For a while I have been working on a paper on democracy, expert knowledge, and economics as a moral science. [The financial crisis plays a role in the motivation of the paper, but the arguments I’m advancing turn out to be only contingently related to the crisis]. One thing I argue is that, given its direct and indirect influence on policy making and for reasons of democratic accountability, economics should become much more aware of the values it (implicitly or explicitly) endorses. Those values are embedded in some of the basis concepts used but also in some of the assumptions in the theory-building.
The textbook example in the philosophy of economics literature to illustrate the insufficiently acknowledged value-ladenness of economics is the notion of Pareto efficiency, also known as ‘the Pareto criterion’. Yet time and time again (for me most recently two days ago at a seminar in Oxford) I encounter economists (scholars or students) who fail to see why endorsing Pareto efficiency is not value-neutral, or why there are good reasons why one would not endorse the Pareto-criterion. Here’s an example in print of a very influential economist: Gregory Mankiw.
In his infamous paper ‘Defending the One Percent’ Mankiw writes (p. 22):
“Discussion of inequality necessarily involves our social and political values, but if inequality also entails inefficiency, those normative judgements are more easily agreed upon. The Pareto-criterion is the clearest case: if we can make some people better off without making anyone worse off, who could possibly object?”
Yet the Pareto-criterion is not as uncontroversial as Mankiw believes. The Pareto-criterion compares two social states, A and B, and makes a claim about whether the act/policy/social change that brings us from A to B is desirable or not. If in B all individuals have at least the same welfare/utility/wellbeing than in A, and at least one of them has a higher level, then moving from A to B is a Pareto-improvement, and the Pareto-criterion recommends the move from A to B on grounds of efficiency.
As many have argued, the Pareto-criterion remains agnostic about the fairness or legitimacy of A as the starting point. The Pareto-criterion also doesn’t attach any importance to distributive issues in either A or B. If those who are living in misery stay equally miserable, but due to some policy or social change the ultra-rich become even more rich, then Mankiw believes that we would all agree that this social change is a social improvement. Given that many citizens (and political philosophers) believe that certain types of inequality are intrinsically bad, I don’t think that we can draw Mankiw’s conclusion.
Moreover, even if we would take the Pareto-criterion as a criterion of efficiency understood as ‘no-waste’ (in the sense of: ‘let’s get a bigger pie if we can, independent of how the pie is distributed’), then the question still remains ‘efficiency of what’? It’s entirely plausible that an economic change that is Pareto-efficient in terms of desire-satisfaction is not Pareto-efficient in terms of a set of basic capabilities, or in terms of equivalised household income. So even if one doesn’t attach any value to distributional issues, our endorsement of the Pareto-criterion may be dependent on the “metric” in which it is expressed.
I’m using this example of the Pareto-criterion merely to support the claim that economists need to think more about the values embedded in their theories – a claim on which any philosophy of economics textbook can give us many more examples and more detail.
Why is it relevant now? In the lively discussion on what kind of science (or something else) economics is which is currently raging on the blogs, we should also consider the view of those who have argued that economics is a moral science. This, in Tony Atkinson’s words (2011. p. 157) means that “Economists need to be more explicit about the relation between the welfare criteria and the objectives of government, policymakers and individual citizens”. Atkinson traces the expression back to Keynes, who had written in a letter that ‘economics is essentially a moral science’. More recent defenders of that view include Kenneth Boulding in his 1968 AEA presidential address, who defended the strong view that economics inherently depends on the acceptance of some values, and thus inherently has an ethical component. And surely Amartya Sen should be listed here too (although I cannot recall having read in his work this precise expression). And also Robert and Virigina Shiller in a 2011 paper in the AER proceedings, who stress that Boulding is asserting that one must understand the ‘human condition’ in order to pursue economics as a moral science. If understanding ‘the human condition’ becomes part of economics, then that surely has implications for the status of inductive reasoning and other non-mainstream methods.
Much has been said in recent weeks about what (if anything) is wrong with contemporary mainstream economics, and what changes are needed. Over at New Apps, Eric Schliesser (post 1, post 2) also pointed to the importance of values in understanding what is wrong with contemporary economics. Schliesser provides a very thoughtful response to Raj Chetty’s piece that is well worth the time of anyone who thought Chetty was right (since, as Schliesser convincingly argues, that’s not the case).
My view is this: economics shouldn’t aspire to be a value-free science, but an intellectual enterprise that combines elements from the sciences with elements from ‘the arts’ done in a manner that makes it value-commitments explicit. Values in economics have many sources. There are values involved in the choice of questions that are asked (and not asked). Value judgements are embedded in the normative principles (such as the Pareto-criterion) that are endorsed. Value judgments flow from the choices in how basic categories and notions are conceptualized (is ‘labour’ only what we do for pay, or also what we do to reproduce the human species?). And value-judgements may be linked to methodological choices that the economist makes, since the methodologies can restrict what one can observe and understand (for example: if you only take into account those things that can be measured, then it follows that non-measurable entities will receive zero-weighs in any evaluation).
I advocate that in as far as economics is not value-free, the valueladenness should be made explicit and economists should be trained to detect the value commitments in their theorizing and conceptual work. My interactions with economists suggest that this is a skill many (most?) of them do not possess, so it is a change needed in the core economics curriculum. Presumably, we can’t turn all economists into scholars well-trained in normative analysis. But there are alternatives possible – for example, economists should collaborate or interact more with scholars trained in normative analysis (I’m thinking primarily of economic ethics, but not necessarily only them), to get professional advice on the value-commitments of their work.
By the way, if there must be a Nobel Prize in Economics, it would be a good idea to give it to Tony Atkinson next year. Not only does he deserve it for his pathbreaking work on inequality measures and other contributions to welfare economics since the 1970s, but it would also help to restore the position of welfare economics in the economics mainstream. Welfare economics could be considered one of the bridges between economics and moral philosophy (economic ethics and normative political philosophy on economics topics being two other such bridges), and therefore deserves and needs more support and attention.