Paul Romer on math masquerading as science

16 May, 2015 at 16:18 | Posted in Economics | 15 Comments

I have a new paper in the Papers and Proceedings Volume of the AER that is out in print and on the AER website …

Paul_RomerThe point of the paper is that if we want economics to be a science, we have to recognize that it is not ok for macroeconomists to hole up in separate camps, one that supports its version of the geocentric model of the solar system and another that supports the heliocentric model …

The usual way to protect a scientific discussion from the factionalism of academic politics is to exclude people who opt out of the norms of science. The challenge lies in knowing how to identify them.

From my paper:

“The style that I am calling mathiness lets academic politics masquerade as science. Like mathematical theory, mathiness uses a mixture of words and symbols, but instead of making tight links, it leaves ample room for slippage between statements in natural versus formal language and between statements with theoretical as opposed to empirical content.”

Persistent disagreement is a sign that some of the participants in a discussion are not committed to the norms of science. Mathiness is a symptom of this deeper problem, but one that is particularly damaging because it can generate a broad backlash against the genuine mathematical theory that it mimics. If the participants in a discussion are committed to science, mathematical theory can encourage a unique clarity and precision in both reasoning and communication. It would be a serious setback for our discipline if economists lose their commitment to careful mathematical reasoning …

The goal in starting this discussion is to ensure that economics is a science that makes progress toward truth. A necessary condition for making this kind of progress is a capacity for reaching consensus that is grounded in logic and evidence. Given how deeply entrenched positions seem to have become in macroeconomics, this discussion could be unpleasant. If animosity surfaces, it will be tempting to postpone this discussion. We should resist this temptation.

I know many of the people whose work I’m criticizing. I genuinely like them. It will be costly for many of us if disagreement spills over into animosity. But if it does, we can be confident that the bad feelings will pass and we should stay focused on the long run …

Science is the most important human accomplishment. An investment in science can offer a higher social rate of return than any other a person can make. It would be tragic if economists did not stay current on the periodic maintenance needed to protect our shared norms of science from infection by the norms of politics.

Paul Romer

One of those economists Romer knows and — rightfully — criticizes in his paper is Robert Lucas.

Lucas is as we all know a very “mathy” person, and Romer is not he first to notice that “mathiness” lets academic politics masquerade as science …


Added 20:00 GMT: Joshua Gans has a post up on Romer’s article well worth reading, not the least because it highlights the nodal Romer-Lucas difference behind the “mathiness” issue.

In modern endogenous growth theory knowledge (ideas) is presented as the locomotive of growth. But as Allyn Young, Piero Sraffa and others had shown already in the 1920s, knowledge is also something that has to do with increasing returns to scale and therefore not really compatible with neoclassical economics with its emphasis on constant returns to scale.

Increasing returns generated by non-rivalry between ideas is simply not compatible with pure competition and the simplistic invisible hand dogma. That is probably also the reason why so many neoclassical economists — like Robert Lucas — have been so reluctant to embrace the theory wholeheartedly.

Neoclassical economics has tried to save itself by more or less substituting human capital for knowledge/ideas. But knowledge or ideas should not be confused with human capital.

In one way one might say that increasing returns is the darkness of the neoclassical heart. And this is something most mainstream neoclassical economists don’t really want to talk about. They prefer to look the other way and pretend that increasing returns are possible to seamlessly incorporate into the received paradigm. Romer’s view of human capital as a good example of non-“mathiness” not-withstanding, yours truly is of the view that talking about “human capital” — or as Lucas puts it,”knowledge ’embodied’ in individual people in the short run” — rather than knowledge/ideas, is only preferred because it makes this more easily digested.

Added 20:55 GMT: Romer has an even newer post up, further illustrating Lucasian obfuscations.

Added May 17: Brad DeLong has a comment up on Romer’s article, arguing that Lucas et consortes don’t approve of imperfect competition models because they are “intellectually dangerous” since they might open up for government intervention and “interventionist planning.” I agree with Brad, but as I’ve argued above, what these guys fear even more, is taking aboard increasing returns, since that would not only mean that policy preferences would have to change, but actually would bring havoc to one of the very fundaments of mainstream neoclassicism — marginal productivity theory.

Added May 18: Sandwichman has a great post up on this issue, with pertinent quotations from one of my intellectual heros, Nicholas Georgescu-Roegen.

Added May 19: David Ruccio has some interesting thoughts on Romer and the fetishism of mathematics here.


  1. For crying out loud! When will somebody notice that comparisons to physics are psychotic! The science of living things starts with Darwin. If economics is about human beings, it starts with Darwin.

  2. On the “mathiness” of it all: “Mincer meets Lerner, Rybczynski, and Solow
    Human capital externalities, factor substitution, and technology differences”
    Erich Gundlach, GIGA German Institute of Global and Area Studies and Department of Economics, Hamburg University, Germany

  3. You are spot on about “human capital.” Romer is on thin ice. In fact, what made “human capital” so attractive and persistent as a concept is its “mathiness.” Its residual is even bigger than in the Solow model and the proxies needed to measure it, much worse. What I believe Romer is arguing is that some economists do “machines” better than others and do not use it as an a priori to foreclose on thinking through models in new ways!!!! Once again, Deidre McCloskey’s papers on formalism in math and its limits to economic modeling remains a better critique than Romer’s “reglement du compte!”

  4. As far as I am concerned, Romer (1990) is an example of “mathiness”. Others. including Ian Steedman, have been very critical with the unconcern of those in the new growth theory with the lack of clarity in the meaning of their symbols.

    • I certainly agree. Some of the critique Romer is mustering against Lucas et consortes does also fall on parts of his own writings. Defending such as hopelessly loose and diffuse concept as Becker’s “human capital” is one very salient example …

  5. The science that never was
    Comment on ‘Paul Romer on math masquerading as science’
    “The goal in starting this discussion is to ensure that economics is a science that makes progress toward truth.” (See intro)
    Reassuring declarations like this appear regularly when economics is in crisis and they have regularly been counter-productive.
    “The claim to the scientific nature of economics is prey to suspicion the moment that it fails to be self-evident.” (Benetti and Cartelier, 1997, p. 211)
    What is in fact self-evident is that economics is a failed science. And this not as recently as the onslaught of some mathematical extremists but at least as the beginning of Neoclassics itself.
    One curious thing to notice is that the thread’s title sends the reader in the wrong direction. The problem has never been that math had masqueraded as science. The real embarrassment has always been that economists have used math in order to masquerade as science.
    However, what comes across is that something has gone badly wrong with math. This, as has been suspected by so many for so many years, is apparently against the spirit and standards of science. And now it falls upon Paul Romer to save the economics world.
    “The usual way to protect a scientific discussion from the factionalism of academic politics is to exclude people who opt out of the norms of science. The challenge lies in knowing how to identify them.” (See intro)
    As matters stand now, some people, nice colleagues, have to be expelled from the temple of economics but Romer is acting in the best interest of the whole profession.

    “Economists have a collective stake in flushing mathiness out into the open.” (Romer, 2015, p. 90)
    Why? What stake? How did the crown jewel of economics become something to be flushed? Let us speculate. Could this be a strategic withdrawal? Is the dispensable ‘bad’ mathematical freak sacrificed for the benefit of the ‘good’ realistic maximization-and-equilibrium kind of guy?
    The Great Recession has now come back on economics itself. On closer inspection, a growing number of non-economists rate it quite realistically as what it is: good for entertainment but not up to the scientific task. The writing is on the wall, Soros spends millions on New Economic Thinking and students defect to Heterodoxy and pluralism.
    Time to close ranks and to make a sacrifice to calm the angry masses. Among the most popular points of critique excessive math has always been at the top of the charts. And the likes of Lucas, McGrattan, Prescott, Moll (sorry, dear colleagues, (Romer, 2015, p. 92)) have simply gone over the top. Their economic absurdities are so drastic that even the representative economist, who swallows almost everything, cannot swallow this ‘lemon’ (Romer, 2015, p. 90). The New Classical mathies have to be flushed to save the core of Orthodoxy.
    This does not suffice. In order to ‘exclude people who opt out of the norms of science’ one can exclude the overwhelming majority economists from economics. Why? Because they subscribe to the same fundamental methodological mistake/error.
    “There is in economics, or at least among the overwhelming majority of its disciples, broad agreement as to what represents the corpus of their subject. This corpus revolves around the concept of maximizing behaviour, whether it be by the individual, firm or institution.” (Blaug, 1990, p. 209) see also (2014a)
    This faulty behavioral axiom has produced numerous offspring: supply-demand-equilibrium, general equilibrium, marginal utility, well-behaved production functions, total income = value of output, total income = wages + profits, marginal product distribution, I=S, and others that fill the textbooks (2014b).
    It is bizarre that economists, who after more than 200 years still cannot tell the difference between profit and income and have still not realized that utility and equilibrium are nonentities like dancing angels on a pinpoint, suggest that there is a valid scientific core that is jeopardized by mathiness.
    There is no such core, Paul Romer’s fallback position is untenable. Scientific incompetence, not mathiness, is the problem of economics. ‘Progress toward truth’ is hindered by Paul Romer and the maximization-and-equilibrium kind of guys themselves. Nothing less than a paradigm shift will do.
    Egmont Kakarot-Handtke
    Benetti, C., and Cartelier, J. (1997). Economics as an Exact Science: the Persistence of a Badly Shared Conviction. In A. d’Autume, and J. Cartelier (Eds.), Is
    Economics Becoming a Hard Science?, pages 204–219. Cheltenham, Brookfield,
    VT: Edward Elgar.
    Blaug, M. (1990). Economic Theories, True or False? Aldershot, Brookfield, VT:
    Edward Elgar.
    Kakarot-Handtke, E. (2014a). Objective Principles of Economics. SSRN Working
    Paper Series, 2418851: 1–19. URL
    Kakarot-Handtke, E. (2014b). The Three Fatal Mistakes of Yesterday Economics:
    Profit, I=S, Employment. SSRN Working Paper Series, 2489792: 1–13. URL
    Romer, P. (2015). Mathiness in the Theory of Economic Growth. American
    Economic Review: Papers & Proceedings, 105(5): 89–93. DOI

    • The only interesting analysis of Romer’s critique is above. The persistence of economics needs to be understood. It is clear that critiques like Romer’s have worked to strengthen and protect the anti-science, anti-empirical nature of the discipline.

  6. What Romer seems to be saying is that the assumptions behind all the maths in economics needs to be questioned and sharpened up. For sure they do . But I think he misses the point. To sharpen up economics, first a lot of the maths has to go. A lot of the subject is simply an artificial construct to allow for the maths – what Keynes calls “pretty and polite techniques”. It is ontological and epistemological inquiry that needs sharpening up. Maths is limited in its ability to understand complex social systems and its forced use actually gets in the way. There is a role for maths in economics, and for models. But to closer to the truth we need a multidisciplinary approach and the subject has to engage with knowledge outside the discipline.

  7. What Romer is railing against appears to be what’s called Authoritative Voice. (And what Krugman calls VSP (Very Serious People).)

  8. “In one way one might say that increasing returns is the darkness of the neoclassical heart. And this is something most mainstream neoclassical economists don’t really want to talk about.”

    You’re aware that Romer’s most cited paper, which, by the way, contains quite a neoclassical model (has optimizing agents, equilibrium and everything), is literally called “Increasing Returns and Long-Run Growth”, right?

    • Yes I am. And?

  9. Noah has a good write up on his blog.

  10. however, I like the piece. It is really a settling of scores against those who dumped on endogenous growth theory model with less than stellar math and empirics.

  11. this is what I wrote to Deirdre McCloskey: As you many know, Romer has been roaming in areas and on topics upon which you had already roamed with better titles and more clarity! In fact, he should be assessed a roaming charge (he, he)

    • If Deirdre McCloskey could read, she would never have slandered Adam Smith as a libertarian. Her critique is exactly like Romer’s in yet another way: it serves to protect economics from getting anywhere near the truth.

Sorry, the comment form is closed at this time.

Blog at
Entries and Comments feeds.