What microeconomists think about macroeconomics

                There is a lot that is interesting in Diane Coyle’s 2012 Tanner lectures. What I want to respond to here is an associated post, where she expresses some of the frustrations that I know many microeconomists have with macroeconomics. These can be summed up rather crudely as: why do we disagree all the time, and yet appear so sure we are right, and don’t we realise we are bringing the whole discipline into disrepute? Diane says that “macroeconomists simply do not realise how low their stock has sunk in the eyes of their microeconomist colleagues”. This may be an exaggeration, but who knows what is said when we are out of earshot. Because I (and pretty well all macroeconomists) do have great respect for our microeconomic colleagues, we have some explaining to do. (Jonathan Portes has an excellent post responding to some of the more specific comments Diane makes in this post.)

                What I want to say in summary is this. Microeconomists are right in many of their criticisms, but what they often fail to see is the root cause of the problem. This is that macroeconomic policy is highly political, with strong ideological implications. Ideology and politics distort macroeconomics as a science. Yet despite this, there is – and for many years has been – a substantial body of analysis that most macroeconomics would sign up to, and which has sound empirical backing.

                What is this substantial body of analysis? It is what used to be called the new neoclassical synthesis (Goodfriend and King (1997) – see here for more background and references). For a closed economy, its details are well represented in Romer’s graduate textbook, for example. This body of analysis has important gaps and omissions, of course, such as a naive and simplistic view of the financial sector. However, as I argued recently, the financial crisis itself showed up this incompleteness, but did not invalidate most of what was in the synthesis. Indeed, events since the crisis have provided significant empirical support for the Keynesian elements of that synthesis. (Of course, there are caveats, some of which I have discussed in earlier posts, and which I will return to below.)

                So why the appearance of constant disagreement among macroeconomists? The first point to make is that the disagreement tends to focus on one part of the synthesis, which is short term macroeconomic fluctuations: their causes and cures. This is not the only part of the synthesis which has the potential for political and ideological controversy, but it is where there is constant popular and media interest. The second point, however, is that politics alone is not sufficient to explain controversy within a scientific discipline. Here I might give climate change is an example: despite strong pecuniary temptations, climate change science is pretty united, and typically dissenting views come from outside the discipline.

                The reason, I would suggest, why many macroeconomists do not sign up to the new neoclassical synthesis is that its Keynesian component conflicts with an ideological view which uses economics as part of its intellectual foundation. That view is that markets generally work well when left free from political interference and that government intervention is almost always bad. Now most microeconomists will immediately point out that microeconomic theory as a whole does not support this ideological view. Much microeconomic analysis is all about market imperfections. But that misses the point. Those attracted to this ideological position use a very partial take on economics for support, and are naturally attracted to that part of the discipline.

                Keynesian analysis is all about the consequences of one particular market imperfection. Conventionally this is seen as the externalities that arise from sticky prices. (I personally am increasing drawn to describe them as the consequences of the existence of money, but that is a different story.) It implies in general the need for constant involvement of one arm of the state – the central bank – in stabilising the economy, and in some circumstances the involvement of fiscal policy to do the same. I have argued that this analysis leaves those with the ideological view that free markets are good and government intervention bad feeling very uncomfortable, so they will try and find ways of avoiding and disputing this part of the synthesis.

                Now nothing I have said so far makes macroeconomics unique in attracting ideologically driven controversy. There are areas of microeconomics which are equally controversial for similar reasons. I would suggest the debate over minimum wages is an example. An outsider looking at this debate might well conclude that labour economists are as equally disputatious as macroeconomists on this issue. Another example from my youth I always like to use is the ‘Cambridge controversies’ over ‘reswitching’. This was a highly technical theoretical dispute, but its ideological implications were such that views on this technical debate were highly correlated with political beliefs.

                I think there is one factor that is peculiar to macroeconomics, and that is the role of ‘city’ economists. There are two elements here. One is that city economists work in an environment which has a strong vested interest in promoting the free market ideological view. The other is that part of the name of the game for these guys is to differentiate the product. These are of course tendencies: there are many excellent city economists who go with the evidence rather than with ideological conviction.

                Why is it important to say all this? First, because the influence of ideology on economics is not confined to macroeconomics. It is also not confined to the political right, so there is no simple bias, but recognition rather than denial is important. Second, the cure – if there is one – is evidence. As Jonathan points out, the recent evidence is clearly with those who accept Keynesian views, but I think there is an underlying problem. The way empirical evidence is marshalled in the microfoundations methodology associated with DSGE modelling makes it too easy for those who hold ideologically based priors to retain and defend those priors, which is one reason why I have tried to suggest that it should not be the only way serious macroeconomic analysis is done.