Peru Concessions: The Overall Approach




James W. Fox


November 2007



            This case study describes the environment and experience of Peru as it privatized and concessioned infrastructure, leading up to the time that USAID provided assistance for this purpose.  It studies the possible impacts of USAID involvement in this sector.  Between 2003 and 2007, USAID assisted the Peruvian government in concessioning three major infrastructure projects:  a new container port at Callao, the 964 km. Northern Highway, and the 867 km. Central Highway.  The three projects are shown in Chart 1. 


            Altogether, these three projects are expected to result in more than $1 billion in foreign investment in infrastructure in Peru.  As of November 2007, only the first two of these concessions had been awarded.  A July 2007 call for bids on the Central highway produced no bidders. Though eight bidders were pre-qualified, none chose to bid after the government made extensive changes to the terms of the concession in the months leading to the award.


Chart 1


            This case study covers the overall approach taken by USAID and general information about the concessioning process in Peru as it evolved over time.  It also draws on the three concession deals that USAID promoted to offer some general conclusions.  The individual case studies should be consulted for the detailed specifics of each transaction.



I.  Introduction


            In 2003, the USAID mission decided to promote concessioning of infrastructure in Peru as part of its overall development strategy.  At that time, USAID’s primary focus was on promotion of alternative crops in coca-growing regions of the country, for which about $40 million was expended annually.  The change in approach followed a judgment that working only in the coca-growing areas to provide alternatives would be less effective than addressing policy and institutional issues that interfered with the creation of productive and well-paying employment in licit activities.


            Since the mid-1990s, Peru has experienced more than a decade of sustained economic growth.  This growth followed a decade of economic stagnation in the aftermath of a severe financial crisis in the late 1980s.  The government elected in 1985 provided substantial fiscal stimulus to the economy at a time when international currency reserves were modest.  The resulting inflation led to a rapid loss in international reserves, inevitable price controls, loss of access to international capital markets, steeper inflation, and finally, economic collapse.  GDP fell by 8.7% in 1988, by another 11.7% in 1989 and a further 5.1% in 1990.  The economy shrunk by nearly one-quarter over this three-year period.  Because of continuing population growth, per capita incomes fell even further – by nearly 30%.  The recovery from this catastrophe was initially slow, but began to accelerate in the late 1990s.  But it was climbing out of a deep hole.  According to Peru’s national accounts, the country only regained its 1987 per capita GDP in 2007.


            The catastrophe devastated the Peruvian government’s budget.  The sharp fall in revenues left the government with the stark choice of firing large numbers of government employees, with all of the legal and political consequences that might follow, or gutting investment spending.  It chose the latter, and Peru suffered for almost two decades from a lack of national government investment in infrastructure. 


            Infrastructure in Peru – even absent these fiscal issues – has always been problematic.  Peru is a huge country.  Occupying nearly 500,000 square miles and ranking 20th in the world in physical size, Peru is far larger than any European country.  Peru’s population of only 27 million puts it among the low population density countries of the world – with a population density level half that of the world average.  The geographically dispersed population naturally complicates efforts to provide roads, electricity and wired telecommunications.  This general characterization is modified by two factors.  First, the infrastructure challenge is simplified by the fact that most of the country’s population lives in the Western third of the country, away from the steamy jungles of the Amazon basin.  Second, a large fraction of the population lives in the Andean mountains, a chain characterized by steep peaks and defiles that challenge the ingenuity of anyone wanting to build roads though it.  This second factor more than countervails the first.  In sum, infrastructure investment in Peru is expensive, and likely to offer lower pay-offs than infrastructure investment in most countries, due to the combination of rugged countryside and low population density.


            This combination of low payoff to infrastructure and slow recovery from the national economic crisis of the 1980s left the country with a relatively poor transportation infrastructure.  Consequently, producers in the Amazon Basin, a relatively empty (though fast-growing) part of the country, face very high costs when transporting products to the large urban markets along the coast.  In turn, only very high-value products – like coca – remain profitable after absorbing the high transportation cost burden. 



II.  Concessions in Peru


            Government ownership of all infrastructure has been the norm in most developing countries in the half-century since “development” became the joint enterprise of poor country governments and rich country donors alongside the multilateral agencies they created to help facilitate this cause.  Until the 1980s, only rarely was developing country infrastructure not publicly owned.  The exceptions tended to be private providers who had provided services long before governments began to assert the right to own all infrastructure.  The movement to government ownership was fostered by donor agencies, particularly the multilateral banks.  They found lending to governments for power stations, roads and telecommunications to be the lowest-risk method for investment.  The availability of long-term credit from the multilateral banks was surely an important factor in extending the length of time that government-owned and subsidized infrastructure was seen as an adequate solution to the needs of developing countries.


There are three main alternatives to government provision of infrastructure: 





            Table 1 shows the principal differences between the use of concession and privatization.  (Other types of pubic-private partnerships should be seen as including a mixture of both methods.)


            As the table indicates, the government maintains far greater rights in a concession than in a privatization, not the least of which is the fact that the government eventually regains control of the asset.  This indeed happened after the port of Callao was concessioned in 1874 for 60 years in exchange for the construction of a modern pier by a British company.  Once the concession eventually ended, the government of Peru reclaimed possession of the port.


Table 1



            The Fujimori government, elected in 1990, began to pick up the pieces of the shattered economy.  It initially opted for privatization of infrastructure as the means to generate needed investment in modernization and in productivity improvements.  This was very much in line with the “Washington Consensus” that was the conventional wisdom among Latin American governments at the time.  Government enterprises were inefficient, overstaffed, and – through operating deficits due in part to failure to charge tariffs that covered costs – a burden on the government budget.  (Again, this burden was sometimes hidden by donor funding of investments by such SOEs.  Donor funding allowed these entities to make investments to upgrade capacity despite their failure to generate internal funding for expansion.) 


            By 1994, the government privatized much of the electricity and telecommunications infrastructure.  However, the political backlash from privatization – higher tariffs, worries about excess profits to foreign firms – gradually made “privatization” a controversial word.  The legislative approvals necessary for additional selling off of public-sector entities became increasingly difficult.  Additionally, there appears to have been concern about how the government was spending the large payments it received from privatizations.


            As a result of the popular backlash, the second Fujimori government (1995-2000) gave greater emphasis to concessions as an alternative that would achieve the principal objectives of privatization – additional investment in the sector, productivity improvement, elimination of a drain on the public purse – without the stigma associated with the permanent sale of national assets to foreigners.  One road, the Arequipa-Matarani highway, had been concessioned during the first Fujimori government, but – beginning in 1998 – the government entered into concessions for the railroads, the country’s main airport, the port of Matarani, and part of another road (National Route No. 5).


            The Toledo government that assumed office in 2001 continued the policy of promoting private management of infrastructure, seeking to use both privatization and concessioning.  But privatization had to be abandoned entirely in the wake of the sale in 2002 of the electric power company that served Arequipa.  The sale of a regional electric company serving Arequipa, with the $167 million sale price going to the national government, led to a revolt by Arequipans – who saw the benefits accruing to the national government while the costs in higher energy prices and reduced employment in the power company fell on Arequipa.  The public outcry led to lawsuit claiming that the power company belonged to the people of Arequipa rather than the national government.  When the city prevailed in the courts, the government was forced to apologize and to abandon all subsequent efforts to privatize infrastructure.  That left concessions as the only available vehicle for promoting investment in infrastructure by encouraging private operation. 


            A 2006 World Bank study of Peru’s experience during this period suggested that Peru had a more efficient concessioning process than most other Latin American countries.  The World Bank study concluded that competition was quite effective in securing the best outcomes in electricity and telecommunications concessions, but that the results were less favorable in transportation, a sector that lacked such competition.  For the Jorge Chavez airport (Peru’s principal link to the rest of the world) the study concluded that the concessionaire simply bid more than was reasonable.


            The World Bank report also comments that transportation concessions have “seemed too open to renegotiation.”  This suggests that personal relationships between concessionaires and the Ministry of Transport and Communications (MOTC) may have been closer, and their interactions less transparent, than the concession model would warrant.


III. The Peruvian Government’s Concession Apparatus


          With the establishment of privatization, concessions and other PPP approaches through a 1991 law, the Peruvian government created a set of institutions to oversee and manage the process.  Two types of institution were involved: 




In principle, the process is set in motion by a directive from the government to the promoter to include a specific piece of infrastructure in its agenda for concessioning.


The Promoter.  For all of the privatizations and concessions that took place from the time of its creation in 2002 onward, the government’s investment promotion unit, ProInversión, has been the action agency.  During this time, ProInversión had general responsibility for fostering foreign investment in Peru.  The process leading to the creation of ProInversión was slow and tortuous.  In 1992, the Commission to Promote Private Investment (COPRI) was established to sell public companies.  CORPRI moved quickly, privatizing telecommunications by 1994, and selling off various electrical companies during 1994-98.  A sister entity, PROMCEPRI, was established to promote privatization of transportation infrastructure, but it did not begin effective operation until 1996, following additional legislation that facilitated such privatization. 


CORPRI was expected to disappear in 2000, with PROMCEPRI continuing to promote additional transport concessions.  Instead, CORPRI absorbed its younger relative in 1998.  It was later fused with other institutions to form ProInversión in 2002. 


Altogether, ProInversión and its predecessors have privatized or concessioned at least 69 firms, most of which were business enterprises, legacies of Peru’s last flirtation with socialism in the early 1970s.  These included an airline, a steel mill, several cement factories, fishing companies, milk companies, hotels, banks, and a stock exchange.  The business enterprises were all sold off, while infrastructure operations were partly sold and partly concessioned. 


ProInversión and its predecessors have gradually become more proficient at providing schedules, clear rules, and supervision of the concessioning process.  “Data rooms” and “white books” are now issued for each process. The former provides bidders access to statistics, maps, and related documents and contracts. These documents included information about the general characteristics of the process, definitions, terms for questions and access to information, requirements for proposal presentation, and criteria for the winner selection.  Publicly available “white books” maintain records of all the information on the concession and privatization award processes from announcement to contract signing.    They included reports from various authorities, economic and financial documents prepared by consulting firms, and all official letters.  Initially, white books were only published when each process was completed, because authorities were fearful that access to detailed information would generate excessive questioning and political repercussions.  Criticisms later induced ProInversión to publish on its website every modification of the rules and contracts, as well as comments received during each stage of the process.


In the specific area of infrastructure, Table 2 summarizes Peru’s privatization and concessioning experience prior to USAID’s involvement in 2003.  As indicated by the table, privatization of the telecom and electricity sectors occurred rapidly between 1994 and 1998.  Transport privatization or concessioning was slower, with most taking place between 1999 and 2003.  Nevertheless, the changes represented major steps.  The country’s railroads were put in private hands, along with the country’s main airport, two highways, and a minor port.


ProInversión was very successful in promoting concessions in Peru, and its newsletters and website show reasonable competence.  ProInversión was the recipient agency for the USAID assistance that promoted the concessions discussed in this paper. Concessionaires and others interviewed for this study give the organization high praise for its professionalism.  ProInversión’s usual approach is to appoint a project director for each proposed concession, with two or three other ProInversión professionals assisting.  Studies are undertaken, consultants are identified and contracted, and the various actions leading up to a public offer of the concession are prepared.  (The case studies of the two highway concessions and the port concession supported by USAID provide considerably more detail on these issues.)  Overall, it is clear that ProInversión has achieved much during a relatively short period.  .


Nevertheless, ProInversión suffers from several weaknesses as an entity that can provide assurances to foreign investors over the long term.  First, its board of directors is chaired by the prime minister, the seven other board members being other ministers with portfolios ranging from finance to agriculture.  Any investment promotion office whose board of directors changes completely every time there is a change in government inspires little confidence in potential investors that their welcome will outlast the current administration.  To the relief of USAID and others promoting concessions, the 2006 change in government had no effect on ProInversión’s role in promoting foreign investment or the country’s commitment to welcoming it.  The Director of the institution continued in office, and government policies in this area were maintained.  Nevertheless, such policy continuity in changes of government in Latin America is more the exception than the rule, as a new administration will typically exaggerate the changes it will make from the “failed policies” of the previous government.


Table 2


Second, ProInversión’s website is unlikely to give lukewarm foreign investors a sense of confidence that it would be a reliable navigator of the numerous problems that inevitably arise when a foreign company makes its first investment in a new country.  Its website lacked an institutional history of the organization that might be useful in reassuring potential investors that a change in the national government would not completely change its mission.  Additional information, especially for English-speaking potential investors, would also have been useful.  ProInversión’s location inside the ugly edifice constructed by the national oil company, Petroperu, during the first oil boom, is a liability rather than an asset – regardless of whatever rent subsidy Petroperu gives to ProInversión.  The security procedures alone for any foreign investor with an appointment with a ProInversión official would be interpreted by many potential investors as a signal that Peru was not yet ready to welcome foreign investors efficiently.


The Regulator.  The Peruvian government chose to establish three different regulatory bodies – one for telecom, another for electricity, and a third for the road, railroad and port infrastructure.  The latter institution was OSITRAN -- Organismo Superior de la Inversión en Infraestructura de Transporte de Uso Publico (or Supreme Authority of Investment in Transport Infrastructure for Public Use). 


OSITRAN was created at the beginning of 1998 as an autonomous government entity to supervise concessioned infrastructure.  Its board of directors is composed of five members, including two appointed by the prime minister’s office and one each from the ministries of finance and transport, and one from the national competition office.


As shown in Table 3, OSITRAN has only a small staff of 50, including clerical assistants.  Only about 40% of the budget is used to pay permanent staff.  Most of the rest is to hire consultants who provide the day-to-day oversight of construction and maintenance activities by concessionaires. Its approved budget for 2007 is 17.2 million Soles, or about $5.4 million. The budget is entirely financed by payments (in most cases 1% of overall revenues) by concessionaires.


Table 3



IV. Previous Port and Road Concessions


Unlike the relatively transparent process used for other concessions, the concessioning of roads and ports prior to USAID’s involvement, has raised concerns.


The Port of Matarani.


            Prior to the concession for the new container port at Callao, only one Peruvian port had been concessioned to a private operator.  This was the port of Matarani, which serves the Southern part of the country around Arequipa, as well as transshipment trade with Bolivia.  Matarani is Peru’s second-largest port, though it only accounted for about 8% of Peru’s maritime trade, or about 1 million metric tons of cargo per year in 2000.  Of this total, nearly two-thirds is typically composed of bulk agricultural commodities and minerals.  (A second port, at Ilo, was also offered for concession at the same time as Matarani, but the concession offer was vacated for insufficiency of qualified bids.)


            This concession is described in more detail in “The Callao Port Concession.”  Briefly, however, the story is as follows.  The concession was won in May 1999 by a Peruvian firm, Santa Sofía Puertos, SA, with the port to be managed by a subsidiary, Terminal Internacional del Sur (TISUR).  The winning bidder committed to pay the Peruvian government $9.7 million for the concession and to invest $4.5 million (later raised to $9.5 million) during the first five years of operation, to pay 5% of gross receipts to the government of Peru, and 1% to OSITRAN.  The concession was to last 30 years, with contractually-specified maximum rates for services during the first five years of operation. 


            Prior to the concessioning, the previous operator, ENAPU, the government-owned port enterprise, had made physical improvements in the port and had cut staff from 315 to 60 to make it a more attractive investment. 


            After the concession was awarded, numerous questions were raised about the process.  A (possibly politically-motivated) congressional committee investigating the transaction concluded that politics had been involved in the decision to concession.  It also noted that numerous changes had been made in the concession contract after the award had been made.  These concerns, together with opposition from organized labor, led the legislature to pass a law prohibiting additional port concessions until a new national ports law had been approved by Congress.  The new law did not pass until 2003, and another year was needed to promulgate the law’s implementing regulations.  A 2003 assessment of the impact of the concession concluded that it was indeed a financial success for the Peruvian government and for importers and exporters as a whole, though some shippers and port workers did lose.


            Road Concessions


            Prior to the USAID assistance, the Peruvian government had awarded concessions for only two roads – the Matarani-Arequipa road in 1994, and portions of National Highway No. 5 in 2003.  This was far less than the government had envisaged in its “Road Development Plan 1996-2005,” which proposed concessioning eleven separate road networks. The plan visualized using cross-subsidies from high-traffic roads to fund concessions on lower-demand roads.


The Arequipa-Matarani Road


This road was concessioned in 1994 as a “one-off” transaction, with none of the institutional framework that has since come to characterize concessions in Peru.  It concessioned a 104 km. road between the port of Matarani and the city of Arequipa.  According to the World Bank (2006), the Arequipa-Matarani road concession was an exception to the generally transparent concession processes in Peru. It lacked both an appropriate institutional framework and a specialized authority.  The Ministry of Transport and Communications apparently drove the process, which lacked both a data room and a white book.  The Bank reported that it had great difficulty finding information on the characteristics and results of the different stages of the project’s bidding process.


Nevertheless, the road appears to be an economic success for Peru.  The concessionaire invested about $10 million in road rehabilitation and improvements during the first seven years of the concession.  The needed improvements were to be recouped through tolls collected from travelers.  However, the concessionaire was also required to transfer 12% of toll revenues (amounting to more than $4 million through 2006) to the Ministry of Transport for use in road maintenance elsewhere in the country.  The concessionaire also paid 1% of revenues to OSITRAN for its oversight of the concession.  Tolls fell significantly after the concession was made, though toll revenues increased because of increased traffic – particularly that of heavy vehicles transporting goods to and from Bolivia.  The road concession also significantly reduced driving times between the port and Arequipa.


            The 1996-2005 Road Development Plan and National Route No. 5


            The 1996-2005 Road Development Plan identified National Route No. 5 as the first to be concessioned under the program.  For various reasons, including the unfeasibility of the proposed cross-subsidy scheme, only a portion of the road (the Ancon-Huacho-Pativilca section, 161 km long along the coast) was concessioned, for a period of 25 years.  No means could be found to include the far less-traveled Lima-Canta-Unish section that went into the Andes mountains in the concession under the cross-subsidy scheme.  Moreover, the concession was not awarded until 2003.  The other ten roads envisioned under the 1996 Road Development Plan remained behind. 


            The winning bidder for the Ancon-Huacho-Pativilca section, Norvial, S.A., was granted a 25-year concession.  With the failure of the government’s effort at cross-subsidies, the award was made on the basis of the concessionaire’s willingness to transfer revenues to the government for road maintenance elsewhere.  As in the above case, the winning bidder committed 12% of revenues for this purpose.  Even with the reduced scope of the project, problems were evident in 2007, four years after the granting of the concession.  The most serious was the inability of the government to transfer needed land to the concessionaire.  (Clearly, completion of land acquisition by the government prior to concessioning is an obvious way to reduce risk for the concessionaire and to introduce greater certainty into the process.)



V.  The Problem of Risk in Long-Term Concessions


            For the potential concessionaire for any of Peru’s concessions, like any major infrastructure project by a foreign investor, the largest problem in valuing the concession boils down mostly to one word:  risk.  A concessionaire makes an expenditure of hundreds of millions of dollars now, in the expectation that this investment will yield profits in the future.  In any infrastructure investment except a power plant on a barge, the concessionaire or foreign investor is likely to have little hope of picking up his asset and taking it elsewhere if the concession goes sour.  Potential bidders for the concession faced three different kinds of risk associated with long-term projects like Amazon North, Amazon Central, and the Callao container port: country risk, contractual risk and technical risk.


            Country risk


            The terms of any long-term concession agreement will reflect the estimate each party makes of the risk of default by the other party.  When the agreement is between a government and a private concessionaire, the concessionaire is likely to face a far larger risk, as the concessionaire must invest money up-front in some kind of physical works, in the expectation that the government will later pay for the investment with an appropriate profit.  If the concessionaire defaults, the concessioned asset can be seized, and the government may seek further redress through the courts by seizing any other of the concessionaire’s assets attachable within the jurisdiction.  Even if this is insufficient to cover the deficiency, the concessionaire is unlikely to be barred from future business in that country.  Reputation effects may also foreclose opportunities in other countries.


            In the case of default by the government, however, the options available to the concessionaire are often limited.  This is particularly the case in Latin America, where the legal system in many of the countries offers little recourse to investors whose assets have been expropriated directly or indirectly by the government.  Over the past several years, governments in Argentina, Bolivia, Ecuador, and Venezuela have all acted unilaterally to restate terms of contracts with private foreign investors.  In the latter three countries, the government took direct steps to abrogate contracts.  The Argentina case is somewhat different in that the government did not challenge the investor’s rights, but nonetheless refused to allow rate increases on utilities that were authorized under the concession contract.  In each of the cases, contracts signed by one government were essentially vitiated by a subsequent government that felt no obligation to live up to those commitments.


            The likely consequence of country risk would be to raise the payout that the concessionaire would require in the early years of the concession.  Revenues promised for later years by the government would be heavily discounted.  Uncertainty regarding whether government will live up to its price commitments encourages a concessionaire to seek profitability in the near term – ideally, before the current government administration departs office. 


            Contractual Risk.


            Any concession will be based on a contract between the concessionaire and the host government.  Typically, such contracts empower a governmental organization charged with regulating the concession with considerable power to interpret the terms of the agreement. 


            As with country risk, the government holds most of the cards with respect to contract interpretation.  Absent a long tradition of judicial independence, the decision of the regulatory body, however much it deviates from the spirit (or even the letter) of the contract, is likely to be determinative.  And Latin America abounds with examples in which government agencies with traditions of competent technical analysis are overthrown by a new administration’s appointments to leadership positions in the agency. 


            As discussed earlier, Peru has a good record of addressing contractual risk fairly over the past few years, though technically-capable institutions like OSITRAN.  But a different outcome in the 2006 presidential election might have changed the behavior of government regulators.


            Technical Risk.


            Technical risk flows from uncertainties in building and maintaining the concessioned highway or port.  In the case of the Callao container terminal, the risks are likely to be of a relatively limited scope – unexpected problems in building the pier, or unexpected environmental conditions that require mitigation during the pier’s construction. 


Road projects in Peru are more problematic.  There are three major kinds of technical risk faced by a firm legally responsible for maintaining a given level of service on a road like Amazon North or Amazon Central (or any of the other highways that provide east-west connections between the Pacific coast and the Amazon basin over the Andes):


§         problems from landslides down steep embankments onto the road

§         slippage of the road down into declivities, and

§         abnormal weather conditions, like earthquakes or the phenomenon of El Niño


            All of these risks are increased by uncertainties and unknowns regarding the geology of the Peruvian mountains.  But the steepness of the ascent from the Pacific coast into the Andes, followed by the steep declines into the Amazon basin, make highway construction in Peru susceptible to far more technical risk than the same activity in most other countries.  The first two of the problems cited above are an annual threat whenever there is unusually heavy or persistent rainfall.  Both rainfall and runoff can cause slippage of the mountainside. 


            Earthquakes and El Niño are more problematic.  Earthquakes and volcanic eruptions are not infrequent in the Andean chain of mountains that characterizes the west coast of South America.  Earthquakes can have catastrophic consequences for specific regions, and their potential toll on highways in any region is utterly unpredictable.  Moreover, such geological phemnomena may produce massive slippages onto roads, or entirely remove road sections.  In addition to being highly destructive, earthquakes are utterly unpredictable in terms of both timing and frequency.


            El Niño, or more accurately, the El Niño Southern Oscillation (ENSO), is a periodic weather pattern where temperature changes in the Pacific Ocean along the coast of Central and South America lead to anomalous rainfall patterns, both in the Americas and in Africa.  Heavy rain may fall in arid areas and drought may impact areas typified by rain.  Such events typically occur in December, around Christmas.  Its periodic occurrence is more certain than that of earthquakes, but the actual year of the next event and its severity is not susceptible to prediction.  Table 4 provides some historical data on the years when this phenomenon occurred, and the severity of the impact. 


Table 4

Chronology of “El Niño” by Year, Magnitude and Changes in the Surface Temperature of the Ocean


Year       Magnitude     Temperature


                 Magnitude    Temperature                                                               Year                                      Change

1578           Very Severe       >8° C

1891           Very Severe       >8° C

1926           Very Severe       >8° C

1932           Weak                    2° C

1933           Severe                 6° C

1939           Weak                    2° C

1941           Severe                 6° C

1943           Weak                    2° C

1953           Weak                    2° C

1957             Severe                 6° C

1965             Weak                    2° C

1972             Severe                 6° C

1977             Weak                    2° C

1978             Weak                    2° C

1983             Very Severe       >8° C

1987             Weak                    2° C

1992             Weak                    2° C

1998             Very Severe       >8° C



            Because of its geographic and climactic conditions, Peru provides an unusually high degree of technical risk for highway concessions.  Landslides can destroy parts of the concessionaire’s road in a variety of ways:  from unstable (and perhaps unanticipated) geology, from earthquakes, from unusually high levels of rainfall, or from temporary disturbances caused by El Niño. 


            While the concession contract will include provisions for addressing these risks, the interpretations made by the regulatory agency are likely, as discussed in the previous section, to be determinative.  Thus, the degree of trust that potential bidders have in the fairness of the judgments of the regulator in addressing the various possible forms of force majeure are essential to obtaining minimum-cost bids from potential concessionaires.  Where the future judgments of the regulator are seen as potentially uncertain, bidders will seek to assure that the concession produces profits very quickly.


Other Risks


            There are risks that do not fall easily into a single one of the three categories mentioned above.  The case mentioned above of the Highway No. 5 concession, where the government was unable to secure rights to the land on which the concessionaire was expected to construct the highway, seems to embody elements of all three.  Technical risk, because the concessionaire’s engineers lacked the ability to carry out the expected construction; contractual risk, because the government failed to deliver on its contractual obligation to make the necessary land available; and country risk, because the evident inability of government to overcome local or regional opposition to the concessionaire’s rights called into question the government’s capacity (or willingness) to deliver on its commitments.


Risk Mitigation


            As suggested above, higher risk in any concession will lead to higher costs, as potential concessionaires seek to protect themselves from loss.  Each of the three forms of risk requires its own approach to risk mitigation. 


            Three ways to address country risk are external guarantees, reputation effects, and concession length.  In the case of Amazon North, the Inter-American Development Bank provided a partial guarantee of government payments to the concessionaire.  This appears to have been moderately useful, as a subsequent concession for a portion of the Amazon South highway was successfully awarded without any external guarantee.  At the country level, a government’s desire for additional foreign investment to assure its reputation as a hospitable recipient is probably the best guarantor of low country risk.  In East Asian countries like China and Vietnam, this search for future foreign investment is a good insurer of the existing foreign investment.  In Latin America, however, history suggests that the current government’s commitment to more foreign investment offers no assurance that a subsequent government may not choose to eschew foreign investment, and therefore freely seize foreign-owned assets.  Nevertheless, Peru has opted for very lengthy – 25 to 30 year – concessions in the belief that its policy stability over time justifies such long concessions.  A country with a less-favorable reputation in international capital markets might have to opt for much shorter concession periods.


            Contractual risk is the most difficult type to address in countries without a long tradition of judicial independence.  Any new government may find ways to undermine the interpretations given to the contract by the regulator, and it may be difficult to sort through the minutae of commercial disputes to understand whether the government or the concessionaire is at fault.  As with country risk, the strongest protection for concessionaires is a government’s desire for a reputation for treating foreign investors fairly.  Nevertheless, a long tradition of judicial independence, by which a concessionaire can expect a day in an unbiased court, is a far better vehicle for reducing contractual risk.


            Fortunately, Peru had more than a decade of experience with privatization and concessioning before the award of the Amazon North highway.  As shown by Table 2, more than 14 concessions or privatizations had occurred during this period, providing Peru with a track record of dealing fairly with foreign investors in such infrastructure activities.  Still, in Latin America, such fair procedures may be only an election away from elimination.  In 2006, a government committed to promoting concessions and foreign investment was elected in a narrow contest.  Had the election gone the other way, the concession and privatization contracts might have been at risk.  The electoral outcome certainly influenced the behavior of companies interested in the IIRSA-North concession, likely leading them to add millions of dollars to their bids.


            Technical risk is the easiest type of risk to address when preparing for a concession award.  Jacob Greenstein’s paper, “Perú: case study of public sector and private sector risk management procedures” discusses the technical risk issues in the specific context of the IIRSA-North highway.  He concludes that a greater degree of pre-award cost-engineering and geological work would have reduced the risk to bidders on the concession, and thereby reduced the award’s likely cost to the Peruvian government.  This involves two aspects:  information about the geology of the area through which the highway is being constructed, and sound engineering estimates of construction costs, based on explicit estimates of unit costs for the various elements of the construction project.  Both types of studies would reduce both the risk to a potential bidder and the cost of preparing a bid.  Both factors should increase competition and reduce the cost to the government in issuing the concession. 


            For USAID or other donors seeking to promote concessions, inclusion of funding for these technical studies may be highly valuable.  An additional $1-2 million on engineering studies made available to bidders is likely to translate to multiple times that value in terms of bidders’ reduction in perceived risk, thereby reducing the real cost of concession.



VI. USAID’s Decision to Promote Concessions


            USAID’s decision to promote public-private partnerships (PPP) or concessions in Peru was not the result of any interest or request for assistance on the part of the Peruvian government or of any extensive understanding of the country’s previous experience with concessions and other PPP.  Rather, it resulted from two factors. 



      USAID believed that there was a possibility that hundreds of millions of dollars in new investment in Peru might be mobilized by USAID funding of only a $5-10 million.  Thus, a micro-level action might achieve macro-level outcomes for the country, and for the level of poverty.  The Director’s previous experience also led the mission to include a performance bonus in the contract with the implementing consulting firm.  The consultant was to be rewarded for actual concessions, not simply for preparing concessioning proposals. 



In the mission’s view, promotion of alternatives to coca could not be limited to convincing farmers to cultivate mangos instead of coca.  Under existing conditions, only farmers wishing to earn a dime instead of a dollar would be candidates for such a switch.  USAID believed, with considerable evidence, that improvements in both the country’s infrastructure and its policy environment were needed if alternatives to coca production were to achieve reasonable levels of profitability – and thereby impact production and employment.  In the infrastructure area, there were two major obstacles to lowering the cost of transporting products from producers in the coca-growing areas to markets in urban areas or in foreign countries. 


For domestic markets, the state of the highways connecting the coca areas, mostly in the Huallaga valley in tropical Peru, was the main obstacle.  Two highways, one in the north and one in the center of the country, provided trans-Andes links between the coast and the tropics.  Both included sections in need of upgrade and required overall restoration – on the order of 1,000 kilometers each. 


Like most other developing countries, road maintenance in Peru was commonly shortchanged for government funding in favor of new road construction.  The latter is politically more popular.  The opening of a new road is a “photo opportunity” and is almost always welcomed with enthusiasm by the beneficiaries.  A failure to maintain roads, on the other hand, leads to a steady, but gradual, deterioration over a period of years that is less visible from a political perspective.  Eventually, however, the road may need to be entirely rebuilt at cost far higher than that of proper maintenance.  Because of the slowness of this deteriorative process, it is not easy to assign blame for the deterioration.  By contrast, the opening of a new road is a highly visible act, and politicians will enthusiastically step forward to accept credit.  Even the rebuilding of a badly deteriorated road would provide an opportunity for politicians to contrast their commitment to the areas served by the road against the neglect of the area by previous governments, as evidenced by the previous deterioration of the road.  (As discussed in the case study of the Amazon North highway, this political dynamic was one of the factors that promoted an enthusiasm in the Peruvian Ministry of Transport and Communications (MTC) for concessions.  Concessions establish a legal obligation for the concessionaire to maintain roads to certain levels of quality throughout the life of the concession.)


For Peruvian products with potential for export to other countries, the state of the roads from the jungle to the coast was only the first obstacle to competing in world markets.  The second was the country’s main port at Callao, near Lima.  Callao is responsible for more than 80% of the country’s seaborne international trade.  The port is notorious for high costs and slow movement of goods.  A half-century after the beginning of the container revolution, Callao had the distinction of being the only one of the ten largest ports in Latin America that has no onshore gantrys to move containers.  The port had to rely on – far slower – shipside gantrys for loading and unloading of containers.  Understandably, ship-owners would much prefer to carry more containers instead of the excess weight of these huge cranes.


USAID pursued the concessioning effort by amending an existing contract with Chemonics International to redirect up to $12 million in funding for this purpose.  (The amount was later reduced to $8 million due to budget cuts.)  This amendment was approved in August 2003.  The contract amendment tasked Chemonics with promoting between four and six concessions. 


USAID and its consultants report that there was little interest by the Peruvian government in pursuing PPP, or concessions specifically, during the first six months after USAID charged Chemonics with promoting the concept.  After that period of apparently fruitless effort, the government suddenly became an enthusiastic supporter of road concessions.  Three factors seem to have been important, though it is impossible to allocate the share each factor contributed. 


First, Peruvian President Toledo became enthusiastic about transport integration within South America.  This concept flowed from a regional summit of South American presidents in 2000, with support from the Inter-American Development Bank.  This produced IIRSA (the Project for the Integration of South America), which proposed a program for infrastructure interconnection among the countries of South America.  It initially languished, but it appears that President Toledo became energized by his Brazilian counterpart, President Lula da Silva of Brazil during a visit to Brazil.  At least according to legend, da Silva complained of the inefficiency of Peru’s main port at Callao and of lack of action on improvements on the three Peruvian roads that were to connect Peru with its eastern neighbors as parts of the IIRSA network:  The three roads were:





            Whatever the motivation, concessioning of two of the roads and a container port at Callao was completed by the time that Toledo left office, and concessioning of the third road was well advanced.


            Second, the closing-off of privatization as a vehicle for downsizing the public sector and generating foreign private investment after the Arequipa judicial decision made concessions the only game in town. 


            Third, the availability of grant money (ultimately $8 million) allowed for far more technical work on concessioning than would have been financed had only Peruvian government resources been available.


            USAID agreed to support concessioning of three of the four projects.  IIRSA South was deemed to be little-connected with the rationale for the USAID program of coca eradication.  That road did not connect to the main Peruvian coca-growing zones in the Huallaga valley.  Moreover, the decision to concession it did not flow from analyses identifying it as a high priority (indeed, they suggested the opposite) but from a quickly-taken instruction from the highest level of government.  The lack of a technical basis for the concession also made USAID leery of supporting it.



VII. What USAID Assistance Did


          The three case studies provide extensive detail on the specific steps taken in each of the three concessions that USAID sought to promote.  The Amazon North case study also includes an extensive discussion of the various forms of risk that are inherent in concessioning.  This section offers only a brief overview of the process undertaken in each of the three cases, and the outcome that resulted.  Chart I, on page 1, shows the scope of the three concessions.  


          Amazon NorthThis concession of a 960 km. highway from the Pacific coast port of Paita to an Amazon tributary at Yurimaguas was concluded in July 2005.  ProInversión had approved making the concession in August 2003.  Nine firms were initially pre-qualified, and four firms were qualified at the late stage.  Three of the pre-qualified firms (Odebrecht and Adnrade Gutierrez of Brazil, and Graña y Montero of Peru) formed a consortium very late in the process, substantially reducing competition.  Only one other bid, from another Brazilian firm, was received.  The three-firm consortium won the concession.  At least one of the reviewers of the two bids considered the second bid to be unresponsive to the terms of reference, and likely to have been submitted as “window dressing” to maintain the veneer of competition. 


            The cost of the concession rose from an early estimate of $150-160 million to $218 million, as the concessionaire identified additional costs.  A significant part of the additional cost appears to have been in the construction of the portion of the road from Terapoto to Yurimaguas to higher technical specifications than expected traffic flows would seem to warrant.  Still, the relatively generous concession for the Amazon North road should have served as a stimulus for additional bidders on subsequent concessions.


            The Callao Port Concession.  The port of Callao, handling about 80% of Peru’s maritime trade, has been a critical bottleneck to Peruvian development for decades.  The World Bank and IDB have long sought to promote the upgrading of the port in any of various ways – through privatization, through concessions, or by direct investment in the infrastructure of the port.  Callao is the only one among South America’s ten largest ports without landside gantry cranes to load and unload containers.  At present, all container ships calling at Callao need to have cranes on the ship – dramatically slowing off-loading and on-loading compared to landside cranes.  This has made Callao one of the highest-cost ports in Latin America.


            USAID promoted a national dialogue on the topic, including participation of all affected groups, drawing on a quantitative study of the drag on Peru’s economy from the excess costs of moving goods in and out of Callao.  It also sponsored a study tour of four other ports in Latin America, including labor unions, shippers, government and other private sector leaders to identify the best options for improving the port.  At least eight technical studies of various aspects of the demand for port services and related topics were completed by USAID contractors, many before the formal decision to concession the port.  USAID consultants helped ProInversión formulate its concession strategy, one that demanded world-class expertise from any bidder. 


            Ultimately, the concession was awarded to Dubai World, which committed to investments of more than $600 million in two phases of construction and expansion of a new container port at Callao. 


            The Amazon Central Highway Concession.  The Amazon Central highway was expected to be concessioned on July 25, 2007.  There was a considerable gathering of people in dark suits in the conference room at the government of Peru’s investment promotion agency, ProInversión, awaiting the opening of envelopes containing offers from any of the eight pre-qualified firms for a major highway concession.  This concession was for the Amazon Central highway, traversing the 854 km. from near Lima, Peru’s main port, through the Andes to the Amazonian basin town of Pucallpa, near the Brazilian border.  This road was the “crown jewel” trans-Andean highway, much more important than the Amazon North road, concessioned two years earlier, and a portion of the Amazon South to be concessioned later.  The expectation had been that this concession would be self-financing, with tolls from road users paying the cost of the improvements mandated under the concession offer. 


            But the convocation produced no offers.  This failure of the country’s most economically viable road concession raises the question of the impact of the assistance financed by USAID to promote this concession. 


            At the micro level (the assistance being provided by the USAID-financed consultants to ProInversión, the investment promotion arm of the Peruvian government), the assistance seems to have been effective.  USAID financed studies of the technical risks inherent in the Amazon Central project, and the Peruvian government further reduced such risk by agreeing to take responsibility for two especially vulnerable sections of the road.  USAID-also financed a “road show” in late 2006 to encourage construction firms in Brazil, Colombia, Ecuador and Mexico to offer bids for the concession.


            At some point in early 2007, however, the Ministry of Transport and Communications (MTC) began to assert a stronger role in the process, arguing that the concession should not be self-financing from toll collections from the highway, but that some of the toll collections should be made available to the MTC for other road projects.  A complex restructuring of the flow of funds from tolls replaced the simple version that had been promoted in the road show.  As potential bidders raised questions about the revision in approach, a stream of revisions of the contract terms was forthcoming.


            The “crown jewel” of Peru highway concessions has so far been a failure.  There is hope that a new concession proposal may be more successful.  Personnel changes are likely to improve the prospects.  The senior MTC official who led the revision of the terms of the concession was replaced.  The ProInversión official initially responsible for managing the Amazon Central concession, but who resigned in protest of the MTC’s revisions of the draft concession agreement, has now been hired by the MTC to promote the concession.


VIII. What USAID Assistance Did and Didn’t Accomplish


            The case studies of the three individual concessions promoted by USAID provide considerable detail and justification that provides a basis for the judgments reached here.  Nevertheless, any judgment about what was accomplished because of USAID assistance must be tentative and should reflect an awareness that USAID was only one of a number of actors whose presence effected the outcome.


            With this caveat, the main findings of this study are the following six propositions.  They are ranked generally in descending order of ease of defense of the proposition to a skeptic.  The first two are easily defended by abundant quantitative and qualitative evidence.  The next three are more judgmental, but draw upon interviews with major actors, both Peruvian and foreign, and from historical accounts (e.g., World Bank 2006) of the evolution of Peruvian concessioning.  The final proposition is more speculative, as it extrapolates from a single successful country case to many different possible circumstances.


  1. There was a significant improvement in the quality and quantity of transport concessions during the period of USAID assistance.


            If one compares the Matarani Port concession, or the National Highway No. 5 concession (or even the IIRSA South concession made during the time of USAID involvement) with the Amazon North and the Callao port concessions, there can be little doubt that the quality of the concessions where USAID was involved was higher than those where USAID remained uninvolved.  The impact of USAID involvement cannot be separated from the efforts of the Peruvian government, but the simultaneousness of the improvements and USAID’s participation suggest a relationship between the two.


  1. The two concessions that USAID promoted were of great value to Peru.


            Based on the available evidence on which to predict future outcomes, this is a “slam dunk.”  The new Callao container port – based on the available projections – will be such a huge success that the value of the three together will be very large.  Projections of the annual savings by Peruvian importers and exporters from the new container port range from $200 million to $1 billion per year.  The lower estimate ignores the potentially large benefits that could come from additional Peruvian exports of products that currently are not internationally competitive because of the high transaction costs at the port.  And these calculations exclude the investment, at no cost to the Peruvian government, of $300-400 million by the concessionaire over the next several years.  The net benefit of the Amazon North highway is likely to be substantial, and the assurance that the road will be properly maintained for the next several decades is a welcome departure from traditions in many Latin American countries.


  1. The concessions would likely have happened without any USAID assistance, but the concessioning probably would have been slower and less technically sound.


            The three concessions promoted by USAID happened primarily because the Peruvian government decided that they should happen.  Nevertheless, in all such decisions from the top, there is much potential for slips twixt the cup and the lip.  The Fujumori government “decided” in 1995 to concession 11 roads.  None were concessioned until 2002, two years after he left office.  Both Napoleon and Hitler “decided” to invade England.  But neither did, because their subordinates could not figure a way to make it happen. 


            In sum, the three concessions probably would probably have happened without USAID, but USAID’s involvement increased that probability, perhaps significantly.


  1. The groundwork for successful concessions was laid by more than a decade of stable and increasingly transparent government policy.


            The success of any concession depends on the country, not the commitment of the government in power.  The experience of several of Peru’s neighbors (Argentina, Bolivia and Ecuador) is a clear demonstration of the impotence of concessionaires in the face of a government determined to ignore its legal commitments. 

            No government can commit its successor to enforce contracts that it subscribes.  Only a tradition of continuity during changes in government can assure this, along with an expectation that the rule of law will be maintained.  Peru’s experience in the 1990s with the early concessions demonstrated this commitment, making bidders in the subsequent decade more willing to make long-term investments in concessions. 


  1. The country’s capacity to undertake concessions resulted from a period of trial and error, and learning by doing. 


            For Peru, the period between the first legislative approval for concessions and the establishment of a competent and efficient system for awarding concessions and overseeing them was more than a decade.  This is reasonable, given the complexity of the legal and technical requirements for well-designed concessions.  The significance is that concessioning needs to be a commitment on the part of the country, and not only of one governmental administration.  As in Peru, the effort by any single government to rush the process so that concessions can be awarded during that government’s tenure is likely to raise the cost and lower the quality of the concessions.


  1. Promoting concessions should be a high priority for USAID.  


            USAID spent a total of $8 million promoting the three concessions.  The impact of these concessions on Peru’s economy and the wellbeing of the country’s ordinary Peruvians will likely be enormous.  Shippers will save hundreds of millions of dollars each year, making Peruvian products more competitive abroad, and making imported goods cheaper for consumers.  In addition, the projects stimulated more than $800 million in private investment in infrastructure. 


            As suggested above, these concessions probably would have happened without USAID support, but they were executed in a way that was more technically sound, and likely faster, would have been possible without USAID support.  As a result, the concessions probably attracted more competition than otherwise possible. Putting a price on the value of this assistance is an exercise in speculation, but a value to Peru of $100 million would not seem unreasonable.  Few investments by USAID offer returns of such magnitude.


            Nevertheless, as suggested above, this is a very difficult area.  Success is likely to require several critical factors:






            As this list of requirements indicates, concessioning is not for everyone.  It is a more promising instrument in middle-income countries than in poor ones, because the institutional structure is likely to be stronger and provide more assurance of continuity over time. 


            Finally, anything but very short-term concessions should be avoided in fragile or post-conflict states, as bidders would typically operate on a time horizon of one or two years.  They would expect to make a profit on the endeavor during that time, so the cost of concessions would be inordinately high – and excessive if the country actually returns to stability.


  1. Still, USAID management of concessions in Peru was imperfect


            After the departure of the Mission Director who stimulated USAID’s involvement in PPP generally, and concessions specifically, the USAID Mission’s interest in this topic seems to have gradually diminished.  The relatively small size of the activity probably contributed to this.  At only about $2 million per year, the work on concessions was a relatively small (and probably relatively trouble-free) part of the USAID portfolio.  Perhaps as a result, this activity seems to have suffered from increasing inattention by the USAID mission.  This is despite the enormous potential of the concessions for the country – probably far larger than most of the rest of the USAID portfolio.


            Speculatively, one might argue that the failure of the Amazon Central concession may have arisen from USAID’s failure to weigh in when the MTC essentially hijacked the basic concept for the concession jointly developed by ProInversión and the USAID contractor.  High-level representations by USAID to the Peruvian government might have prevented this outcome.  At worst, the failure of such an attempt would have allowed USAID later, after the failure of the concession process, to say “I told you so.”  But no USAID staff were present at the envelope opening for Amazon Central, and they were only vaguely aware of the date of the award.


            In sum, the Peru experience suggests that success in PPP requires a change in USAID mentality – away from managing money and toward promoting positive change.