Thursday, 12 January 2012

Economists in government: what are they good for?

Since the financial crisis, economists as a profession have come in for considerable criticism, both for failing to predict the crisis and for disagreeing violently with each other about what to do about it. As far as macroeconomics and finance go,  it is difficult to dispute that there is a strong case to answer. But the criticism has often extended beyond these topics to a broader critique of what is sometimes described as “economic imperialism”,  which could loosely be described as the view that economics was not just about markets or supply and demand, but was a method that could be applied to almost any social policy issue:   the “science of choice”.   This in turn translates to a view that economists have become too dominant in advising and informing policymakers.   [This case is set out well in Ben Fine's book "From Economics Imperialism to Freakonomics"].

In many respects I find this critique quite convincing. But in this post I'd like to focus on the narrower question of what economists within government (or acting as policy advisers) can in fact usefully do to make policy better.  Not surprisingly, given my 23 years in the civil service, I do think there are important roles, although in my experience they correspond almost not at all to what outsiders, or indeed academic economists,  think economists in government spend their time doing. I would identify three functions (and macroeconomic forecasting is not one of them!).

First, and perhaps most important, is to explain, in language which intelligent non-economists (and most Ministers and at least some journalists fit this description) things which are true but not obvious. I base this description on a famous incident when Stanislaw Ulam, one of the greatest mathematicians of the last century (and one of the inventors of the hydrogen bomb) asked Paul Samuelson (Nobel Prize winner and the “father of modern economics”) whether economics, or indeed any social science, had ever come up with a theory which was both true and not obvious. Several years later, Samuelson pointed out that the theory of comparative advantage fits the bill.  And it is important to note, as Samuelson did, that while understanding and proving the theory is trivial, explaining it convincingly, in a concrete not abstract context, to a harassed Minister (and in a way that he can then explain to the general public) is very far from being so.  And while most informed commentators do now understand comparative advantage, at least in the abstract sense, there are a number of other equally important concepts that I would put in the same category:
-     
  •  The lump of labour fallacy.  I explained this to six successive Secretaries of State for Work and Pensions, usually in the context of immigration, but also in relation to people on incapacity benefit and older workers. To their credit, they all got it, and were prepared to go out and defend policies that were consistent with the view of the labour market that underlies it.   I thought then, and still think now, that this was probably the most useful thing I did, from a public policy perspective, in my six years as Chief Economist at Department for Work and Pensions;
  • Modigliani-Miller.   The fact that the return on an investment should, if markets are working, be independent of the way it is financed, is particularly relevant given the current debate about capital requirements for banks.  If every policymaker understood that when bank lobbyists complain that “equity is more expensive, so requiring more of it will reduce lending and raise borrowing rates” they are arguing the contrary, usually without any explanation or justification, then they might be treated with the derision they usually deserve.
  •   The paradox of thrift. I have already written here about David Cameron’s apparent failure to grasp this one, so clearly the profession has failed in this respect.
Of course I’m not saying that Modigliani-Miller, say, holds always and everywhere; these are results from simple theoretical models, not descriptions of the real world. But that doesn't mean they aren't useful in a real world policymaking context.   Each of these models gives  a very strong conclusion based on a very clear analytical framework, based on a set of assumptions; if you want to dispute the former, you need to explain which of the latter don’t hold and why, what that means, what your alternative framework is, and (in government) the policy implications. That is an extremely strong and useful discipline in policymaking; and it something economists are well placed to explain and to implement.


The second contribution economists can make is that we bring our own particular “bag of tricks” to the examination of public policy problems.  In this, everyone will have their own favourites (and some are not exclusive to economics) but I would identify the following, at least, as being particularly useful and well-developed: 

  • opportunity cost; if something (labour, investment, tax revenue, whatever) is used for something it can’t be used for something else;
  •  general equilibrium; that while individuals, companies and even in some context countries can usually "take everything else as given" (for example, the general level of prices and wages) this is not the case if you are looking at the economy as a whole. For obvious reasons, this concept is almost irrelevant in business decisions, but is essential in policymaking;
  •  correlation is not causation.  Of course other social scientists know this perfectly well, but it is perhaps most relevant in economics, both macro and micro; it is no accident that the best-known application of this principle to policy in the UK is named after an economist ("Goodhart's Law").


The final, and most difficult because least well-defined,  role of economists (and other "professionals" within government) is to assess critically, synthesise, and translate into policy, the research consensus in a particular area. This is difficult for (at least) two reasons.  First, it requires judgement.  For every paper in social science you can of course find one giving the opposite result.  Academic standards - where something was published, the author's contribution to the advancement of theory - don't necessarily tell you what to take seriously and what to dismiss from a policy perspective.   Second, while doing this need not be a political exercise, it is inherently and inevitably an  ideological one; where you come out will be conditioned by your views, attitudes and background.   There is nothing wrong with this, but we need to be aware of it.

Again, a few examples of where this process has helped inform good policy and where it hasn't are informative:
-         
  • labour market policy, a broad consensus emerged during the 1990s, based to a large extent on the work of academic labour market economists in the UK, on the important role of active labour market policy.  This consensus was over time, via economists in government, fully internalised and translated into policy, by and large successfully;
  • by contrast, in financial economics,  a similar consensus emerged on the benefits of financial innovation (in particularly, better allocation of both capital and risk) and the role of "light-touch" financial regulation in facilitating such innovation. Again, it was fully internalised and translated into policy here and in a number of other countries, with somewhat less successful results;
  • finally, a more contentious and live example is "expansionary financial contraction", where an academic debate has obvious real-world policy implications, and no consensus currently exists . It is of course my view that the empirical evidence on this is pretty clear - fiscal contraction, under most circumstances, and certainly those obtaining in the UK today, is contractionary.  I would argue that economists in government should not use the fact that there is controversy in academia, and that the issue is politically charged, to avoid taking a position on the balance of the evidence.  

To conclude, as I said, one thing I don't think government economists should be doing is forecasting the economy.  That's not because I think (as many academic economists do) that forecasts are a waste of time; you can't run economic, fiscal or monetary policy without a forecast, so they are at least a necessary evil.   But there is no reason at all to think that government economists will be better than external ones at forecasting, and a number of political economy ones why they might be worse.  Government economists need to be able to explain forecasts, scenarios and their implications to policymakers - but actually making them is definitely not their comparative advantage.

[These thoughts are based on my discussion, at a recent Institute for Government event, of Vicky Pryce's paper "The DismalScience".]

8 comments:

  1. Looking back from today, would Samuelson have replied that his theories might be obvious or false, but they were still almost as dangerous as Ulam's?

    ReplyDelete
  2. "... correlation is not causation."

    This is the weakest part of scientific method. Usually economists stop here and forget about two really important requirements to be matched in the hard sciences:
    1. no correlation - no causation
    2. causation is always expressed as correlation

    Samuelson's answer together with the author's explanations lack one important part - Ulam's answer and the responce from the hard sciences. So to say, this is a one way road which has been chosen by economists - they state something and deny any criticism on the ground that nobody understands them correctly.

    ReplyDelete
  3. Does the lump of labor fallacy imply that new labor-saving technologies should never lead to shorter working hours? Discuss.

    ReplyDelete
    Replies
    1. No, it doesn't. Economists would indeed expect that labor-saving (ie productivity enhancing) technologies would lead people, over time, to work less and take more leisure. That is indeed what's happened most places most of the time, and it is a great puzzle why it hasn't happened in the US. The point is thought that working shorter hours doesn't in itself create (or destroy) jobs for anyone else.

      Delete
  4. On that lump of labor fallacy again: Me and my wife would love to work six hours a day three days a week going hard and in my free time build my own house, cultivate a garden, cook and eat at home, take care of my own children and my parents when they get old, and continue on working (part-time) for as long as I am able, even into my 70's or beyond. A lot of other Americans feel the same way.

    See Gallup poll..


    If companies could make money by it why wouldn't this be a good idea?

    It would require coordination, planning, infrastructure.

    ReplyDelete
  5. I don't get his argument that economists in government should NOT undertake forecasts of the economy because expertise resides outside. A sort of "specialisation" and "crowding out" private forecasters argument. But what about other considerations e.g. avoiding the contractual 'hold up' problem if we relied outsiders? Also where do we stop?

    ReplyDelete
  6. I'd add something rather more mundane to your list. Economists in government are generally highly numerate, and that is a very useful counterpart to those civil servants of a more legalistic mindset. I suppose one could call this a part of the "bag of tricks", but it is really more about general skills than having a particular set of conceptual tools.

    ReplyDelete