MEXICO “Gross misculation” messed pikes — World Bank


MEXICO “Gross misculation” messed pikes — World Bank

Originally published in issue 20 of Tollroads Newsletter, which came out in Oct 1997.

Page:4

Subjects:financing concession policy

Agencies:CAPUFE

Locations:Mexico

Sources:Francisco Javier Alejo Lopez

MEXICO

“Gross misculation” messed pikes — World Bank

A World Bank analyst, Jeff Ruster, writes in a Bank “Viewpoint” paper that “gross miscalculation” of costs and revenues by investors and lenders and poorly designed concession arrangements by the government led to the many failures of earlier Mexican toll road program. Last issue (TR#19 Sept 97 p15) we reported a government recovery program. The World Bank report (“A Retrospective on the Mexican Toll Road Program 1989-1994” No 125 Sept 97 from Sandra Vivas 202 458 2809 svivas@worldbank.org) says the program did more than double the country’s toll network from 4500 to 9900km. 53 concessions were awarded in the 5-years covering 5500km of pikes of which 5120km were in operation in 1992.

$13b were spent including $6.7b from local banks, $2.5b of government grants or equity and $3.8b of concessionaire equity. About $5b of the $6.7b of bank loans for the pikes were in default before the recent government move, the World Bank report says, a situation which simply “shut off” capital flows to the sector, despite calculations that another 6500km of toll roads may be justified.

A major criticism of the failed concession program is that competition for award of concessions was limited to “a handful of local construction companies that were more interested in the construction work than in the longterm viability of the projects.” These companies saw profit margins of 35 to 50% in the construction work. The government’s disdain for project viability was also evident in the 15-year concession term and the provision that if traffic exceeded a “guaranteed level” the term of the concession would be reduced, but if it fell below such the level the concession would be extended — enfeebling incentives to garner traffic. In the event Ruster writes 5 of 32 projects were above the ‘guaranteed’ level, 10 were in the range 50 to 100% and 17 were 50% or more below it. On average revenues were 30% below projections, Ruster writes.

Many of the projects bids were required to be prepared so quickly there was no time for proper cost estimates and revenue studies even if the bidders were inclined or able to conduct them. Fieldwork for traffic surveys was limited to 1 to 2 weeks. Bank lenders waived normal conditions for granting loans, based on the tacit underderstanding that the government would bail out failing concessions. Everything invested in the projects, all the phsyical assets created, became legally the property of the government of Mexico, a provision which may be have led the lenders to see financial responsibility as laying in its supposedly deep pockets, another possible incentive to recklessness. Actually both the federal and state governments failed to deliver on many of their undertakings. State governments many time defaulted on contracted loans and equity contributions, while permits were agonizingly slow to be granted. Toll rate flexibility was too little, government permission being needed for every change.

Further the contracting process produced a fragmented toll network, with some high priority segments never being offered for bid: “The piecemeal pattern of contracting reduced the near-term attractiveness of the toll roads to longdistance traffic, particularly to truckers.”

The report says the haste to build led to construction starts before design, even without right-of-way fully acquired, which enabled protestors or property owners to extract huge concessions from the concessionaires. Meanwhile a stream of change orders issued from the ministry of transp. Expensive equipment lay idle as permitting and routing problems were resolved. One highway designed and originally approved with 4 pedestrian crossings ended up having to be built with 60.

Many of the projects went forward on the basis of loans 10 percentage points over going rates paid by the government. Lack of clear contractual arrangements and of independent engineering supervisors made many projects more expensive than expected. Still the system built the highways for $2.6m to $2.8m/km compared to initial estimates of around $1.7m/km.

Low traffic is attributed in part to the over-high tolls. Francisco Javier Alejo Lopez, chief exec of CAPUFE, the government toll operator says that the private concessions generally pitched their tolls for cars at around 75c/km — not the 10c to 25c/km reported in TR#19, Sept 97 p15. He says reductions being implemented in the government bail-out arrangements are 35% for trucks and 15% for cars. (Tempted to joke — these are toll rates pitched to drug trade incomes!) The World Bank report notes too that traffic suffered too from lack of ancillary services on the new roads such as fuel stations, rest stops, restaurants and other services, and poor marketing of the pikes. The maintenance of adjacent free roads also clearly undermined the viability of the toll roads.

Ruster does not criticize the inital concept which was to use tolling to substantially upgrade the major border crossings with the US and five main highway corridors:

• Nogales-Culiacan-Tepic-Guadalajara-Toluca-Mexico City: 721km

• Nuevo Laredo-Reynosa-Monterrey-San Luis Potosi-Queretaro-Mexico City: 480km

• Ciudad Juarez-Mexico City-Puebla-Oaxaca: 340km

• Mexico City-Veracruz-Sayula-Ocozocoautla-Arriaga-Puerto Madero: 428km

• Tuxpan-Pachuca-Mexico City 222km

The encouraging conclusion to be drawn is that the financial failures do not necessarily reflect any fundamental non-viability of toll road projects in Mexico or even of the projects built. Properly designed and permitted under a better concessioning system many should have worked, and could still be made to work. (Contact Jeff Ruster 202 473 9175 jruster@worldbank.org)