The above discussion leads to the obvious question of why the French government would structure its concession process and agreements in such a way as to not maximize the concession price, or in other words, to leave money on the table. The answer to this question lies in the concept of the "public interest" which, for the purposes of this paper, we define as being equivalent to the economist's concept of social welfare. Social welfare includes the welfare of all agents involved in or affected by a policy or situation; consumers (users), producers, workers, and taxpayers. There is a fifth party if the policy has relevant external effects (for instance congestion, pollution, etc.). This social welfare concept can be expressed by the following function:
PI≈SW = CS + β PS + γ WS + δ TS + EE ,
where SW stands for Social Welfare, CS stands for Consumers' Surplus, PS stands for Producer Surplus, WS stands for Workers' Surplus, TS stands for Taxpayers' Surplus,10 and EE stands for External Effects α, β , γ, and δ are the different weights given to the welfare of each of these groups. When policymakers make decisions with redistribute implications, they are implicitly weighting each of the categories. If one does not want to consider distributional considerations, we can assume that α = β = γ = δ. In this way, transfers between groups do not change aggregated social welfare. It is worth noting that even in this trivial case transfers tell interesting stories about winners and losers.
With respect to toll road concessions, the following social welfare relationships hold:
(a) If lengthening the concession period results in a higher concession price, then local taxpayer surplus increases.
(b) Higher tolls paid by consumers results in extra profits for the producer -the concessionaire-, and this, in turn, increases taxpayer surplus because of the resulting higher concession price
Merely transferring cash from consumers to taxpayers does not increase overall social welfare. In the cases of the Chicago Skyway and the Indiana Toll Road it is clear that the city and state, respectively, gave more importance to taxpayers' welfare than to consumers' welfare. It is easy to explain how this occurred given the fact that all local taxpayers are voters in local elections (regional taxpayers are voters in regional election), whereas only a fraction of Skyway users vote in Chicago, and many users of the Indiana Toll Road are from out of state.
Using this framework, we can identify some tangible measures to contrast the two approaches.11 From the perspective of the public owner (e.g. the French Government, the City of Chicago, the State of Indiana) and with respect to taxpayers' surplus:
a) A larger vs. smaller concession price is better, regardless of the use of the sale proceeds.12
b) A large concession price that is financially unsustainable ('the deal goes "south"') may or may not be a bad thing depending on the "take back" provisions.
c) A shorter concession term vs. a longer concession term may or may not be better depending on how the public owner assesses risks of ownership (e.g. the contingent liabilities of on-going maintenance, bond payments) vs. the value of future operational flexibility.
The much higher prices paid for the Skyway and Indiana Toll Road concessions result in the taxpayers in Chicago and Indiana being winners, as contrasted with the French concessions and taxpayers.13
With respect to consumers' (toll payers') surplus: (a) Lower tolls vs. higher tolls are better (b) Better maintained and operated (efficient, safe, service-oriented) roads are good.
The most striking difference between the French and the US approaches relates to the toll setting formulas to be followed by the concessionaires over the term of the concessions. It is worth recalling that toll increases in the French concessions are limited to 70% of CPI after an initial period when tolls increases are pre-agreed. The toll setting formulas for the Skyway and Indiana Toll Road also have pre-established tolls for an initial period and then allow tolls to increase annually by the greater of 2%, change in CPI, or change in nominal GDP per capita.
The following graphs compare the actual toll schedule versus the CPI in the case of the Sanef concessions (Sanef and SAPN), and the Chicago Skyway and the ITR concessions. To make comparisons possible these graphs show toll rates only until 2028, which is the year when the Sanef concessions end.
(Insert graph 2 around here)
(Insert graph 3 around here)
(Insert graph 4 around here)
(Insert graph 5 around here)
From the graphs it is readily seen that the projected toll increases for the Sanef concession are smaller than the change in CPI. After 2009, when the 70% of CPI formula becomes applicable, there is a continuous decrease of tolls in real terms (adjusted for inflation). At the end of the concession, tolls have decreased by more than 10% in real terms. As a consequence, toll road users have won purchasing power (or an increase in consumers' surplus). In the case of the SAPN concession, real tolls exhibit a similar pattern, although it takes longer for consumers to win purchasing power, since the real decrease of tolls begins in 2014.
This situation is very different in the cases of Skyway and ITR. Graph 4 shows that actual tolls on the Skyway increase faster than CPI because of the pre-agreed increases. After 2017, when the pre-established toll schedule ends, the real increase of tolls continues because of the nominal GDP per capita factor of the formula.14 Graph 5 tells a similar story for the Indiana Toll Road.
Finally, Graph 6 shows the toll schedule throughout the full term (99 years) of the Chicago Skyway concession. The difference between the increase in tolls, based on the projection of the concessionaire and the increase that would result from using only CPI in the toll formula is significant. By the end of the concession term real tolls have more than doubled, assuming an average annual increase in CPI of 3%, and more than tripled assuming an annual average increase of 3.5%.
(Insert graph 6 here)
As previously observed, the concession model used by the City of Chicago and the Indiana Toll Road resulted in maximizing the concession price at the expense of consumers' surplus. In order to quantify the effect on consumers, we calculated the concession price for Skyway using the "French" toll regulation of 70% of CPI, while maintaining all of the Skyway concessionaire's other assumptions, for example 99 year term, Halcrow traffic projections. (See Case 3 in Table 5.) The total price paid would have been US$1.088 billion (or US$742 million less than the actual concession price). Alternatively, if we calculated the price using toll increases tied to 100% of CPI, the concession rent would have been US$1.330 billion (or US$500 million less than the actual concession price).
The US$742 million difference between the actual concession price and the hypothetical concession price resulting from the 70% of CPI toll regulation is made up of two pieces:
(a) US$500 million of lost of purchasing power for consumers (the effect of the difference between greater than CPI increases and increases limited to CPI); and
(b) US$242 million of opportunity cost resulting from consumers not sharing in the productivity/efficiency gains (the difference of tolls beginning indexed to 100% of CPI versus 70% of CPI).
Hence, consumers are losers in the Skyway (and Indiana) concessions, relative to consumers in France; the Skyway users not only lose purchasing power because tolls increase faster than CPI, but also do not share efficiency/productivity gains as is the case with the French concessions.
From the perspective of the region (the locale served by the road) and with respect to external effects:
(a) Pricing strategies (for example, toll setting, variable pricing, etc.) that optimize regional mobility (measured either in terms of aggregate traffic moved and/or in terms of level of regional congestion) are good.
(b) Tolls that reflect externalities and compensate the parties that pay for these externalities are good.
Both the Skyway and the Indiana Toll Road have strong monopolistic characteristics which give the concessionaires significant latitude to set tolls so as to maximize profits. There is no motivation for these concessionaires to set tolls to optimize regional mobility or to internalize external effects unless these activities enhance their bottom line. As a consequence, residents in the regions served by the road are likely losers from this concessions approach.15
___________________________________________________________________________________________
10 The definitions of the various surpluses are:
- Consumers' surplus is the difference between what a consumer is willing to pay for a good and the actual price of the good.
- Producer surplus is the difference between what a producer obtains from selling a good and the cost involved in producing it.
- Workers' surplus is the difference between what workers receive from working and the wage that workers would be willing to accept in order to keep the job.
- Taxpayers' surplus is the difference between the utility derived from public services and the costs implied by taxation to pay for the services.
11 We do not go deeper into the workers perspective, which relates to workers surplus. Given the relatively small work force associated with tollways (and therefore the small impact on the overall welfare equation), we believe we can safely disregard this variable.
12 The use of the proceeds is a factor in whether these sales are 'good public policy', but it is not a differentiator between the two approaches.
13 Theoretically, this taxpayer "windfall" is not a risk even if the price is not financially sustainable, but practically there are likely to be costs borne by the taxpayers in the case of a default by the concessionaire. Also, there is a situation in which a higher price and greater leverage could have a long-term detrimental effect on taxpayers; in the case of the Skyway, the City can terminate the concession early, but only upon payment to concessionaire of the fair market value of the concession but not less than the value of the debt outstanding. To the extent that concessionaire keeps the asset fully leveraged, this termination option will be expensive for taxpayers.
14 History of macroeconomic indicators in the US shows that increases of CPI are lower than changes of nominal GDP/capita. Table A-2 in the appendix displays the recent history of CPI and nominal GDP/capita in the US in the 10 years before Chicago Skyway was privatized.
15 It is worthwhile noting that traffic conditions in the corridor make this problem more important in the case of the Chicago Skyway concession than in the ITR concession.