CBA is rarely used to evaluate public investments or policies. Environmental regulation is usually implemented by imposing limits and bans in the absence of evaluation. This is the case, for example, for the net zero CO2 emissions target set by the EU for 2050, the ban on selling internal combustion vehicles from 2035 and, more generally, for air quality limits.
But, even when CBA analyses are indeed carried out, some doubts come to the surface.
Most studies are conducted by organizations interested in obtaining the government funds for the projects under scrutiny. They risk nothing and are not accountable for their mistakes. The same applies to the ultimate decision maker (a politician or a bureaucrat).
In this context, it is not surprising that the estimates contained in the analysis are usually wrong and almost always biased: costs are underestimated and benefits overestimated. About twenty years ago, the Danish scholar Bent Flyvbjerg collected data on over two hundred major projects spread across the world. Well, in the case of railways average actual construction costs exceeded estimated costs by 45 percent, while the actual number of passengers turned out to be 39 percent less.
Deviations were more contained in the case of highways, where funding is often fully or partially private: actual average costs were “just” 20 percent above estimates, and the traffic deviation was overestimated by 9 percent.
In a similar vein, benefits for the consumers are also miscalculated and usually overestimated. The benefit for people who switch from car to train is measured as the difference between time and cost by car and time and cost by railway transportation. Yet, many studies do not take into account the “hidden” cost of the fact that the train does not always take you door to door. The entire trip, therefore, involves additional transportation and costs.
There’s also a significant issue with quantifying the cost of CO2 emissions. In the case of other pollutants, experts simply consider the estimated damage. In the CO2 case, however, the benefits of lower emissions (for example by shifting freight from road to rail) are estimated by taking into account the zero-emissions target imposed by the EU to be obtained before 2050 and a dynamic abatement target that at 100€ per ton of CO2 in 2016 and reaches €800 in 2050. Last but not least, the benefit of each ton of CO2 that is not emitted thanks to a project is assumed equal to the cost that would have been necessary to abate it (and not to the damage it causes).
Not surprisingly, very few public investment projects have been dumped because of poor CBA scrutiny.
One of these projects comes from Italy. In 2019, the government formed by the Northern League and the Five Star Movement appointed a commission (of which the author was a member) to evaluate five major planned infrastructure investments. The evaluation gave a positive result for two of them and a negative one for the other three. While the positive judgments triggered no major reactions, all political parties except the Five Star Movement strongly criticised the CBA studies with a negative outcome, and the new railway line between Turin and Lyon in particular. Strong critiques also came from business associations, unions, academia, and almost all media outlets.
The main arguments against the Turin-Lyon CBA evaluation were: (1) the fact that some works had already been done (in fact, the money already spent should not be taken into account); (2) an alleged error in the methodology adopted, that is, taxes should not be considered since they are “only” a transfer between the consumers and the state. In reality, the lower taxes paid are included in the benefit to the consumer (for whom they amount to a cost reduction) and precisely for this reason, CBA must consider it.
People also objected to the fact that the more trucks were shifted to rail, the worse the outcome became. In fact, many critics failed to notice that the shift is due to the different level of taxation that the two modes of transportation involve.
A similar situation can occur when renewable energies are subsidized (or taxed less). The consumer may prefer renewables over fossil fuels, but the community as a whole ends up worse off, despite the environmental benefit. More generally, as pointed out long ago by Pascal Salin, experts cannot truly estimate the damage suffered by the victims when their property rights are violated.
To summarise: in theory, cost-benefit analysis is less arbitrary than dogmas, intuition, political preferences or interest-group power. In practice, however, it can be easily manipulated by those who are merely interested in expanding their power or pursuing their private interests. One can only hope that CBA studies are not rigged “too much”, and ensure that at least part of the investment is private so that profits and losses reward good decisions and sanctions poor decisions, at least in part.
Photo by Ricardo Gomez Angel
