Economic value of Alta Velocidad Española

Great Britain is not the only country where high speed rail investments have been pursued without effective transparency mechanisms or sanity checking. In the May 2012-updated ‘Economic evaluation of the High Speed Rail‘ Ginés de Rus noted the lack of information available on the Alta Velocidad Española projects.

Before dealing with the evaluation of the Madrid – Seville HSR line, it is worth mentioning the scarcity of published information on costs and demand. The lack of transparency concerning the economic profitability of HSR in Spain is remarkable. The economic effects on the HSR lines circulated by the Spanish government have no relation whatsoever with the economic appraisal of projects, but rather with the type of impact studies where investment is not a cost but a benefit, including the multiplier effect, labor is not an input but an output, and transfers and relocation of economic activity are not considered, etc.

The de Rus evaluation went on to put the probability of the Madrid — Sevilla AVE line having a positive net economic value at zero.

The economic evaluation of the Madrid – Seville line is based on the cost and demand information contained in de Rus and Inglada (1993) but this is the only common content of that original CBA and the present in media res CBA of the first Spanish HSR line. In what follows we describe the CBA of the Madrid – Seville and interpret the result of the density function of the net present values obtained through a risk analysis of the project.

Madrid – Seville was the first HSR line constructed in Spain. The 471 km line started its operation in April 1992. The investment period lasted from 1987 until 1993. Total investment was €2.1 billion (1986) and these costs were distributed during the construction period according to the real disaggregation of these costs per year as reported in de Rus and Inglada (1993). Investment costs are inclusive of indirect and income taxes. The project is assumed to have an economic life of 50 years and for simplicity we ignore any potential residual value.
[…]
The expected 1987 net present value is negative, and is equal to €-2.27 billion in 2010, equivalent to 55% of the construction costs (€4.1 billion). The risk analysis shows that the probability of having a positive NPV is zero. In fact, the best result shown in the density function of NPVs is €-1.7 billion. Although one should be cautious, given the limited information available, the cost-benefit analysis of the line is conclusive in terms of the negative social value of this investment.

[…]The sensitivity analysis…shows the results are quite robust to a lower social discount rate (3%) or an increase in the time cost component of the generalized costs of all the alternative transport modes by 25% to account for willingness to pay for HSR higher comfort, service quality, or whatever. In both cases, the probability of a positive NPV is zero.

The evaluation also looked at the more recently completed high speed line, between Madrid, Zaragoza, and Barcelona.

We are unaware of any comprehensive cost-benefit analysis of the Madrid – Barcelona HSR line. In this section we undertake a full cost-benefit analysis of the investment in the Madrid – Barcelona HSR line. It is not an ex ante CBA as the construction started in 1999, and was completed in 2007, but it became partially in service between Madrid and Zaragoza in 2003 (services between Madrid and Barcelona in 2008). Nevertheless, although we are conducting this study in 2011, the official information available for the CBA is not as good as one would expect for a public investment. However, we have gathered some data from different sources and, based on some reasonable assumptions for the reconstruction of the unavailable information, have been able to calculate some sort of in media res NPV.
[…]
We assume demand is distributed uniformly during the day (no peak hours). The load factor follows a random uniform distribution within the range 0.6-0.7 [i.e. 60 – 70%]. The number of daily services required, given the load factor, the contingency factor, the length of the route and the hours of operation are computed according to Campos et al. (2009). Trains operate 16 hours per day and do not exceed the maximum number of kilometers per year (500,000 km).

Other general assumptions affecting key parameters are the following: the annual growth rate of income is taken from the National Institute of Statistics until 2009. From 2010 to 2015, the estimated growth rate is equal to 1% according to the IMF. For the rest of the evaluation period, the annual growth rate follows a random uniform variable between 1% and 4%, independent between years. The net present value is calculated at the beginning of 1999 with benefits and costs expressed in €1998 prices deflated with the CPI of the National Institute of Statistics. The social discount rate of the base case is 5% and the shadow multiplier of public funds is 1.

The HSR travel time between Madrid and Barcelona is around 3 hours, and many passengers benefit from these time savings, although as we will see, the picture is quite different depending on which transport modes passengers were initially traveling. Moreover, it is revealing to distinguish between the changes in generalized costs and the differences in resource costs, ignoring transfers.
For the period 2003-2009, the demand is based on real data while for the rest of the project life we have projected the number of passenger-trips of previous years, assuming that the demand-income elasticity is equal to 1.

In 2009, the HSR transported 5.5 million passenger-trips in the whole line: 2.6 million between Madrid and Barcelona; 1.9 million between Madrid – Zaragoza or Zaragoza – Barcelona and the rest between some other O-D.
[…]
In the case of the O-D Madrid – Barcelona line, the origin of HSR passenger trips is as follows: 43% (plane); 27.3% (train); 16.1% (car); 3.2% (bus) and 10.4% (generated). In the O-D Madrid – Zaragoza and Zaragoza – Barcelona: 2.7% (plane); 49.3% (train); 20% (car); 1.3% (bus) and 26.7% (generated).
[…]
The expected 1999 net present value is negative, equals to €-5.26 billion in €2010, as shown in Tables 4 and 5, and this is 66.4% of the construction costs. The risk analysis shows that the probability of having a positive NPV is zero. In fact, the best result shown in the density function of NPVs is €-4.25 billion (see Figure 14). Although one should be cautious, given the limited information available, the cost-benefit analysis of the line is conclusive in terms of the low value for money of this investment.

The results obtained in the economic evaluation of the Madrid – Barcelona HSR line reflect several undisputable facts: demand is extremely low. Only 2.6 million passenger-trips travel the whole length of the line. The other 3 million only cover half of the length or less. The majority of passengers were initially traveling by conventional rail, or by air, where the time benefits are quite low. The massive fixed costs of the line (63.8% of total costs) and the fact that variable costs are higher than the benefits from time savings and generated demand show that this investment requires a higher volume of demand to be socially worthy.

The sensitivity analysis…shows the results are quite robust to a lower social discount rate (3%) or an increase in the time savings benefits of 25% to account for willingness to pay for HSR higher comfort, service quality, or whatever. In both cases the expected NPV is negative and the probability of a positive NPV is zero. The results show the importance of the avoidable costs in other modes of transport. Our assumptions are quite favorable to this resource cost savings. If this were not the case, the economic implication would be much more serious.

The report also mentioned Kageson and Westin’s conclusions on the environmental benefit of upgrading existing lines.

The comparison between the negative externalities during the construction of the line and the positive externalities from the diversion of traffic from road and air to HSR is further analyzed in Kageson and Westin (2010). The results do not support the belief that investing in HSR is good for the environment. Using a Monte Carlo simulation Kageson and Westin estimate, using more than 10 million annual one-way trips, the traffic volume required to balance the CO₂ emission from the construction of a 500 km line. The authors conclude that from a climate point of view it may be better to upgrade existing lines and to try to make people substitute air travel by modern telecommunications rather than investing huge amounts of resources in making us travel faster and more.