26 April 2013

Whats the Use of Economics?

Teaching the Dismal Science After the Crisis

Diane Coyle (ed)
2012, London Publishing Partnership, 197 + xxii pages, £14.99
ISBN 8601200901627

Reviewer: Peter Sinclair, Emeritus Professor of Economics, University of Birmingham

How much blame do economists deserve for the financial crisis that broke in 2007-8? Why was it foreseen by so few? Is the economics we teach at fault? How and where does what we teach need to be altered? Are economists in fact asking the wrong questions? Or giving wrong answers to the questions? How should the research agenda change? Where and how should we modify the models we use? Why are economics students discontented? How can we produce better young economists?

These are now suddenly revealed as really pressing questions. All economists are thinking seriously about them, informally by ourselves or in small groups. But until now no-one has organized a detailed symposium or volume devoted specifically to exploring those issues. Until, that is, the appearance of the volume under review.
What’s the Use of Economics? Has a fine scene-setting introduction by its editor, Diane Coyle. Twenty-two chapters follow. They differ in punch, length and depth. Yet all have something interesting to say. Each offers penetrating analysis and criticism of the discipline of economics, varying in acerbity. Most of the authors are academics with, collectively, a formidable stock of wisdom and expertise on which to draw. Six work in the United States. By discipline, one is an economic and financial historian, another an expert on financial journalism, a third a sociologist by trade, and the rest are economists. Outside academia, the UK Treasury and the Bank of England are well represented with two powerful essays. And there are pithy and practical pieces from the worlds of banking and economic consultancy.

Many contributors complain that economic history has been unjustly evicted from the teaching of economics, and should resume its place there. Ten of the essays urge this. Dissent is sparse. There is also eloquent support for including the history of economic thought, from Alan Kirman for example, in his erudite and thought-provoking piece. But what aspects of the history of economies and economics should we include? Precisely where should we include them? What should we omit to make space for them? And have we got enough people able to teach them?
For undergraduates, and most Masters’ students, many models can be unremittingly arid. They cry out for motivation. A few minutes devoted to a non-technical aside, before broaching the details, works wonders in a lecture. Such asides command attention. They are remembered. They might refer to current events, or to historical phenomena, or indeed, sometimes, to the ideas of an early economist. Hume is a must on money and external payments; Cournot on mineral water, as a preparation for Nash equilibria; Dupuit on the benefits from well-chosen infrastructure projects; snippets from Adam Smith on a vast range of issues; the ancient Greeks on why useless gold commands such a higher price than indispensable water. Students need not read those authors, though some will wish to, and gain if they do. Furthermore, it is odd that inter-university cooperation, common in research, is all too rare in teaching. Why not include an economic history option by video-conference?

The case for including the history of economies and economics in a top PhD or Masters programme, not only as a field option where resources permit, is urged powerfully by Wendy Carlin.The case for adding them in Masters’ courses, preferably through inter-university collaboration, is strong (as it is for more discussion of practical policy). But when it comes to PhD courses, your reviewer confesses to reservations. The opportunity cost of building either into the core would be considerable. As would the financial cost of mounting an option that, on present indications, unless it were shared with many universities, hardly anyone would take. Glaeser notes that young scholars are increasingly generalists, combining a broad range of theory with econometric applications. Today’s best work usually straddles the old (and imprisoning) borders between micro, macro and stats. So should major graduate schools specialize in different areas of economics? No. Yet the case for comprehensiveness and uniformity is moot. At the Tepper School at Carnegie-Mellon, new economics PhD students concentrate at once on writing research papers, rather than getting bogged down in many arduous years of comps. Tepper is unique - and highly successful.
History might not be economists’ only lab. But it remains our main one. Interpolation, courage and caveats might allow us to push key macro data-sets back many decades, enabling us to get a better glimpse of low- frequency events. Data points are useful. Yet so is span. Data that stretch back only a decade or two may help in some tasks. But their brevity, and the tranquillity of the times until late 2007, lulled financial risk managers and econometricians alike into false security. Deep history can be really useful. Harold James aptly entitles his essay “Finance is History”.

Was it not just the narrowness of data, but the discipline of economics itself, that failed? Ben Friedman and Andy Haldane offer finely crafted analyses of the financial crisis, the role played by gaps in orthodox thinking, and what we should do now. Friedman laments the paucity of attention paid in mainstream academic economics to financial frictions, institutional lending, credit aggregates, and non-rational expectations. For Haldane, policy-makers failed to spot the significance of increasingly integrated capital markets; and he agrees that fashionable theory was at fault in downplaying the importance of money and credit.

Friedman and Haldane are right. Yet financial frictions, capital market integration and non-standard expectations were far from absent in the academic literature. Still, they were rarely in the limelight. Often they were dismissed as awkward and insignificant complications. The Bernanke-Gertler financial accelerator model was developed before 2000, and would spawn a big literature. A not dissimilar fate awaited much good work on risks in both banking and international capital movements: it carried modest prestige and influence. The analyses of learning and bounded rationality by Sargent, Marimon, Marcet, McGrattan, Evans, Honkapohja and others, much of it pre-2000, painted a far subtler view of expectations. This work was respected. But largely sidelined.
Most policy-maker thinking, many university courses, and much academic research demonstrated at least partial adherence to the claims that expectations were rational, asset markets efficient, and financiers best watched only lightly. In a steady state, at least, how could wise people make systematic forecasting errors, or let assets stay mispriced? (The objection, that steady states were not the norm, was often ignored).

In other changes, public economics, traditionally the basis for optimal taxes, began to morph into political economy, and treat political behaviour as optimizing by selfish actors. And in the mid-1980s the expanding State stumbled in the West, soon followed by China, India and Russia. No longer the omni-competent solver of problems, the State had itself become a problem. The ideological vacuum was filled by 1970s economic doctrines, rarely disputed, that had emanated from Chicago. Fifteen years of benign macro data, Japan excepted, would reinforce all this. In financial regulatory policy, its most telling effect was the repeal of Glass-Steagall in 1999. But for that event, the bank-and-housing bubble that burst nine years later could hardly have happened.
One chapter, by Paul Ormerod and Dirk Helbing, attacks the orthodoxy best reflected in standard Dynamic Stochastic General Equilibrium (DSGE) models. They strongly prefer Herbert Simon’s approach to decision taking to the standard one. They call for a wholesale redesign of macro theory. Is DSGE just one of those culs-de-sac - like turnpike theorems, implicit contract theory, quantity rationing, or the Cambridge attack on capital - with which the history of our subject abounds? But DSGE is a broad church. It includes the Mankiw-Reis DSGE variant with ‘rational inattention’ - an idea barely distinguishable from Simon’s key notion of ‘satisficing’. Nor, oddly in my view, does this chapter complain about that most convenient but downright implausible of DSGE premises, Calvo pricing. A very different view is given by Jagjit Chadha. He offers a spirited and thoughtful defence of the panoply of macro theory, urging its retention. He gives the reader a fine undergraduate lecture, with some interesting diagrams that derive from the ideas of past authors but throw valuable light on recent events.

Roughly midway between fierce attack and robust defence comes a lengthy, probing, elegant and highly persuasive chapter by John Kay. Models simplify; they can’t capture everything, he stresses; and Robert Lucas was right to say that people could not have predicted ahead that Lehman’s would fall in September 2008. If they could, it would have collapsed earlier! But asset-price anomalies are what drive capitalist economies forward, Kay argues; and for him, induction from data observation is a far better lens for analysis than ideology. Kay mixes praise with sharp criticism for Chicago’s economics. But it would be churlish of this reviewer not to nuance what Kay says here by adding that easily the most magisterial account of the empirical failures of the efficient-markets hypothesis, published in the Journal of Finance in 2011, comes from the pen of none other than John Cochrane (of Chicago). And I am not fully convinced by Kay’s doubts about ‘bolting on’ an extra feature onto standard models. Putting financial frictions into DSGE models is an urgent activity, on which many scholars are now engaged. Ditto, learning, and credit or money. Such extensions offer more promise than adopting a wild ‘agent-based’ model where, it seems, anything can happen, but nothing can be explained.

Whether inductive or deductive, and favourable or suspicious towards markets, how should UK universities teach Economics, particularly at undergraduate level? Jonathan Leape, Michael McMahon, John Sloman and Alison Wride offer four pieces in the final section of the volume on this subject. Between them, they offer numerous suggestions. Among them, we see the case for problem-based learning (Sloman). We learn about a new multidisciplinary course at the LSE (Leape). And McMahon sketches an interesting incentives-focused model with interaction and net benefits for academics and students.

What makes the debate illuminated by these four papers (and others that come earlier, especially Carlin’s) more urgent is the publication of Britain’s increasingly detailed National Student Survey. In this, 122 UK universities are examined, department by department. The most recent results appeared on 28th September 2012. For economists they make worrying reading. Final year undergraduates’ responses about teaching quality rarely flatter Economics. Consider four other subjects with which Economics is broadly comparable: History, Law, Mathematics and Politics. Take the 30 pre-1994 universities that quantify teaching quality in all five disciplines, as recorded in the Sunday Times supplement of September 29th. In none of these does the score for Economics lead the others. In half, Economics comes bottom. When the focus is limited to three subjects, Economics, Law and Politics, which now encompass 39 of these universities, Economics is in third place for 26. And among the 26 come all the Russell Group institutions included, except for just one (Liverpool).

Does this matter? With the big increase in university tuition fees, it must. Economists in older universities are lucky that the demand for our courses has been so strong (particularly from abroad). Numbers taking A-Level Economics have shrivelled in recent decades (albeit with some reversal in the latest years), diverted, especially in challenged schools, into Business, Law, Media Studies, Politics or Psychology. If Economics students are known to be unhappy, applications and enrolments may fall. British business, commerce and policy-making could all suffer as a result.

Why are the ratings so poor? Economics is tough, but has juicy job prospects. So economists might sacrifice course enjoyment for an expectation of more pay later. Then incentives for staff in Economics departments stress research (and publications in peer-reviewing journals with very low acceptance rates) well above teaching, and probably more so than in most other disciplines. The five UK Economics departments included in the latest ranking of the world’s 30 best by IDEAS - for research reputation - have an average teaching quality score in Economics below 70% (as against an average of 80% for the other four subjects of History, Law, Maths and Politics). There is also an unpalatable possibility: Economics might simply be taught less well, by worse teachers.

Very abstract micro theory, favoured in hiring for its supposed research-rating benefits, is hard to teach excitingly to undergraduates. Recruits may baulk at teaching applied economics or macro; when they have to, students can spot the distaste and condescension. Furthermore, Economics instructors, told to put research first, may fall back on what they recall from their PhD courses, and feed their students a dull diet of ageing papers. Interesting current research and events, as well as good older ideas and telling historical phenomena, get played down, or cut out. Worse, tedious exam questions, based on rehearsing stale material, reward no-one but the assiduous parrot. Courses with such defects train no student to think. Practical problems stay unaddressed; there is no thrill at approaching the frontiers of the subject, and precious little scope for creativity. Few lecture courses have these defects. But it would be wrong to pretend there are none, either in the UK or abroad. The switch in departmental income from research to teaching brings large and painful costs of adjustment - but also, in time, the prospect of better-trained and happier undergraduates in departments that embrace the change.

Enthroning the student-customer may spell demotion for the decreasingly scarce, cheap or humble senior administrator. Economics trades with almost every discipline - Psychology, Engineering, Law, Mathematics, Geography and History, for example; so why is it often closeted in cumbersome officialdom with Management, Political Science and Sociology? Customs unions are almost always inferior to free trade. Extra layers of bureaucracy depress staff, slow up decisions, stifle initiative under mindless uniformity, and smother incentives by aggregation. Economics departments thrive best in universities with a more stellar structure, where the department, through its head, can deal directly and nimbly with the centre. Similar arguments may apply for national research councils.

Many culprits share some blame for the financial crisis: credit agencies, Greenspan’s ultra-low policy rate after 9/11, politicians who repealed Glass-Steagall, bankers, risk managers who mistook data frequency for span, regulators, and those who either bought or sold credit default swaps and other opaque or dubious derivatives. But economists are junior partners in crime too. Economists relied on models where serious errors and financial disasters simply could not happen. Despite both old and newer evidence to the contrary, they took, and helped others to take, too an optimistic view of future trends and financial markets. Economic research can take wrong turnings. The edifice of contemporary economic theory needs not replacing, but refurbishment, and improvements in how it is taught. This is the main message of the excellent volume under review. And it is the right message.