Values in economics

A philosophical aside

I feel obliged to declare my biases up front. I’ve spent almost as much time studying moral philosophy as economic theory. Somewhat ironically, I decided to study economics after I had read Karl Marx’s selected works. If you want to understand social change, Marx remains a compelling starting point. I very quickly came to the same conclusion as the Polish political theorist, Zbigniew Pelczynski – that Marx’s greatest failings were his prescriptions, not his analysis.

In the simplest terms, Marx argues that the economic deployment of technology determines social structure. This is the fundamental organising principle of capitalist society. In this sense, Hayek and Marx occupy similar space. Hayek marvels at how order can emerge from a decentralised system which no individual controls. Marx analyses the structure of this order, and how it evolves. I was brought up by academic liberal christians in the 1970s and 1980s, I was interested in social reform, so I had to study economics. On the same journey, I also studied moral philosophy. Growing up in catholic Ireland, it was assumed that ethics is the domain of religion. I wanted to know if there could be a secular ethics. So I embarked on a journey studying economics and moral philosophy.

Marx is often seen as epitomising the view that intellectual objectivity is impossible. The crude version of this would see all ideas as reflecting ‘class interests’ or local, personal and cultural bias.

It is no coincidence that probably the most important moral philosopher of the last hundred years, Jurgen Habermas, came from the German Marxist tradition, and was obsessed with precisely this issue: how can we critically assess prevailing beliefs if objectivity is impossible?

Habermas achieves an extraordinary synthesis of this tradition with the best of analytical philosophy, integrating the work of the greatest Anglo-Saxon moral philosophers, mainly from Oxford, but also from the United States. Economists should be aware of his conclusions, because they provide clarity.

Habermas shows that in discussion we raise three distinct claims of legitimacy: claims of truth, truthfulness and rightness.

Claims pertaining to truth are essentially empirical claims. For example, questions such as ‘Do humans have bounded rationality?’ are questions which are either true or false. This doesn’t mean we have all the data to resolve the question, but the nature of the claim is simply true or false, (or probabilistically so). This is also the case with a question such as, ‘Is raising the minimal wage Pareto optimal?’ In this case we almost certainly don’t have the data to answer the question with certainty, whereas we know that rationality is bounded. But the claim has the same status. It is either true or false.

Claims pertaining to truthfulness are also of a distinct form. For example, it is reasonable to ask George Osborne: ‘Do you really believe austerity is needed to keep gilt yields from rising, or are you just saying that to undo Gordon Brown’s policies and neuter local governments?’ Empirical evidence can of course be marshalled to determine truthfulness, but the rebuke is not an empirical one – it is dishonesty.

The final type of claim raised in human discussions pertains to what is ethically right. This is in fact an assessment of actions, not thoughts. (Some people try to argue there are thoughts which are ethically wrong – but that doesn’t usually survive serious analysis).

An example of an economics question that raises the claim of ethical rightness is, ‘Should we raise the minimum wage?’ As with truthfulness, empirical evidence will be critical to that judgement, but it alone cannot answer that question. We might agree on all the evidence regarding minimum wages and still disagree over the right policy.

This is easier to see if we consider a clear cut example of a policy where there is an overwhelming consensus on the ethics, ‘Is it right for people to drive under the influence of alcohol?’ Empirical evidence was no doubt very helpful in encouraging changes to the law (although probably less so changes in social attitudes). Legislators would have received data on the numbers of people critically injured or killed by car accidents involving drunk drivers. Presumably advocates of greater degrees of leniency would have referred to rural areas where there is an absence of public transport etc. But the ethical judgement is not empirical. It is about deeming the value of life above the value of enjoying alcohol. For example, if one party in the debate says ‘I know what the evidence shows, but I don’t care about the interests of others’, it is clear to most humans that this argument is amoral. It is clearly outside the empirical domain.

How does this pertain to the issue of values in economics? Economists should drop phrases like ‘value-laden’ or ‘inherently-biased’. Theoretical and empirical issues in economics are questions of true or false – inherently, they are not ethical questions (how we choose to devote research resources may be, however). They should – and eventually will – be rejected by evidence. What we choose to do with any of the ideas – either as individuals or policy-makers – are not just empirical questions, but ethical questions, which are attempts to assess what is right or wrong. Ethical discussions are centred around the interests of those affected, and how to weight these interests fairly. Often, we transfer these discussions to impartial third parties to reduce bias.

Ownership bias

I should be very clear that this is not to say that individual economists are not biased. They are. Some are extremely biased, some come very close to the scientific ideal.

How do I know they are biased? In all science, there is an ownership bias. Individuals over-weight ideas which they believe they own. Physicists may the worst, here. But this is true in all areas of thought.

The other reason we know many economists are biased is that they declare it. For example, economists often make proclamations such as ‘I am a Keynesian’ or ‘I am a free-market economist’. Very often the same people claim to be scientists, which is hilarious really – particularly coming from advocates of rationality. I’ll forgive Samuelson for saying ‘I’m eclectic because Mother Nature is.’ But he should really have added the caveat, ‘at least I think she is on current evidence’.

The other way we know that some economists are extremely biased is that we can predict their ’empirical’ conclusions a priori. For example, if Robert Barro produces a new empirical paper on fiscal policy, I predict it will show limited, if any, impact. In this sense, I suspect there are areas of US economics, in particular, which are extremely biased. However, suggesting that Robert Barro is biased, which is a variant of raising the claim of truthfulness, does not alter the fact that his work is either true or false. The suggestion that it is biased may raise the probability it is false, but does not determine it. For example, I do not believe that real business cycle theory explains many recessions – and the evidence supports this – but the ideological preferences of its originators are neither here nor there. I also find RBC intellectually thought-provoking and perhaps even relevant in certain circumstances. In one sense, it is actually helpful if a researcher’s bias is worn on their sleeve. It is tantamount to saying ‘this is a thought-experiment which is almost certainly false’.

Intellectual progress doesn’t always happen as intended. Eugene Fama is now fond of claiming responsibility for the emergence of behavioural finance. Let’s be honest, his views on fiscal policy approximate parody. Alesina’s are also borderline. To some extent this is also true of real business cycle theory. In the same way that Friedman tried to argue that false assumptions are fine if the macro predictions work, one could argue that if the macro predictions fail, the assumptions must be false.

In conclusion, economists can take away several observations from moral philosophy. Whenever an economics discussion is occurring, participants are likely to be raising three distinct claims which should be not be conflated. Is a researcher honest or biased? Is the hypothesis true or false? Given everything we know, is the policy ethical? These are distinct questions, which should not be conflated. They are also frequently indeterminate.

(provoked by a series of brilliant posts by Simon Wren-Lewis and some great questioning by Beatrice Cherrier)