Cost-benefit analysis and its limits

What is CBA? The case of a public investment

Economists have developed methods for assessing the advantages (benefits) and disadvantages (costs) of a given investment, attempting to evaluate them in monetary terms and calculate a “socio-economic return”. The idea is to mimic the private calculation of the profitability of an investment: the financial director calculates the cost of an investment, the future operating expenses, the revenues generated and draws up a financial balance sheet, which can be expressed synthetically by the Net Present Value (NPV ) indicator.⁷. Sometimes, however, he is satisfied with a “rate of return on investment”, expressed in simplified terms as a number of years: in how many years are the initial expenses “reimbursed” by future net earnings?

These are the concepts used in CBA, bearing in mind that the advantages and disadvantages to be compared (and monetized for NPV purposes) are not all “marketable”, i.e. are not necessarily the subject of monetary transactions. As a result, the “values” to be used are not derived from observed prices, but from more indirect methods. We’ll come back to this in section 4.3 of this fact sheet.

N.B.: CBA methodological guides also enable risk analyses to be carried out, which we won’t go into here.

The benefits of CBA

CBA is part of a long-standing French tradition of “rationalizing” public decisions through economic calculations. The engineer Jules Dupuit is credited with laying the foundations for these calculations, which he introduced in 1844 to provide a “measure of the utility of public works”.⁸ Since then, numerous academic and administrative studies have been carried out, which we won’t go into here. French law requires the use of a socio-economic assessment for civil public investments.⁹ For several decades, France Stratégie (now the Haut-commissariat à la Stratégie et au Plan) has been updating the administration’s doctrine and the guides to be used.

CBA has three major benefits, which we develop in 2.1, 2.2 and 2.3.

The CBA encourages the use of opposable, verifiable ex-post costings.

These figures limit the arbitrariness of political decisions, which can be based on whim or fantasy, or even corruption (passive or active) or clientelism.¹⁰. Without going that far, public projects may not correspond to clearly identified social needs, or may meet them in a questionable way. CBA makes it possible to objectivize the reasons for carrying them out. In dictatorial regimes, examples abound of public funds being used for the benefit of oligarchs. In democratic regimes, however, examples of “white elephants” that fail to meet socially and/or economically justified needs abound in all countries. Examples include airports in Spain, a railroad line in Greece, and an urban development project in the Gonesse triangle in France. The case of the Notre-Dame-des-Landes airport is emblematic of the value of the objectivation approach (see box).

CBA provides a framework for negotiation between project stakeholders

CBA requires that the benefits and costs of a project be made as explicit and objective as possible, even before they are monetized.

While there may be abuses on the part of political power, criticism of a project may be for individual reasons more or less disguised as general interest. A freeway or wind turbine project close to a dwelling obviously has a negative impact on the inhabitants. It may even lead to expropriation. In such cases, the expropriation procedure becomes more cumbersome, and the prefect must declare the project to be in the public interest (DUP) on the basis of a public inquiry. The CBA allows us to take a step back and look at the project from the point of view of the general interest, and by setting out the interests of the project (and its possible variants, see next point) it can help stakeholders to understand the ins and outs of the project. This analytical work is very useful: if it is not done or not understood, criticisms can turn into attacks or even violence(see the case of the A69 in the box).

Finally, the CBA encourages us to think of alternatives

CBA is in fact quite relevant for comparing alternatives. Revealing them can sometimes lead to the abandonment of a costly project for a less costly one with similar benefits. The Nantes airport conflict is a case in point. It’s worth noting that the main alternative against which the project is compared is the counterfactual, i.e. the situation without the project. But as Luc Baumstark and Jean-Paul Nicolaï point out¹⁴ “the calculations envisaged cannot avoid the need for a real foresight exercise, as this reference situation is, most of the time, quite far from a business-as-usual situation.”

In conclusion, we can say that CBA responds to a very legitimate social demand and that it is desirable that it not only be produced but, for large projects, discussed in depth with the stakeholders, which in practice is not the case. So it’s not the logic of the socio-economic balance sheet that poses a problem: we’ll see that it’s the monetization that poses major conceptual and practical problems. We will also insist on the fact that CBA has no theoretical justification when it is not a question of comparing variants (or marginal differences between two situations or projects). The case of climate policies will illustrate this.

The conceptual limits of CBA

What costs are we talking about?

CBA speaks to common sense, which spontaneously understands its objectives: reducing costs for equivalent benefits, or increasing benefits for equivalent costs.

However, the notion of cost can cover completely different notions, the addition of which (via the monetary unit) poses fundamental problems. We refer you to The Other Economy’s article on the concept of costs in economics. In short, there are two notions of cost:

• cost seen as a loss of satisfaction (economists call it “disutility”);
• cost as a drain on a limited resource.

In addition, any cost calculation must specify who bears the costs:

• a person or legal entity (company, association, public administration);
• a community (the nation, for example).

The result is a summary table for easy reference.

When applied to CBA, these distinctions should help avoid confusion. Are CBA costs costs for the State (or a regional authority)? Project investment costs? Cumulative discounted costs (including future operating expenses)?

Can we legitimately use economic calculations to inform decisions that are fundamentally ethical or political?

Beyond the initial criticisms of the notion of costs, there is a deeper criticism to be made of CBA. By explicitly putting a price on human life or nature, the economist gives the impression that he is outside the moral or ethical field, and is making a rational analysis. On the contrary, he is in fact pulling off an ethical and philosophical coup de force. Let’s take the example of human life. CBA is used in many fields, for example in healthcare¹⁵. The COVID crisis was the occasion for debate on this issue. Economist Christian Gollier¹⁶ with his colleague James Hammitt, for example, wrote¹⁷ Because we are not prepared to sacrifice everything to increase our life expectancy, this means that our life has a value, and that this value is finite”. The same economist used CBA to economically evaluate the containment policy during the COVID crisis¹⁸.

Admittedly, our individual and collective resources are limited, which means we have to make trade-offs between different types of expenditure, including reducing road deaths or increasing our medical capabilities. This does not mean, however, that we place a finite value on life or health, and that this value is equal to the expenditure considered.¹⁹.

As Jean-Jacques Rousseau wrote:

The politicians of old spoke incessantly of morals and virtue; our politicians speak only of trade and money. One will tell you that a man is worth as much in such and such a country as he would be sold for in Algiers; another, following this calculation, will find countries where a man is worth nothing, and others where he would be worth less than nothing. They value men like herds of cattle. According to them, a man is only worth as much to the state as he consumes in it.”²⁰

The value we place on things and people has sentimental, emotional, philosophical, anthropological and even religious components that cannot be summed up in monetary terms. We may put an end to our lives following the death of a loved one, or go into a deep depression; none of this has anything to do with the money we’re willing to pay to reduce the risk of death linked to an infrastructure or public policy. As Aurore Lalucq and Jean Gadrey say in their bookFaut-il donner un prix à la nature ? Nature has no price, but protecting it has a cost”. When it comes to human life and health, Jean Gadrey also writes: “Human life and health have no price, but taking care of them has a cost.”²¹

If we accept the inescapable idea that we must provide ourselves with the means to limit the collective expenditure in question, we won’t be able to do so by resorting to such theoretical artifice in a doctoral fashion. In democracies, these deeply debatable calculations cannot replace collective deliberation, the outcome of which is not the expression of a social value made explicit by the level of expenditure decided upon, but rather, more modestly, the expenditure we are prepared to make.

The operational limits of CBA

Can we put a price on non-market quantities?

Let’s take the case of a freeway project that would divert part of the traffic from a trunk road. The expected benefits are fewer injuries and fatalities, time saved for users, decongestion of towns (when the freeway allows bypassing). These benefits are not directly monetary. The costs of the freeway are, in part: initial investment, annual operating costs and maintenance work. It also implicitly creates transfers of “land rents” (land near freeway exits increases in value, while land bordering the freeway loses value). But some of these costs are also non-monetary: the freeway artificializes the soil, encourages traffic and therefore_CO2 emissions, and is a source of noise for local residents.

To draw up a “socio-economic balance sheet” for this freeway project, we need to assign a monetary value to the non-market benefits and costs. This is what economists do. In France, official guides exist²² giving a price to a death avoided, to an hour of time saved, to a tonne of_{CO2 avoided} …

The first difficulty, mentioned by economist Nicolas Bouleau, is “to attribute prices to quantities that are not marketed, in particular, but not exclusively, with regard to environmental modifications.²³. Economists have proposed several methods within the framework of “neo-classical welfare economics”²⁴to evaluate intangible impacts and non-market goods. The development of environmental sciences and the sociology of public decision-making has made it clear that decisions relating to urban areas are so complex that these methods are of little use, and also that the disturbance of natural sites and water management require much more elaborate decision-making procedures.”

Nicolas Bouleau’s criticism is expressed above in terms of feasibility and reliability; it is the fragility of price estimates for non-market goods and services that he questions.

The second difficulty is more fundamental. Can we put a price on nature? We refer you to the fact sheet Should we put a price on nature, which explores this fundamental question.

The economically intractable problem of choosing the right discount rate

Comparing the costs and benefits of a project is usually an intertemporal process: capital expenditure is incurred at the outset; operating expenditure and other costs and benefits are incurred throughout the life of the project, right up to its completion, and any costs of decommissioning and environmental restoration. To compare flows that do not occur at the same time, economists use a discount rate. This concept is presented in detail in the “Discount rate” fact sheet; we’ll say just a word about it here. The value chosen for this rate is decisive for the result of the calculation, and its choice is conventional. This convention cannot be based on purely economic reasoning: it necessarily involves ethical and political considerations.

CBA is only legitimate for comparisons at the margin of a reference situation.

The final difficulty is that CBA is adapted to comparisons between projects or project variants, in relation to an overall situation that is considered unchanged by the project, and therefore seen as marginal in relation to this overall situation. What is evaluated are marginal differences. Take, for example, the Montagne d’Or gold mining project in western French Guiana. ²⁸

As economist Claude Henry writes²⁹ :

Back in 2010, The Economist magazine, whose main concerns are not ecological, warned: “If the Amazon rainforest were to go up in smoke – a scenario that is not inconceivable with a little more clearing and a little more warming – it would spew into the atmosphere the equivalent of ten years’ worth of emissions resulting from the burning of fossil fuels.” A march to the abyss if ever there was one: it would tip us into a hostile climate without return. Considered in isolation, the Montagne d’Or project does not appear to be truly dangerous in this respect. But it makes no sense to consider it in isolation. It would be just one link in a chain of destruction running through the entire Amazon rainforest: mines, dams, legal or illegal clearing to make way for soya or cattle farming, and to appropriate the timber, arson attacks to facilitate some of these operations. The Montagne d’Or operation – on French soil and with the blessing of the French government – would provide invaluable support for all these destructive ventures and for those responsible for them. Yet the stakes involved in conserving the Amazon rainforest are such that it is now a categorical imperative (Immanuel Kant, Critique of Practical Reason). A categorical imperative does not authorize the usual political compromises, it is incompatible with the “cost-benefit” methods of evaluation customary in economic studies, nothing that violates it can be held to be marginal.”

CBA applied to public policy: the case of climate change

(See also : Global warming: what impact on growth?)

Nobel Prize-winner William Nordhaus was instrumental in perfecting this approach.³⁰

The basic idea is to draw up a socio-economic balance sheet of the disadvantages and advantages of a policy to combat climate change. There are several ways of doing this. W. Nordhaus has drawn up economic balances for a series of warming trajectories, and deduced the economically optimal trajectory (the one with the best balance).

The economist considers that benefits and costs are measured in GDP points.³¹ The costs of combating climate change have an impact on GDP, but so does global warming. A macroeconomic model is therefore needed to make these assessments. W. Nordhaus developed the DICE model³² – available online – with this in mind.

The costs of global warming are assessed on the basis of a damage function, which links climate change to production losses, and therefore to GDP losses. With this logic in mind, it is possible to perform an optimization calculation to determine the optimal trajectory, the one with the best economic balance. To make this balance, which is intertemporal (costs and benefits are distributed over a long period), we need to use a discount rate(see discount rate data sheet) which compares a gain or a cost today and for each future year.

Two outputs can be derived from these calculations:

• the trajectory that optimizes the socio-economic balance ;
• the social cost of carbon corresponding to this optimum (see box below).

What do these works conclude? Antonin Pottier³⁶ In the arguments of the juries for the Bank of Sweden prize, we find a graph showing the “optimal” trajectory of_CO2 emissions according to Nordhaus, a trajectory that goes from 35 Gt_CO2 per year in 2015 to around 15 Gt_CO2 at the end of the century. It takes a little research to realize that Nordhaus’ optimal solution leads to a warming of… 3.5°C in 2100. Hardly a slight change from the 4°C of the no-action scenario!

Nordhaus’ work has been severely criticized, notably by economist Steve Keen³⁷. To sum up, the main problems with this type of exercise are as follows:

• Expressing the outcome of a policy in GDP points (or in utility derived from consumption) is ethically questionable. The underlying logic of the neoclassical school of economics thus poses a fundamental problem: it is deeply questionable to make the maximization of the “utility” (see The Other Economy ‘s fact sheet on utility) that consumers derive from their incomes the basis of a policy.³⁸ as the criterion for social well-being.
• we don’t actually know how to assess the socio-economic consequences of climate change.

The effects of climate change are very heterogeneously distributed, both geographically and socially, and it is impossible to reason as if the losses of some offset the possible benefits of others in the absence of redistribution mechanisms. Moreover, they are “non-linear”, and tipping points can occur.³⁹. More pragmatically, past observations on the costs of climate change cannot be used as a basis for assessing what they will be in the future, due to this non-linearity.

• inter-temporal comparisons are made using a de facto arbitrary discount rate⁴⁰ as soon as it is positive, it weakens the weight of long-term costs and benefits. This has the effect of minimizing the effects of climate change, and raises major ethical issues (the relative weight of present and future generations in the decision) that cannot be summed up in a single figure.
• macroeconomic models have not proven their ability to represent the economy and its evolution, especially over the long term. In particular, GDP growth in these models is determined to first order by exogenous labor productivity growth.

All in all, it’s clearly a sham. Mathematical formalism and the words of physical science are being used in a field where experimentation is either impossible or non-existent, while camouflaging eminently questionable ethical choices.

Conclusion

In conclusion, CBA is useful in well-defined cases. It enables us to objectivize the terms of the debates necessary in a democracy, to study projects and especially variants between projects whose socio-economic impact is limited. By highlighting the positive and negative consequences of a project for stakeholders, the terms of these debates can be clearly defined. Their monetization, in the case of impacts on Nature, is far more debatable, as is the use of a single aggregate indicator.

We need new tools for reasoning in the face of the climate and ecological crisis we are experiencing. If we need to limit our public spending, because trees don’t grow on trees, we can’t do it by drawing up accounting balance sheets based on conventions, however elaborate they may be.

Shouldn’t we therefore adopt stricter rules based on the carbon budgets that the French government has defined for the coming years? Should we not simply abandon projects that do not reduce the GHG emissions of the activities concerned, or whose impacts on biodiversity are clearly out of proportion?
We need new methods which, without denying the importance of budgetary trade-offs, make them conditional on consideration of non-negotiable planetary limits. This is the case with so-called “robust” approaches. ⁴⁷