One of the most ambitious projects of the Chilean government is the transformation of some 1,400 km of the so called Pan American Highway (currently just a B road in most parts) into a four-lane tolled motorway. However, President Frei has recently promised to substantially improve the railways, raising their commercial speed from some 40 to more than 100 km/hr. On the other hand, air traffic has been increasing at 15% per year and new operators are increasingly joining the market. Finally, the interurban coaches, which have easily the highest market share, are fiercely competitive and offer an outstanding quality of service. Thus, forecasting is not a simple matter in this case. This paper discusses the data this case. This paper discusses the data collection, model structure, specification and estimation results of a system of forecasting models designed to evaluate the tolled motorway project from a social point of view. Stated Preference (SP) rank data for choice among car, coach, train and also plane in many O-D pairs is being collected for a huge sample of road users travelling between 50 and 1,000 kms, using mail-back questionnaires with block designs for various journey length strata. At the same time revealed preference data for bus and tram users with car access is also being gathered. Finally, binary route choice SP data is being collected for various cases: tolled bypasses vs city centre crossings, tolled motorway vs paved, gravel (B or C) and mountain roads. The model system has to combine all this consistently in order to forecast the motorway demand and examine the social worth of the project under several growth, car ownership, toll charge and public transport fare scenarios. The paper concludes by summarising the methodological contributions of the study and discussing the importance of its main assumptions on the results.
Samenvatting