Friday, 16 September 2011

Economics for ethics

Ethics and economics have a troubled relationship. The public is generally under the impression that ethics is about being nice or fair to other people, while economics is about the machinery of translating individual selfishness into general wealth. One should not ask what each can say to the other, but which one we should choose.

Strangely enough this is also approximately how most ethicists and economists think about the relation between their disciplines, as a result of a tacit agreement to perpetuate mutual ignorance and antipathy. Ethicists think economists are clumsy buffoons with an impoverished view of human nature and morality, obsessed with incentives and markets as the answer to everything. Economists think ethicists are obsessed with discovering mystical intrinsic values, at the expense of systematically thinking through their real world relevance. These are caricatures with some truth to them. But to the extent that they prevent ethicists and economists from taking each other seriously, they block the real scope for mutual learning.

Although the learning can and should go both ways, it has generally been a one-way road. Ethicists have been reprimanding economists about their lack of ethical perspective since even before economics was founded as an independent science. They have rightly criticised economics for evading its responsibility for explicit ethical reasoning in favour of a scientistic identity of strict neutrality; its impoverished and flawed normative theories (rational choice for individuals; welfare economics for society); and the arrogance with which economists such as Gary Becker and Steven Levitt have promulgated a narrow methodology to understanding 'all we need to know' about the human condition.

The trouble is that because ethicists have so many reasonable complaints about economists they rarely take what economists do and how they think seriously. So this post will follow a path less travelled and ask what economics can do for ethics, and without ever talking about markets. First, ethicists should recognise that a great deal of economics is concerned with ethical issues, often ones surprisingly neglected in mainstream ethics. Second, theoretical economists have developed a sophisticated approach to ethics - a form of logistical consequentialism - together with specialised analytical tools that can be directly employed in mainstream ethics. 

Economics is about ethics

Understood properly economics is an ethical science, an important branch of applied moral philosophy. For it concerns how to understand, manage and fulfil the heterogeneous and often conflicting values, interests, and capacities of large numbers of individuals operating within the constraints of limited resources in a particular community. That system-level attention to the key aspects of heterogeneity, conflict, and scarcity within a community should be a central concern of moral philosophy, but it generally isn't (except for some political philosophers). 

Most moral philosophy focuses instead on different levels: meta-ethics (the nature and meaning of ethics); normative ethics (articulating theories of the right and the good, like utilitarianism or deontology); and applied ethics (the rights and wrongs of controversial issues like abortion and animal rights). Society and its complications is generally absent. This lacuna frequently leads moral philosophers to wish away or parenthesise (idealise) those aspects of heterogeneity, conflict and scarcity unavoidable in a moral community.

Economists as social scientists have been working on this area for a very long time, and have much of importance to say about it. For example, economists have addressed the centrality of scarcity by developing the concept of opportunity cost. Given scarce means you will not be able to do everything you want. For example, you may not be able to give $500 to the Somalia famine appeal and also take your long planned trip to visit your ailing grandmother. The cost of choosing one rather than the other is not the money. Money is not important: it merely represents your purchasing power constraints and only an idiot would try to acquire money for its own sake [previously]. Rather, the cost of choosing charity over filial piety is the full value (including the full ethical value) of the alternative you thereby give up. Moral philosophy rarely considers the full costs of actions in this broad sense and is the poorer for that. 

The economist's understanding of values in terms of choice under constraints also leads to another interesting conceptual innovation from which moral philosophy can learn. Economists present individuals as having not simply values, but preference relations between those values. In other words, people are assumed to evaluate not only whether things are valuable or not, but also to work out how valuable they are in relation to other valuable things, to work out 'better than' comparative relationships. 

This places the individual in a strong position to make choices in a rational way, in those rather common circumstances when we cannot have everything we would like. In contrast, much of moral philosophy is concerned with merely identifying sources of intrinsic value, for example of nature or communitarian identity. That leaves out entirely the question of how to relate these valuable objects to other important values, such as fairness, autonomy, well-being, and so on. Moral philosophers here assume an investigative identity (in what does value consist?) that seems to come at the expense of practical relevance (what's the right thing to do?).

Economics can provide logistical support to ethics

Economics' particular contribution to practical ethics is a sophisticated form of consequentialism concerned with the systemic mapping of both the implications of our ethical claims and the unintended but foreseeable consequences of our actions. It does this by 'modelling' what might happen to the things we are concerned about under various counterfactual conditions.

Deontological rigour

Economic modelling allows us to test the robustness of particular ethical principles as principles of action and thereby to clarify our practical moral reasoning. The deductive form takes axioms we believe to be plausible and develops tight formal arguments ('proofs') that follow from them using mathematical techniques economists have honed over the past hundred years. Such analytical consequentialism can be used generally to identify and clarify the ambiguities and difficulties of moral reasoning whenever the claims can be formulated in a formal (mathematical) way.

For instance, social choice theory is a sub-discipline of economics concerned with aggregating the interests (or preferences, or judgements, or views) of different persons (or groups) in a particular society. i.e. it is concerned with heterogeneity and conflict. It works by showing the logical implications of the combination of various apparently plausible ethical premises ('axioms'). Sometimes the analysis shows that the premises are inconsistent in ways that were not perceptible by the standard philosophical means of testing them against our intuitions and common sense (as in Kenneth Arrow's famous (im)possibility theorem). Sometimes it shows that premises we found agreeable lead unavoidably (by deduction) to unpalatable conclusions (as in Amartya Sen's Impossibility of a Paretian Liberal). Thereby we are brought to reconsider the plausibility and/or ethical status of our initial ideas.

Better utilitarianism

Economics is not only good at analysing the logical implications of concepts and values. It is also good at quantitative logistical analysis, at figuring out what is likely to happen if we try this or that well-intentioned idea. Economists model the interaction of significant causal mechanisms to map the likely long-term extended consequences of changes in agent behaviour or government regulation in a complex but closed system. By doing so economists can identify and assess the unintended but predictable consequences of a sensible seeming proposal. 

In a closed system seemingly good things can have bad consequences. Keynes famously identified the 'paradox of thrift': that the aggregate result of individuals acting sensibly (even morally virtuously) can be catastrophic for the society as a whole. On the other hand, an aggregate improvement can be catastrophic for some individuals, as when an economic boom causes price increases which leave those who didn't benefit from the boom much worse off than they were before. 

Economic modelling can be used to map the distributive effects of such economic events, or government policies, and thus to identify vulnerable individuals and socio-economic groups in advance, allowing a society that cares about such things to intervene. For example, carbon taxes may be effective, but they are also unfair because they have a particularly big impact on the purchasing power of the poor. We shouldn't sacrifice the poor to achieve our carbon goals, important as they are. Economics can identify ways to compensate the poor so that we can have our cake and eat it too. More generally, such methods can be used to map the impact of events, laws and policies on the opportunities and circumstances of apparently uninvolved parties, as when considering laws against child labour or pollution controls.

Sometimes bad things may have good consequences. This is the sneaky Mandevillian side of economics, the analysis of how private vices can generate public virtues. For example, Kaushik Basu (chief economist to the Indian Government) has proposed legalising paying harassment bribes (when officials demand money for doing their job, such as issuing identity documents or schooling certificates). People have little choice about paying these, but they really don't like it. If they could get what they needed and then report the crime, many of them would do so. Knowing this raises the risks (cost) to officials of demanding a bribe in any particular situation. So, counter-intuitively, legalising paying bribes may be expected to reduce levels of corruption.

Economists take human agency seriously, if at times narrowly. This leads them to a rather different perspective on politics and policy than most political philosophers. For political philosophy the problem of government concerns making and scrutinising proposals for just rules and institutions. Once these are found it is assumed that everyone in society will recognise their justice and comply with them in the spirit they were intended. (Rawls, for example, generally assumes full compliance with his social contract, though, under the influence of economists, he also included the 'difference principle' to motivate more productive people to work harder.) 

In contrast, economists are concerned with the actual effectiveness of rules and institutions, and that requires incorporating how they interact with the opportunities and interests of wily human agents. Their view of human nature may be shallow and distorted, yet it may still be more realistic in these circumstances than the homo justicius employed by political philosophers. For instance, economic modelling can be used to test whether the intended form of a regulation provides agents with financial incentives to act in ways that undermine the intentions of the policymakers (such as the service-fee basis of doctors' earnings in the US; or disability welfare payments). And it can be used generally to explore whether a 'just' rule is likely to be successful in getting people to act as they are supposed to. For instance, drug bans in sports; no fault divorce laws and women's equality; corporate social responsibility codes; or changing voting rules from first past the post to the 'more democratic' proportional representation. Economists try to work with people as they are, not as how they ought to be.

Finally, quantitative logistics can introduce a sense of proportion, whether about the benefits of intellectual property protection (contested) or the benefits to the world's poor of removing all immigration barriers (incredible). That can helpfully inform and focus our public deliberations as a society, as well as the work of professional academic ethicists, towards issues that really matter.

Interestingly, although many economics methods are descended from utilitarian moral philosophy, they can make contemporary utilitarians look rather shallow. Take Peter Singer's famous moral parable of the shallow pond.* The argument goes like this: You're walking by a pond and see a child drowning. What do you do? Jump in and save him of course! Ending world poverty is like that. 

Gosh, is world poverty really that simple to solve? If only we had known. The economist would ask for rather more information and logistical analysis before jumping into Singer's shallow pond of good intentions. Economists get called boring, cynical, and even cold-hearted for taking this approach, but it seems a more consistently utilitarian position to take. Utilitarians, after all, are supposed to maximise the good they actually achieve, not their good intentions. For utilitarians, inefficiency is a vice. To be a good - efficient - utilitarian you need to understand and employ the best logistical techniques, and the place to get them is economics.

Perspicuous contrasts

Economists are shy about introducing explicit moral arguments of their own and much prefer to analyse the value systems they can derive from their observations. But that doesn't mean that their analysis of the values we reveal in our actions has no moral bite. Without promoting any values of their own economists can confront us with the moral implications of our own actions in a way that can be extremely discomforting (no one likes to have their complacencies and hypocrisies exposed).

The disciplined rigour of cost-benefit analysis can be used to reveal striking discrepancies in what we are prepared to pay to save a life in one context (e.g. air travel or terrorism) rather than another (e.g. driving accidents or alcohol related illnesses). What we pay doesn't have to be money. For example, we can apply  the same method to state surveillance. If we wouldn't give up privacy rights to reduce minor risks of death in other contexts (like installing government cameras in every bathroom to save people from bathtub slips), what rational reason do we have for giving up all our privacy to the government to reduce the risk of terrorism from almost nothing to possibly slightly less? This kind of perspicuous contrast reveals the vacuity of the popular political argument that the loss of our privacy is a small price to pay for preventing terrorism and saving lives.

Elsewhere, health economists have shown that the main determinants of bad health are factors like inequality, poverty, social exclusion and so on. Not especially a lack of health care. So if we as a society really value good health, we should reconsider why we direct so much government spending and subsidies to health care interventions chasing incremental increases in longevity for terminally sick people when spending that money on poverty alleviation and preventative interventions would be so much more efficient. Of course we may have other concerns, including moral concerns, than merely maximising the number of people who are able to live the healthiest lives possible. But such an analysis is an open challenge to society to justify our policy choices (and to bioethicists to justify the narrowness of their research focus on things like euthanasia and informed consent).


Conclusion

Economists are used to being told that what they do is non-ethical or even anti-ethical. I hope they may take heart from seeing their subject as already an ethical science, with its own sophisticated resources, perspective, and contribution to make to ethical thinking. Economists do not have to put up with the sneering disdain of ignorant philosophers. Nor do they have to surrender the ethical ground to the 'real' experts.

But this essay is not intended to let economists off the hook. Economists should be much more sensitive to the perspectives that ethicists can bring, and much less eager to retreat into positive theory and scientific neutrality (as they have been more or less since the Marginalist revolution). Not only in the direct technical advice they give governments but also indirectly in their training of undergraduates and their choice of research questions and techniques, what economists do and say matters for society in a way that subjects like physics does not.

Economists greatest ethical problem is not their lack of ethical interests and analysis, but their belief that economic scientists shouldn't talk about ethics. And they need to talk about their ethics, to each other, to the public, and to professional ethicists. The challenge for economists is to make their ethical science better. To understand both the capacities and limits of their economics ethics, to incorporate that understanding more fully into their work, and to articulate this to sceptical philosophers and public alike.

Ethicists on the other hand should be much more humble in their interaction with economists. If they listened as well as they talked, they would see that economics has developed sophisticated analytical tools that can be usefully employed in many areas of moral philosophy (as economists like John Broome and Amartya Sen have been demonstrating for some time). They would see that economics can deliver insights into significant moral issues of our time, particularly those involving heterogeneity, conflict, and scarcity in a community (for example, global warming).

The single most important bridge between ethics and economics relates to the pragmatism that drives the logistical techniques of economics. Ethics at its best is all about perspective, about standing back from our first order intuitions about a situation or relationship in order to properly appreciate its subtleties and nuances. Its central question is, Shouldn't we think about that too? Economics at its best is all about proportion, about looking closely at how important things really are when considered as part of a complex closed system. Its central question is, How much does it matter? Good judgement needs both sense and sensibility, both proportion and perspective.





*I just read a great paper on this, though not by an economist: Scott Wisor, “Against shallow ponds: an argument against Singer’s approach to global poverty,” Journal of Global Ethics 7, no. 1 (April 2011): 19-32. (Gated, unfortunately, but that's academic publishing for you.) I was also impressed by Peter Singer's interview by the economist, Tyler Cowen. With regards to utilitarian analysis, it seemed to me that the economist wiped the floor with the moral philosopher.

9 comments:

  1. But here is where economics falls short: Normative claims depend on revealed preferences. These revealed preferences need to satisfy certain properties, such as choice set independence (a version of transitivity). Human behavior does not necessarily satisfy the axioms of revealed preference.

    In short: the descriptive basis for economic "ethics" does not pan out. To sort out this mess you need to go back to ethics, or else look to psychology and neuroscience.

    ReplyDelete
    Replies
    1. Here's the thing though: the big-ticket item in everybody's life is the existence and proliferation of the public sector - it takes roughly half of everything that the median agent earns (once taxes, charges and future obligations to repay government debt are taken into consideration); it makes rules about things like what we may eat/drink/smoke/snort and what volume our toilet flushes 'need' to be; it can, if it wants, decree that we have to dress up in a costume and go kill our analogues from other countries.

      And the basis for this menu of coercion is.... what exactly?

      Well, economics tries to set up a situation in which it may make sense - that there may be
      (a) considerations other than 'pure' allocative efficiency (due to utility interdependence); and/or
      (b) markets in which allocative efficiency is not achieved because of the existence of costs or benefits that are not accounted for in private decision making.

      Put simply, economics is alive to the fact that the distribution of resources that occurs under allocative optimisation can be perturbed in a way that increases overall welfare (under certain conditions).

      But here's the rub: both of these things rely on quantifiable, cardinal, interpersonally-comparable utility. That's a concept that I think is nonsense, having studied it formally and in depth: my conclusion is that individuals could, in principle, quantify their own utility, but that the resulting measures would not be interpersonally comparable - hence not summable-across-society, and hence not a basis for coercive reallocation of resources.

      But anyhow - back to the 'base story': the rationale for .gov is the correction of market failures of one or other of the types above Correcting them enables one to add up Harberger Triangles of surplus that would not otherwise exist... and declare that social welfare has improved. Then, .gov funds itself via progressive taxation (because of the diminishing marginal utility of money) and Bob's your uncle.

      (Obviously the story ends before we get to rent-seeking, corruption, perverse incentives, the principal-agent problem, and the fact that only States 'do' industrial-strength war).

      Now... what's the 'ethical' justification for the existence of the State? I ask the question in all seriousness: I have no training in Ethics as a 'stand alone' discipline, and I despise dilettante autodidacticism (something every trained economist encounters ALL THE TIME).

      With my deep and abiding antipathy (nay. hostility) towards the internet's legions of dilettante autodidacts firmly on the table, I'll have a go at answering my own question.

      The ETHICAL basis for the State - the biggest thing in most people's lives - is that it sets up systems that result in things being 'fairer' or more 'just' or some other such term (sadly, Google searches for "Ethical justification for government" result mostly in things written by economists).

      In other words, the ethical basis for the State is that the system under anarchy does not produce an optimal outcome and that said system needs to be perturbed (by force) towards something 'better'.

      Again, there appears to be the unstated notion that there has to be some set of supermen who
      (a) can reliably discern what 'better' is (with or without the help of the mass of citizenry); and
      (b) can budge the system towards 'better' in sensible ways that have a chance of actually hitting their targets without engendering a vast laundry list of predictable but 'unintended' consequences.

      Seems to me that when trying to answer the 'big ticket' item in modern industrial life, Ethics masks the inherent logical paucity of ITS justifications for .gov behind straw-man attacks on Economics' justifications. I would love to be proved wrong.

      Delete
  2. @Anonymous

    Realize, the author is arguing that the rationality model of economics is not an empirical prediction, but a normative principle. Simply put, it's not that economics claims that individuals act rationally, but that we should think about normative problems _as if_ behavior was rationally. Any other model for human thought is simply unacceptable, because it assumes too much of other people -- particularly that they're a bit dopey, which is the whole point of behavioral research in economics. People may be dopey, but you're not allowed to plan for them as if they were dopey. This is the argument from neoclassical economics; it's a close cousin to the categorical imperative.

    In actuality, the state of economics isn't nearly as messy as people think it is. Yes, we have new kinds of crises every time, but blaming economic engineering for it is like blaming penicillin for cancer: yes, cancer rates have risen alarmingly, and it's largely because people don't die of old-timey younger-onset diseases anymore. The current global economy is, truly enough, going through a kind of chaotic events, but actual human suffering does not nearly rival the kind of crises we've learned how to treat -- like the Great 1930s Depression --, let alone crises that resulted from a paternalistic, "people are less than rational" attitude to economic planning that has plagued periodically socialist economies. There are no bread lines. There's little famine in the world.

    ReplyDelete
  3. A wonderful, and in part ironic, assessment of ethics and economics. Ethics is, after all, about social living and, perhaps about the relative "deserts" of present versus future generations (so, for example, environmental policies are ethical because they involve claims of future generations on us).

    I have a problem with rational choice, however, because it seems to imply freedom of will. Capitalists love to argue in favor of freedom of choice; but, in fact, we don't choose but rather follow (rationally) our own presuppositions and needs. In Aristotelian terms, we seek to flourish. That's where freedom leads and that's enough to prefer the market to politics when we divide up chances to flourish among people -- since it is those chances that are scarce and a proper subject for ethics and economics.

    But ethics asks us to be decent to people, at least insofar as we can assure them a minimally acceptable life without harming ourselves appreciably. That sort of consideration leads to thoughts about labor and compensation that are decidedly not matters of efficient production. This, even when the cost of decency might be the reduction in either jobs or profits. How does one trade those things against one another using economics?

    ReplyDelete
    Replies
    1. "Ethics asks us to be decent to people" - first you need to define what 'decent' is.

      You indicate that you think that involves "assur[ing] them a minimally acceptable life without harming ourselves appreciably".

      As a recipe for self-directed individual action, it is almost possible to disagree: in economics we refer to this as "utility interdependence" (that agents may derive happiness from - gasp! - other agents' happiness).

      But from there, "Folks should be nice to one another" almost always transforms to "You know what? I don't like [gays/Muslims/weed/alcohol]: folks ought to be forced to be more like us, amirite?." and from there quite rapidly to death squads and executions without trial, jury or the right to defend oneself.

      Also, you seem to have a distorted view of what "rational" means (as do most people).

      It doesn't mean that agents are Cylons (or Daleks); it doesn't mean that they have 'perfect information' (or even that the central tendency of their expectations of the future is accurate); it doesn't mean that they only count the costs and benefits to themselves.

      "Rational" (in the context of economic models of optimising agents) means that agents will use the information set that they have to hand (which includes guesses about future variables, including their own preferences) in forming their decisions.

      One corollary of the approach is that agents will not (generally) deliberately make a decision that they expect to make them worse off.

      Point is, it's about what agents expect, which is not necessarily what actually happens; they can be wrong, and systematically so, even in environments where being wrong is costly (e.g., Gruen and Gizycki's work on 'anchored' traders in Foreign Exchange markets).

      There is also a lot of stuff on 'bounded rationality' - where agents can rationally choose to remain ignorant of relatively low-cost information that would help them refine their expectations (i.e., would help them reduce the error in their expectations-formation process).

      Delete
  4. Thanks for these comments. Sorry for the brief replies.

    Anon. Normative principles are about how things should be, so by definition they can't also be descriptively true.

    Diego. Thanks. Nicely put.

    Dennis. Aristotle didn't think we would necessarily follow our true nature and act so as to flourish. In general economics is about maximising decency rather than moral excellence (people as they are, not as they should be i.e. liberalism). Adam Smith is good on this. As to your 2nd point, economics is quite capable of complex constrained maximisation problems such as you suggest. For example, John Rawls used the work of economists on social choice extensively in his Theory of Justice to come up with his Difference Principle.

    ReplyDelete
  5. Dear Beard: Might I point you in the direction of the book Debt: The First 5000 years? You can find a decent introduction to its scope in this interview. I think it offers an interesting angle on why these two topics so often seem to be incommensurable.

    ReplyDelete
  6. Um, surely Ethics aims to find the course of action a person ought to rationally take in a given situation. If so doesn't that by definition make all economic models that assume human rationality flawed in so far as they are able to make ethical judgements. They might, as you have shown, be able to show what a rational individual would do given their current normative structure, but the inability for the model to speculate as to what the most rational normative structure is poses problems for the model's usefulness to philosophers.

    ReplyDelete
    Replies
    1. What do you mean by rationality? If you're referring to the distinction between instrumental rationality (econ - how to most efficiently achieve one's goals) and full practical reason (including reflection on one's goals) then I would agree an important dimension of human life is excluded from consideration by economists.

      But then I never suggested that economics could replace how we think about ethics. Rather that there are some areas of ethics, and some methodological tools, that economists are actually quite good at and mainstream ethicists should be willing to learn from.

      Delete

Comments will be moderated for civility and relevance