web stats

INSIDE, vol. 2, no. 1

Karl Whelan (“Thoughts on Teaching Economics After the Crash”) has a thoughtful comment on the debate over pluralism and economics education. What follows is very critical, so I will first point out that I appreciate his honest engagement with the project and agree with a fair amount of what he says. Nevertheless, I think his piece is also characterised by some common misconceptions about heterodox economics and pluralism in economics education which need correcting. I will also note, as I do throughout, that Whelan has been more than ready to acknowledge the shortcomings of his piece. But this should not mean those shortcomings are not worth exploring in further depth.

Student Concerns

Whelan makes a solid effort to understand and address students’ concerns, and he agrees that more real world application is needed in economics degrees. This is undoubtedly a good thing. However, I still get the feeling Whelan believes arguments about the shortcomings of economics are fundamentally misconceived. For example, he argues that “first-year biology students moaning about not hearing anything about nanotechnology” would not be covered by the BBC, which implies that this is analogous to the current debate about economics. He only concedes economics is different because it is “part of the fabric of people’s daily lives,” so students should be better equipped with knowledge to deal with issues they feel are important.

The concerns of economics students are deeper than this. The problem is that, unlike in other subjects, in economics you do not start by building an in-depth understanding of the different parts of the economy, how to study it, what is important in economics, and then go on to study those problems which are most easily modeled. Instead, you begin with a watered-down version of the most advanced, abstract theory. As I’ve previously (“Not my idea of a Simplification”), despite being ‘basic’ relative to economic research, many of these models remain difficult for students to grasp because the concepts involved are not intuitive in nature (utility, capital etc.) and not enough time is available or devoted to unpacking them. Most time as an undergraduate is wasted solving mathematical problems or reciting theories by rote. This may culminate in a look at some of the more applied or complex areas of economics in your final year, but even this rarely gets close to the kinds of models used in research.

In other words, the reason economics education fails is not because it doesn’t address the most complex and/or interesting areas of research within the discipline; it fails because it insists on teaching students a rigid and hard to grasp methodological framework at the expense of making sure they understand key issues within the discipline. Students cannot use the framework they are taught to answer interesting or complex questions without taking the mathematical difficulty far beyond what is expected at undergraduate level. The result is a plethora of algebra and regurgitation which pushes out room for critical engagement. (Note: I am not conceding for a second that all the problems are ironed out at research level. Even at this level the framework remains rigid, and is forced to incorporate interesting questions in a relatively ad hoc manner to avoid mathematical indeterminancy. But perhaps that point is best saved for another time.)

Whelan goes on to argue that although economic theory often assumes people are “cold analytical calculating machines” (an obviously false assumption), this does not make the models generated by these assumptions irrelevant:“Take introductory micro, for example. We can derive models of supply and demand from first principles based on rational choice theory. But the basic insights of the model don’t rely on these assumptions. Do we need to assume hyper-rationality to argue that people probably purchase less of a good if its price goes up? Are we assuming people are maximising computer-like drones if we argue that the price of something going up makes it more attractive to supply it?”

The argument is actually suggestive of a far deeper issue: there is no need for economics students to derive microeconomic theory from “first principles”. The demand-supply model stands or falls on its ability to explain empirical reality, not on its relation to arbitrary axioms about consumer behaviour. The propositions that “people probably purchase less of a good if its price goes up” and “the price of something going up makes it more attractive to supply it”, stated by Whelan as if they are obvious, seem to me to highlight how reliance on axiomatic theories can push out habitual empiricism. I am not suggesting Whelan’s statements are necessarily false, but that textbooks which spend time expounding the so-called ‘law of supply and demand’ spend little time offering rigorous empirical evidence supporting the theory, or discussing situations in which itmight not hold. Whelan rightly argues for devoting “a bit less time to teaching theory” and a bit more time spent on empiricism, but stops short of a more radical recommendation – such as ejecting relentless deductivism in favour of building a critical, empirical understanding of economic theory.

Heterodox Economics, Old and New

Whelan is uncharacteristically lazy in his treatment of heterodox economics. He first makes the all-too-common error of suggesting an equivalence between heterodox economics and ‘old theories’. This a strange assertion when heterodox economics is a contemporary evolving discipline with its own journals and books. Not to mention that being ‘old’ does not mean a theory is wrong. Newton’s Laws are pretty damn old, but they still have a lot of relevance. To be sure, due to the complex and social nature of the subject, nothing heterodox (or mainstream) economics has – or ever will have – the level of empirical support as Newton’s Laws. But this is actually a reason to be suspicious that old theories may have been rejected for reasons other than absence of empirical confirmation. Indeed, a relatively recent issue of the Cambridge Journal of Economics was devoted to refuting this ‘Whig’ interpretation of the history of economic thought.

Whelan goes on to make a series of puzzling claims about heterodox theories. In general, he states without argument that “heterodox” quantitative models such as stock-flow models or Steve Keen’s brand of post-Keynesian macro are really just standard quantitative macro models with some ‘tweaks.’ All I can say is that if you consider no microfoundations, no optimisation, disequilibrium and a completely different way of closing a model (balancing accounting entries) ‘tweaks,’ then everything is a tweak of modern macroeconomics! Whelan then attacks a straw man by saying the models “are also subject to the exact same critiques about abstract theorising and unbelievable assumptions as standard models and are just as technically difficult.” In fact, most heterodox economists – particularly post-Keynesians – do not attack abstraction or mathematical difficulty per se, but attack contemporary economics for gratuitous and inappropriateuse of abstraction and mathematics. There is a big difference.

In particular, Whelan claims Stock-Flow Consistent models are “unstable and hard to work with”, which is odd given they mostly use basic algebra and, well, aren’t “unstable” in any obvious sense.Andrew Lainton has given a good run-down of why SFC models are probably a better way to think about economies than marginalism, primarily because money flows are actually observable by economic agents, but ‘marginal utilities’ and similarly abstract concepts are not.

Whelan also criticises Steve Keen’s Minsky model as being “highly stylised and seem to me to feature odd definitions and assumptions”. Concerns about Keen’s communication of his ideas notwithstanding, it seems to me that ‘odd definitions and assumptions’ just means ‘assumptions not equivalent to those used in the mainstream’. In particular, Keen’s argument that aggregate demand is equal to income plus the change in debt is just a disaggregation, perhaps confusingly worded, but ultimately not that hard to understand. One interesting and relatively straightforward extension of Keen’s model is this paper by the mathematicians Grasselli and Costa Lima, which contains interesting results relating to instability, debt levels and ‘Ponzi financing’.

Whelan’s final puzzling claim is a caricature of Minsky. “If Minsky’s argument was that financial crises are an innate part of Capitalism, will economics be a branded a success when the next crisis occurs if we have been teaching our students about Minsky and that such a crisis was inevitable?” This is bizarre. Minsky’s financial instability hypothesis does not claim that crises are inevitable and there is nothing we can do about them except look smug when they happen. Instead, it describes the process by which financial crises develop and unfold: the relatively calm period causes people to become (rationally) overconfident and overextend themselves with investment and borrowing. It’s obvious that, yes, if people (especially regulators) were more aware of this possibility, they would be more skeptical of the new financial instruments which supposedly lowered risk. On top of this, Minsky’s arguments and prose are relatively straightforward. (I’d certainly recommend them over Brunnermeier and Shin, both of whom would be largely impenetrable to undergraduates.) All in all, Whelan’s attitude toward heterodox economics exemplifies economists’ difficulty engaging with non-mainstream literature, which itself is a symptom of the narrow nature of the discipline.

Bit and Pieces

Whelan also makes a number of points which I either partially agree with or have discussed in detail elsewhere, so I won’t spend too much time on them. He rightly argues that there is nothing in the conclusions of mainstream economics which implies ‘free markets’ are always and everywhere the best policy: there are numerous ‘interventions’ advocated by mainstream textbooks and economists: fiscal stimulus, environmental regulations and competition policy to name a few. Nevertheless, he does not ask the more fundamental question: Why use the perfectly competitive baseline – and the related idea of an ‘untouched’ market – at all?

The perfectly competitive market as a baseline for analysis is illogical if the real world departs from this baseline in virtually every conceivable way, just as illogical as using a cow as a ‘baseline’ for a horse. Framing issues as ‘governments versus markets’ ignores the institutions and policies which shape capitalism in the first place. Furthermore, it’s entirely possible this outlook contributes to a loosely ‘conservative’ worldview by setting up a Panglossian vision as the baseline for analysis, and may explain the common use of ‘economics’ to support free market type policies. Ultimately, the perfectly competitive market has become completely unfalsifiable, since the conditions under which it supposedly occurs will never take place in the real world. It’s time for economics to move on.

Whelan also discusses the relevance of the financial crisis to economics, with two main claims: first, that forecasting crises is impossible; second, that economists understand key mechanisms driving the crisis. These are points #3 and #7 in my article “Is the Economic Crisis a Crisis for Economics?” I’ll recap them briefly. First, forecasting a crisis by its exact date may be impossible, but making conditional predictions based on key observations and mechanisms is not. Whelan appears to address this point when he states models may be good for ““What if?” analysis i.e. if X happened, then what usually happens to Y?”, but he actually just moves the question backward one step by assuming X itself is random and unpredictable. In fact, it’s entirely possible for X to be a gradual, observable phenomenon (say, the build-up of private debt) and to know there will be catastrophic, sudden consequences (say, a financial crisis), but not to know exactly when this will occur. Second, mainstream economists do not understand how crises happen. They understand what happens when an otherwise smoothly functioning model is hit with an exogenous shock or change in parameters. This characterises both macroeconomic models – from IS/LM to DSGE with financial frictions – and microeconomic models of banking/finance like Diamond-Dybvig or Akerlof/Romer.

Conclusion

Whelan concludes with a familiar refrain: that economics is “more complicated than any of us can understand”, and there are “no magic formulae” that will solve the problem once and for all. This is reasonable in itself, but I would like to join Roger Farmer in his frustration with this kind of ‘all models are incomplete/wrong’ fall back that crops up time and time again in economics. The problem is that these are truisms. Students and heterodox economists are not complaining that economists’ models fail because they don’t incorporate every single aspect of the economy or don’t predict perfectly. They are claiming that economic models fail to (as Whelan puts it) “help us get a better (if still incomplete) understanding of the world than we would have if we weren’t using the model to organise our thoughts.” Critics of economics do not want economics to be perfect; they want it to be relevant and interesting. This can never happen if there is no willingness to abandon models which are untrue or practices which are unnecessary and outdated.