Paul Romer on mathiness and orthodox economics methodology
Recent criticisms of the mathiness of many economists has raised the question within the blogosphere of whether a fundamental fault-line has now punctured economics orthodoxy
Matthew Watson, Professor of Political Economy and ESRC Professorial Fellow, University of Warwick
Over the course of the summer Paul Romer has set the economics blogs alight with his accusation regarding the so-called mathiness of many of his peers. We can be pretty sure that this is not going to be a candidate for the Oxford English Dictionary’s Word of the Year for 2015. But it is noteworthy, nonetheless, for the way in which it appears to strike right at the heart of orthodox economics methodology.
It is necessary to be clear from the outset, however, what mathiness is not. Most economists just don’t get why most non-economists question their reliance on mathematics in the theoretical modelling process. Paul Samuelson’s defence of a mathematically-oriented methodology continues to leave a deep imprint. He argued that mathematics is merely a language and that every economic statement should therefore be directly translatable into mathematical form. Moreover, the advantage of doing so is that the mathematical translation produces a more succinct, precise and rigorous version of the statement than could ever be possible using normal words.
Given this assumption, the view from outside economics that economists’ work contains too much mathematics can never really make a lot of sense to those inside economics. And this is certainly not Romer’s point. He does not want to see any less mathematics being used in economics. Neither does he have any obvious axe to grind with a particular type of mathematics. His concerns about mathiness only involve the need to use the mathematics for the right purpose.
Romer’s specific complaint is that too many articles now make it through the refereeing process and into the economics journals where the mathematical and the economic elements of the argument do not sum to a coherent whole. Mathiness arises from trying to provide a dubious economic argument with a veneer of mathematical respectability, but where the content of the mathematics is unnecessary for, and maybe even contradicts, the economic claim being made. A whole host of big names, Romer says, have recently been guilty of trying to protect the perfect competition view of the macroeconomics of growth by using fancy but ultimately superfluous mathematics to deflect attention from the questionable economic meaning of the argument.
Many people from outside economics might well nod knowingly at this point. After all, their charge of too much mathematics often turns on the worry that this becomes a means of asserting the perfect competition view of the world as the only legitimate starting point for economic enquiry. The outsiders’ objection to too much mathematics and Romer’s objection to too much mathiness might therefore seem to be two peas from the same pod.
Not so fast, though. Romer sees nothing wrong with the way in which the perfect competition view of the macroeconomics of growth might act as a stalking horse for an ideological preference for free-market economics. Anyone who thinks that a critic of mathiness must also be an advocate of heterodox political economy is therefore likely to be disappointed.
Romer’s solution to the mathiness problem is merely to re-state the case for the most orthodox positivist philosophy of social science. We never learn whether he thinks that it is a sin to use misleading mathematics to perpetuate a worldview that promotes inequality and entrenches privilege and poverty in equal measure. We are only told what an affront it is to the classical model of scientific explanation to continue holding an opinion when the accumulation of relevant data suggests something else entirely. The possibility is simply never entertained that the assumption that every problem of social organisation has one single right answer is as much a diversionary tactic as mathiness itself.
‘The goal in starting this discussion’, Romer explains, ‘is to ensure that economics is a science that makes progress toward truth’. Defined out of existence immediately, then, is the alternative idea that economic theory must necessarily be plural because political discord will always be present over the desired form of economic institutions. Whilst economics might be seen to flourish only insofar as it represents the full breadth of this discord, for Romer it can flourish only when everyone gives up their opposition to the one true answer.
Romer has decided that there is no option but to ‘exclude people’ who will not let the facts speak for themselves, by which he seems to mean those who do not read the available evidence as much as he does in support of his alternative to the perfect competition view of the macroeconomics of growth. Persistent disagreement is not to be treated as a signal that you and I simply hold different opinions about the social objectives that economic institutions are designed to meet. Rather, he says, it is ‘a sign that some of the participants … are not committed to the norms of science’.
Romer’s faith in mathematical economics as the protector of the norms of science is thus countered only by his fear that others will be less trusting of it if the tendency towards mathiness is not nipped in the bud. ‘[A]fter readers have been disappointed too often by mathiness that wastes their time, they will stop taking seriously any paper that contains mathematical symbols.’
And here we perhaps come to the crux of the matter. There has been a collective intake of breath on the economics blogs at how daring Romer has been to call out some of the biggest names in his profession, including a number of people who have won Nobel Prizes for their work in macroeconomics. Given that Romer himself is widely regarded as an outstanding candidate for a future Nobel Prize, the ghost at the feast in many of the blog commentaries is whether he has shot himself in the foot in this respect by not showing more deference to existing Laureates. Certainly, the accusation of mathiness is not a compliment to their work.
However, if we shift focus from the identification of the problem of mathiness to its cure, a very different reading emerges. In defending mathematical economics and linking its integrity to a related defence of the classical model of scientific explanation, we might just have been privy to a first draft of Romer’s future Nobel acceptance speech.Print page
Articles and comments posted on this blog reflect the views of the author(s) and not the position of SPERI or the University of Sheffield.