Cost and benefit
The New Statesman had an article (in August) criticising the New Approach to Appraisal (Nata) rules for road building in the UK. Basically, Nata is a cost-benefit analysis framework for assessing proposals of road building projects, introduced by Labour in 1998.
The author comes down hard against Nata. He’s right in that some of the estimates seem not very plausible, such as a barrel of oil to cost $35 in 2025.
Other examples are difficult to appraise given the limited information in the article:
Take section 3.5.1[1] of the Nata rules. This awards extra points to schemes that generate more traffic because more cars and lorries on the road mean more fuel sales – and hence more tax revenue for the government. By contrast, public transport schemes, which take motor vehicles off the road and so reduce fuel sales and tax revenue, have points deducted.
More tax revenue for the government is indeed a good thing, so on its own this rule is not necessarily a problem. But is there a rule that looks at the adverse effects of increased motor traffic? Noise, pollution, and risk emanating from motor traffic result in a loss of livable public space to everyone outside of cars (including drivers while they are not driving). Is there a rule that looks at the benefits of more public transport, such as decreased necessity for people to buy and maintain cars at great cost, decreased drunk driving, and so on? Tax revenue is not a goal in itself, it is a goal because the state can use it to fulfil other goals, that is buy things that are good for its citizens. A solution that fulfils these goals directly might well be preferable.
Sometimes the Nata rules neglect a potential change in behaviour altogether. For example, a proposal gets credit if it is predicted to shorten journey times:
Every minute saved for a car driver is valued at 44p [...]
[...]
A minute saved on a cyclist’s travel time, for example, isn’t worth 44p but just 28p. A bus-user’s time is valued at 33p a minute. The implicit assumption is that cyclists and bus-users make less contribution to the economy than car drivers.
Again, on its own this assumption may well be justified. It is plausible that currently, on average, cyclists are in a lower income bracket than car drivers, and if you value economic contribution in terms of revenue, the time of someone who earns more is more precious. The first problem is of course that the money generated in an economy isn’t everything. Some jobs are widely recognised to be underpaid compared to what they are actually worth to society (caring for infirm patients, or raising children). But in addition, the assessment seems to be based entirely on the status quo, without any regard for the possibility of change. However, as more people take up cycling, the difference in contribution to the economy between drivers and cyclists is likely to shrink. And of course, how many people take up cycling is in no small measure influenced by decisions about road building proposals.
Cost-benefit analysis is especially controversial when monetary value has to be assigned to goods that are not usually thought of as being part of the economy, such as clean air or the beauty of a landscape. Have a look at this excerpt from the article:
Nata assessments have also always avoided costing the most damaging aspects of new roads, such as the impact on landscapes, noise and pollution and, of course, carbon emissions. This means that, even though a road might be an environmental disaster, there are no estimated cash costs to be set against the claimed economic benefits. Instead the planners give a qualitative assessment, using terms such as “moderate”, “severe” or, worst of all, “very large adverse”.
These qualitative judgements have in the past been enough to frustrate the road builders. Alistair Darling rejected “improvements” to the A303 that would have carved a new road through the Blackdown Hills, an area of outstanding natural beauty on the Devon-Somerset borders, after seeing in the Nata assessment that it would have a “seriously adverse” impact. He apparently did not want to be the minister who overruled such a negative assessment.
It should be no surprise that ever since that decision was taken, Treasury and DfT [Department for Transport] officials have been working to get rid of such “emotional” analyses by designing a system to assign monetary values to landscapes, tranquillity and biodiversity. The ostensible aim is to make the system more “objective” and number-based – but the crucial issue is what values are assigned to qualities that are inherently priceless. How much might the last dormouse in Wiltshire be worth? Soon Labour’s minions may be able to tell you.
Note how in the first paragraph, the author seems to criticise lack of inclusion of “non-monetary” costs in the formal part of the Nata assessment, while by the last paragraph, he casts precisely that inclusion in a negative light.
Many people have a visceral reaction against a formal market valuation of some goods, goods that are not usually traded in a market. Of course, the problem is, what is the alternative? Personally, I’m not sure that a purely qualitative judgment is necessarily better. In the example of the A303 it resulted in a decision after the author’s taste, but that is no guarantee that it will do so in the future. What if some planners categorise environmental damage as “negligible” where most residents would disagree?
Of course, the devil is in the detail. In CBA, value is usually assigned to any good, even clean air or natural beauty, by asking people about their willingness to pay (WTP) for the good. This can be problematic, for example if the value elicited from a poor person and that elicited from a rich person get compared directly, in currency terms. A poor person stating that they would pay £ 10 and a rich person stating that they would pay £ 100 can well be manifestations of the same underlying emotional value held by each of them for the good in question — should the rich person’s value enter the equation as 10 times greater than the poor person’s? However, there are alternatives. WTP can be measured as percentage of income, for example, or currency units can be weighted.
In general, I think it is better to try to make an explicit quantitative assessment of such values than to rely on the intuition of a small panel. However, some of people’s intuition against such quantitative assessments may reflect that the quantitative process, while more open and objective to experts, is less accessible and open to criticism by laymen than a one-off, verbal assessment such as “considerable damage” or “negligible”. The latter is immediately visible and understandable, and can be openly criticised by anyone using everyday language. A formal WTP assessment will be shrouded in a report, disguised in numbers, and resting on many implicit assumptions and hidden decisions about methodology (such as WTP in currency vs percentage of income), and understanding and criticising it would require research and more formal expression of criticism. What would such a perceived (or real) difference in accessibility to criticism mean? I’d argue that translating it into a rejection of a formal quantitative method would be shortsighted, and that we could instead strive to make the quantitative process more rigorous and more easily accessible even to laypeople.
But even if we elicit the value with perfectly sound methodology, many people will remain uneasy because they think that some values are absolute and should not be traded off under any circumstances. Of course, this is bogus. If extinction of the last Wiltshire dormouse would mean revenue for the treasury of billions of pounds, and the NHS could use that money to find a cancer vaccine, wouldn’t that have been worth the money? When people refuse to trade off some values (nature, beauty) against money, they may neglect that that money could fulfil other goals of the same emotional and “moral” significance.
