundled the project, it would have reopened all the concerns about Costanera Norte. It would have also meant that concessionaires had wasted three years of preparation and considerable expense because of this government failure. Rather than recognize their failure and scrap the project, the government would in effect force it down the market's throat by making the concessionaires interested in the other projects also produce Costanera Norte.(32) In the end, another path to "success" was chosen.
The plan was resurrected on May 3, 1999 with a new focus. Rather than force the market to accept the Costanera Norte proposal by packaging it with the other two urban concessions, the Ministry of Public works set new dates for the concessions, and radically altered the Costanera Norte plan (which will be first in line for the new bidding), effectively transforming it into direct provision. The Chilean government will now assume far more risk and the required investment will be lowered, with a maximum cost set at US$300 million in order to assure bidders what their costs will be. This will provide greater firm profitability, especially because the state will also guarantee a minimum amount of traffic flow. In addition, special legislation has been submitted to provide for strict policing and fining of toll violators on Costanera Norte. In effect, these guarantees amount to subsidies and removal of most market risks such that there is little resemblance to true market provision. In this case, semi-privatization has been transformed into essentially direct provision. In the same way that governments regularly borrow from their citizens to finance projects, and often use private contractors to do the work, Chilean planners will simply use the bidding system to select lenders and contractors to complete the ailing project. Besides political motivations in an election year, the government hopes to boost lagging employment in the construction sector by forcing the plan into existence.(33)
Correspondingly, it is of considerable interest to note the announcement on April 21, 1999 that some municipal governments will be starting a new project called Costanera Sur, which will directly compete with a large segment of Costanera Norte. The new public highway will be built on the south side of the Mapocho River, across from Costanera Norte, and will thus dramatically change the demand characteristics of the Costanera Norte proposal, almost necessitating moving it back into the realm of public provision.(34)
Social Losses From Knowledge Problems In Planning And Forecasting
A number of costly knowledge problems due to government planning failures have already been noted. One relevant concern is that it may be possible for a firm to profit from a highway concession and still leave the state holding the bag for potential risks or other costs, as was the case with the Mexico City-Acapulco highway recently. The private operator might abandon (or threaten to abandon) the infrastructure at some point and require the state to step in to maintain it, creating inefficiency and new taxes, and further beleaguering projects with government failure on account of the knowledge problem (cf. Holcombe 1995, pp.15-18). Such a strategy would be rational and could be clandestinely planned by a profit seeking firm, making it difficult for planners to foresee. There may be other rent seeking costs too that highway concession policy cannot avoid. For instance, the propensity to concentrate projects in the hands of a few larger firms over time might be unknowable to planners, and could also indicate rent seeking activity or even regulatory capture.
Good ideas and good intentions do not always lead to good planning. Inadequately informed planners have failed to implement congestion pricing schemes well, notably in Singapore (Mohring 1998, 194) where congestion pricing on one road backfired and caused greater congestion on other roads. Gómez-Ibañez (1998, pp.107-108, 111) explains that failures, such as in Singapore, result from an insuperable knowledge problem. Congestion pricing policy is difficult to implement because planners must know customer demand and cross price elasticities in order to price roads correctly. Moreover, even if these elasticities were found, still more knowledge would be needed as any price change would immediately result in some people changing their quantity demanded (i.e., road use) and thus change expected road revenues. Planners would have to know much more than just elasticities in order to estimate the outcome of any policy. Furthermore, Gómez-Ibañez (p. 124) points out that planning knowledge is limited in other areas as well. Planners do not even have an objective means to allocate joint costs between trucks and autos on highways. Any such judgments are based on equity choices rather than efficiency. Consequently, planners can not hope to have the amount of social knowledge needed to implement efficient plans such as congestion pricing (Hayek, 1945).
Semi-privatization has also met with scholarly criticism on account of its inefficiencies. Werner Hirsch raises some doubts about the effectiveness of contracting-out for road production, noting flaws in the theory of semi-privatization pointing out the limited social efficiency they produce. He argues that the popular notion that contracting-out reduces costs is "fraught with serious dangers" because (1) reduced costs are not necessarily due to increased efficiency, (2) biased samples for cost data are often used, leading to pro-contracting results, and (3) cost efficiency should not be the only thing considered when deciding to contract-out services (Hirsch, 1995). It is possible that government failures have led to both poor technology choices and poorly defined property rights that augment highway concession problems.(35)
The cost per mile of Costanera Norte could be reasonable especially considering the tunnels and trenches that would have to be built. However, the fact that the project entails constructing a relatively narrow urban road, that the project cost for labor is likely lower in Chile than in the United States, and that the average land acquisition cost is relatively low for an urban project (much of the highway will be built on vacant state-owned land along the Mapocho River), Costanera Norte might actually be relatively expensive and socially inefficient. Also, some potential concessionaires anticipate that actual project costs will be significantly higher, exacerbating inefficiencies. Therefore, it is not clear that semi-privatization has wrought the most efficient cost of production for the Costanera Norte project. Costs might even supersede those that would be expected under direct government provision.
Costanera Norte does not seem to be a marketable plan, unless one takes the perspective of a rent seeker. Alternatively, there seems to be considerable interest in building ring roads around Santiago. This fact suggests that planners have not effectively chosen the best alternative, i.e., likely not the alternative that would have risen in the market. How can planners expect to make even close to optimal choices when faced with the knowledge problem?
Accordingly, Costanera Norte may have simply been a planning error. Labbé, who in 1997 doubted the project will ever be completed (and he was right, insofar as the pure concession version is concerned), would prefer investing in light rail or some of the more cost-effective highway projects proposed, in order to alleviate congestion with the least amount of environmental damage. Weber observed that just by replacing several traffic signals on the existing roadways with overpass-underpass systems, a similar goal could be achieved for the eastern portion of the project at far less cost. Consequently, it appears that Costanera Norte is an example of a sub-optimal choice based on inadequate planning information that was contorted by political process problems.