Conference Report

The Society’s Annual Conference took place this year at the University of Surrey, 21st -22nd April. This report of the event is written by Saugato Datta of The Economist.

Three days spent attending talks on a variety of topics in academic economics at the Royal Economic Society conference promised to provide, if not a column or two, then at least substantial food for thought.

The sessions, on everything from research methods in development to macroeconomic policy to climate change sounded promising, and suitably varied. In the midst of the Great Recession, I actually found it refreshing that while there was reflection about the economic crisis and efforts to use research to illuminate what was going on, this did not take over the conference to the exclusion of everything else

Another of the pleasures of the conference was the way in which sessions devoted to seemingly very different topics ended up addressing related issues, with the contribution of one set of speakers illuminating the concerns that had been raised by another.

For example, Pinelopi Goldberg’s talk on her research on the effects of India’s trade liberalisation provided evidence about the mechanisms through which it worked using detailed firm-level data on the composition of their inputs and outputs. By itself, this would have been enough to pique my interest. But when Esther Duflo and Mark Rosenzweig discussed the merits and limitations of randomised experiments as a research methodology in development economics, Dr Goldberg’s work helped provide a concrete example of the kind of policy evaluation that simply was not amenable to randomisation, but about whose effects it was clearly important for economists to be able to give concrete and useful answers.

In this way, it really illuminated the more abstract discussions about methodology to be able to look at instances of one or other method of research in action, in order to be able to tell more clearly what the costs and benefits of different ways of doing research were. At the same time, the discussions about methodology provided a useful lens through which to judge examples of research like Dr Goldberg’s. This inter-relatedness and the effortless segués between very different areas and styles of research ensured that the conference was considerably more than the sum of its parts.

Here, then, are some impressions from the sessions and lectures which I particularly enjoyed, with no particular attempt at comprehensiveness or exhaustiveness.

Frank Hahn Lecture, David Laibson
After several changes of title, to which RES President John Vickers alluded while introducing him, David Laibson’s talk was still listed in the final programme as ‘The Financial Crisis of 2008’. In the event, Dr Laibson treated the audience to a bravura exposition of his interpretation of recent economic events, through the lens of what he called ‘Bubble Economics’. He provided a clear, accessible, and I thought, convincing interpretation of the global crisis that had its roots in asset-price bubbles (not only about the oft-discussed house price bubble, but also about a residual equity price bubble) whose genesis and effects could be understood using several phenomena that behaviourial economists like Dr Laibson find characterise human behaviour. He argued that this provided a cohesive explanation of the economic events of the last decade, and that the welfare costs of this bubble were very large.

Dr Laibson explained how the rise in house prices in places like Phoenix, AZ and Las Vegas, NV, were simply not amenable to an explanation other than that they were part of a bubble. The coastal cities of the United States — New York, Boston, and Washington might be genuinely cramped, but Dr Laibson’s description of Phoenix (aided by a satellite map that showed how Phoenix was essentially surrounded by farmland and desert) made it clear that no such argument was possible for that city, where there was ‘nothing but space to build’. So the doubling of house prices in Phoenix, when anyone looking around should have known that there was an essentially flat long-run supply curve, provided Dr Laibson for the perfect segué into discussing the psychological biases that underlay the behaviour of market participants.

The psychological biases relevant to the bubble, he argued, were extrapolation (where people look to the past to predict the future — a generally useful heuristic but a terrible one, Dr Laibson argued, in financial markets), return chasing (which follows from extrapolation), excessive optimism, and herding or social proof, which together are ‘all the things that belong in a general model of how we come to believe in a world that doesn’t exist’.  The upshot of all this, said Dr Laibson, is that when people are prone to behave in ways that psychologists and behavioural economists believe they do, it takes only a very small set of mistakes to generate the wrong asset prices, or the doubling of house prices in a place like Phoenix.  The bubble, he estimated, had a value of between $15-20 trn.  He then argued that given all available estimates of the marginal propensity to consume, consumption in the US rose exactly in line with what a standard model would imply in response to the bubble in asset prices.

All this excess consumption was paid for by running huge trade deficits. Here, though, he took issue with Ben Bernanke’s savings glut explanation for the consumption boom, which argues that the developing world’s high savings found natural recipients in the Irelands, Icelands and USAs of the world. Laibson argued that standard models would imply an investment boom rather than a boom in consumption. Given the path of interest rates, capital stock and investment, he argued, the bubble holds up well as the causative mechanism.

What now? Once the bubble burst, Dr Laibson’s calculations predict a drop in consumption of 4 to 15 per cent — with the upper end far beyond anything that has been seen so far. The primary welfare costs would come from a combination of resource under-utilisation (essentially, the recession we are now in), and consumption volatility, which is particularly painful because people have to cut back when their consumption is low, and so where every additional unit of consumption would actually add more to their utility. In sum, he reckoned that these two channels would account for the lion’s share of the welfare costs, calculating them to add up to around $5.3 trillion. In contrast, surprisingly, he found that the welfare costs of inefficient investment — all those unsellable condos — is ‘a rounding error’, costing about $250b.

Denis Sargan Lecture, Pinelopi Goldberg
Pinelopi Goldberg quoted Paul Krugman, who wrote that ‘the purpose of international trade — the reason it is useful — is to import, not to export. That is, what a country really gains from trade is the ability to import things it wants. Exports are not an objective in and of themselves; the need to export is a burden that a country must bear because its import suppliers are crass enough to demand payment.’

This was certainly not the dominant tone of the discussion in India in 1991, when the IMF foisted a widely unpopular trade liberalisation on the Indian economy as part of the conditions for a $2.5 billion bailout loan. But it turned out to be an ideal reform to study the effects of trade liberalisation, because the IMF’s diktats regarding across-the-board tariff cuts meant that jockeying by industries trying to head off greater import competition, which usually means that which industry’s tariffs are cut has a lot to do with lobbying abilities, was minimised.

As part of those reforms, India slashed tariffs on imports from an average of more than 80 per cent in 1987 to 39 per cent by 1994. Non-tariff barriers were reduced too, from an average of 87 per cent to 45 per cent. Not surprisingly, India’s imports did boom, growing by 200 per cent between 1991 and 1997. But the effects on Indian manufacturing were not what the merchants of doom had predicted: the output of the manufacturing sector grew by 200 per cent over the same period. And by looking carefully at what was imported, what it was used for, and what Indian firms made, Dr Goldberg came to the conclusion that cheaper and more readily available imports actually contributed to the growth in India’s domestic industrial growth in the 1990s.

In essence, the reduction of tariffs on all sorts of goods meant that Indian manufacturers gained access to a variety of intermediate and capital goods which had earlier been too expensive for them to use them profitably. The increase in imports of intermediate goods was much higher, at 227 per cent, than the 90 per cent growth in consumer goods imports in the thirteen years to 2000. The prices of imported intermediate goods that were already in use before liberalisation declined by 4.7 per cent per year after the reforms. This allowed manufacturers to turn to lower-cost producers of inputs they already used.

Also, some inputs might have had no perfect domestic substitutes, so being able to import them opened up entirely new ways of producing something that was already produced, but using an inferior input variety. But even more important, some inputs that were unavailable or uneconomical before could be indispensable for particular products. So over time, removing restrictions on imports could broaden the range of what domestic firms made, and spur more research activity aimed at developing new products, the most important channel of technological progress.

Very detailed firm-level data about inputs and products allowed all these ideas to be tested in the Indian case, and the results were illuminating. Of the surge in imports of intermediate goods that followed liberalisation, fully two-thirds had simply not been imported when India had a more restrictive trade regime. India was not just importing more intermediate goods, it began importing new ones. And this led to an explosion in the variety of goods that Indian manufacturers made: the average firm made 1.4 products before liberalisation, but by 2003, this had increased to 2.6. Detailed data about what inputs firms used to make which products allowed the link to be made explicit. Furthermore, the increases in variety were largest for industries whose input tariffs were cut most.

Interestingly, the theory would also predict that some products would cease to be produced, so Dr Goldberg expected an increase in both additions and removals from product lines. In practice, it turned out that Indian firms simply did not drop products nearly as often as they added them, or indeed anywhere near the rates seen, for example, in America. Creative destruction, in the Indian context, did not seem to involve much destruction. Dr Goldberg speculated that this might be because India is such a diverse market: there is always a segment lower down in the economic hierarchy which is quite happy to buy products which richer consumers turn their noses up at.

RES Presidential Lecture, John Vickers
As John Vickers pointed out in his Presidential lecture, there is a strong tension between competition policy and property rights, with a whole host of legal questions, many with very important economic implications, which are not amenable to being resolved simply by looking at the law on the books. Some might argue that the very essence of property rights was excludability. But on the other, he said, one could imagine that competition would never arise in a market with a powerful incumbent unless competition policy intervened. So when, if ever, Sir John asked, should competition law require a dominant firm to share its property with its rivals?

He outlined a model that suggested that a rival is less likely to develop new products if it cannot share in the profits from the dominant firm’s invention. If the leading firm is free to licence its technology on strict terms, it reduces the profits of rivals who have to pay. Rival firms are indeed encouraged to innovate because of the prospect of a larger payoff. But on balance, the incentive to innovate will be greater where access is granted more freely, because profits now are more valuable than profits in the future.

Another argument for reining in dominant firms is that the competition to innovate tends to be keenest where the battle to be the dominant firm is very close, with firms running neck-and-neck. If market leaders license their knowledge on easy terms, the pay-off from research and development is reduced. But smaller firms may also be able to catch up quickly. Despite lower profits from any new inventions, they will be sought more aggressively when competitors are similarly placed.

Finally, he argued that the strongest reason to trespass on property rights would arise if the dominant firm used them to stifle ‘follow-on’ innovation. An ideal set-up would reward the most important inventions without stifling those that would build on them. This would require the original inventor a share of profits from related products that his invention might lead to, but with enough left over to make rivals’ R&D viable.

Sir John peppered his talk both with details of relevant cases, as well as anecdotes about the way in which the US Supreme Court functioned. One learnt, for example, that while Justice Scalia and Justice Bryer were from the so-called ‘conservative’ and ‘liberal’ wings of the court respectively, they both tended to concur in cases related to trustbusting, arriving at the same relatively non-interventionist position from very different theoretical underpinnings. But the factoid that may have interested those in the room most was that Justice Bryer had recently published an article in the Economic Journal, something that most of the audience would aspire to but few may have achieved.

Economic Journal Lecture, Gilles St Paul
As Gilles St Paul pointed out, most models of reform tend to focus on the conflict of interest between groups that stand to gain or lose from the competing policy proposals. In reality, he argued, there is also a lot of disagreement about the working of the policy: conflicting views (and not just conflicting interests) matter. Examples of such conflicting views could include disagreements about the size of the Keynesian multiplier, the cost elasticity of labour demand, or the effect of returns to education on people’s incentives to acquire skills. But where, he asked, did these beliefs come from?

What, for example, did the people of France believe? (The choice of country was clearly heartfelt: I got a sense of great personal frustration with the belief systems of the French and their cussed intransigence on important matters). France, Dr St Paul pointed out, is among the countries in the world with the greatest hostility to the market economy. Those views, he argued, are shaped in part by an educational bureaucracy whose beliefs differ substantially from those of broader society. Schoolteachers were self-selected in terms of being those with either very negative views about the market economy to begin with, or those whose views had been coloured by bad experiences with the private sector, so they would not be representative of the general population.

In addition, once a person became a teacher, she would gain little additional experience about the market economy, and would therefore not update her views (though surely working in the private sector is not the only way to garner such experience, and even French schoolteachers must shop). But in any case, in Dr St Gilles’s model of occupational choice, those initial criteria through which people selected into the teaching profession would transmit a particular set of beliefs to students.

Finally, he argued that economic institutions also affect the capacity to learn from experience . A society which has a negative view of markets is likely to favour a more rigid labour market, which would in turn reduce individuals’ experience of the market economy, and thus their ability to learn from experience. This then reinforces the importance of the educational system in the way beliefs are formed, thus validating people’s initial presumption against the market economy. In equilibrium, beliefs and institutions reinforce each other.

Special Session, Experimental and Non-experimental Approaches to Development Policy
Esther Duflo is often referred to as one of the founders of the ‘new development economics’ centred around the use of randomised experimental designs in developing countries, so it was interesting, to say the least, to hear her say that she found the term itself ‘somewhat strange’. Having got that much out of the way, Dr Duflo argued that experiments were good for identifying causal effects. But as important, she argued, was that the availability of the experimental approach to development economics fostered creative approaches to test theories, and were capable of challenging theory by uncovering results that standard theories would not deem possible, to explain which new theories would have to be generated. In her view, the problem of external validity, which is often brought up by critics of randomisation-based approaches, was a bit of a red herring in that it applied as much to other approaches to empirical work in development.

Randomised experiments would be most useful, she argued, where they were designed explicitly to design variation that would allow a parameter of interest to be teased out. But experiments are only as good as the questions they ask, may not be a good way to identify general equilibrium effects, and scaling up is an ever-present issue. Experiments, she said, are ‘neither necessary nor sufficient’ for a good piece of research.

But just as Dr Duflo argued that one of the great contributions of randomisation was to have attracted bright young students to what used to be a relatively boring corner of economics, Mark Rosenzweig, worried that the discipline also needed to think about whether the attractions of experimental approaches could crowd out interesting non-experimental work. Agreeing that one of the great benefits of the experimental approach was that it was quicker and that there was little scope for quarrels about identification, he argued nonetheless that this came at the cost of asking more limited questions.

In any case, Dr Rosenzweig argued, there were plenty of important questions where randomisation was simply not feasible: how, for example, would one randomly allocate the productivity of schooling? And even if one could - Dr Duflo gave the example of research that tried to manipulate, if not the actual returns to schooling, then what people thought those returns might be - then would this raise ethical concerns?
Both agreed, in the end, that there were uses for the experimental approach done right, and that there were limitations, even if there were clearly differences between which side of the dividing line between costs and benefits certain features lay on. But Dr Duflo’s response to a question from the audience, about where additional research funds should be used, was a fitting end to this fascinating discussion: take all the money that is wasted on ‘utter uselessness’, she argued, and divide it up between well-thought out experimental and non-experimental work.

Special Session, Growth
Chang-Tai Hsieh argued there has been a good deal of progress in accounting for the proximate determinants of growth, but less about where differences in these factors (human capital, physical capital, etc) came from. Dr Hsieh argued that total factor productivity played the most important role in explaining differences in income per capita between countries, both directly as well as through its effect on the accumulation of both kinds of capital. And TFP varied so much, he argued, most likely because of misallocation of resources. Looking at the distribution of marginal products in manufacturing and services, for example, gave the sense that more and more resources were going into a sector where productivity was stagnant or falling. What I enjoyed most about his talk was the way in which possible (and plausible) explanations for variations in things like human capital accumulation were made to face up to the evidence (and then, often, rejected).

But, as Pinelopi Goldberg pointed out in her response, saying that TFP, which is essentially an unexplained residual, was the crucial factor in growth was essentially tantamount to saying how little we have actually learnt. This is a view with which I have much sympathy; I do believe that a combination of the sort of work Dr Goldberg is doing on trade, and the ‘new new’ randomisation, which seeks to uncover parameters of interest rather than just evaluate programmes, holds the key. Dr Hsieh said that he was frustrated by the existing micro evidence. I hope that a few years down the road, he will have more reason for cheer.

Finally, a few comments on the general sessions. I stumbled upon a few I really enjoyed, particularly a talk about the economic origins of ethnolinguistic diversity as well as some entertaining discussions of the economics of blogs and violence in Canadian hockey.

But in some instances, I (and in at least one case, one of the speakers) were more than a bit foxed by the choice of talks that made up a single session. Several seemed to have been thrown together in one session mainly because they had not fit anywhere else. This made for some entertaining segues, but might be something for the organisers to try and minimise. The other thing, of course, was that I felt a bit sorry for some of the general sessions which had virtually no audience. Perhaps holding the conference in a university with a large number of graduate students would make the general sessions a bit better attended. That said, the University of Surrey proved a remarkably pleasant setting for a few enjoyable days of listening to economists talk about economics.

Page Options