The Society’s Annual Conference took place this year at Royal Holloway, University of London, from 3rd – 5th April. This report was compiled by Richard Davies of the Economist.

It is hard to judge the state of economics at the moment. From one perspective it is booming. In Britain, the number of students choosing economics A-level, flat before 2007, has been growing 10 per cent a year since then. In America, economics sits 5th in a ranking of the most popular undergraduate degrees. Even PhDs, maligned in other subjects, remain popular in economics.¹ As far as the numbers go, economics is on the up.

But at the same time the subject seems in disarray. At the very top of academia debate takes the form of personal attacks between Nobel Laureates, played out in newspaper columns. Away from the limelight, any honest macro researcher has to concede the pre-crisis policy consensus-inflation targets, fiscal rules and freely-floating exchange rates-has been shattered. Much of microeconomics is in better shape, but in staple applications taught to undergraduate level (income inequality, for example) economies are moving in the wrong direction.

The truth is that economics is as confusing as it is popular. I went to the Royal Economic Society’s annual conference in search of some answers.

Learning to use the new tools
Perhaps the most important change in macroeconomics is the way post-crisis monetary policy works. It was fitting then that the conference opened with a session on ‘macroprudential policy’, chaired by Andrew Haldane of the Bank of England. This is central banks’ prized new tool-in reality a suite of balance-sheet levers (rules on capital, liquidity or funding) that are being created to lean against financial cycles.

Hélène Rey provided evidence on how important getting macroprudential policy right in financial upswings is. She pointed out a strong correlation between banks’ asset growth and leverage. The implication is that equity is very sticky: banks do not raise equity to match increased lending. This was not just a problem with unsophisticated small banks, she said, there was leverage pro-cyclicality at the world’s biggest banks too. The implication is that, in boom time, central banks should tighten capital ratios or liquidity requirements to offset this.

Whether the tools will work in the same way in a downswing is far harder to say. And it is that much harder question that today’s central bankers face: how to deploy macroprudential policy in a bust. Andrew Haldane set out the challenge, which he thought was partly psychological. The problem is that the correct response is ‘countercultural’: when recovering from a banking crisis the natural tendency is to tighten standards but the right thing to do might be to loosen.²

Martin Wolf and Hélène Rey agreed on what macroprudential policymakers should aim to do. First, they must protect the financial sector from the economy by making sure that banks can withstand a downturn in real activity (i.e. an increase in non-performing loans. But the tool has to protect the economy from the financial sector too, by offsetting credit-fuelled booms.

But while there was agreement on the aims, views were divided on the prospects. Both Franklin Allen and Martin Wolf highlighted potential clashes between central banks' two levers. Mr Wolf worried that from a macroeconomic perspective rate-setters might want to loosen policy, while from a stability point of view the capital-setters might want to tighten it. In the real world, these conflicts could be big, he thought. It is a problem that may arise in the latter half of 2013, given the combination of moribund economies and frothy-looking markets.

The panel seemed to agree that a banking system with more capital would reduce these conflicts (interest rate setters would not have to worry so much about financial stability). But the prospects for a better capitalised system were poor, Wolf thought, since the Basel III ‘risk weightings’ used to work out capital ratios are near useless. A simpler capital system would be better. Hélène Rey proposed a different solution, something she called ‘finer radars’ — essentially more granular financial stability warning systems. It seems a good idea, but the implication of fine radars would seem to be finer tools. That could mean varying loan-to-value ratios for different types of mortgage, or different capital weightings for loans in different parts of the country. These decisions would have distributional consequences. They might be tricky for independent central banks to make.

Rules and discretion
As the conference kicked off in earnest it became clear I was facing a huge capacity problem. There were 139 sessions-almost 200 hours of economics. To my regret given some of the exciting titles (for example, I missed Giovanni Prarolo’s paper on the rise of the Sicilian mafia) I was forced to dip into just a fraction of the sessions.³

Some interesting micro studies were motivated by concerns about low productivity. One idea is that low workforce effort and motivation are to blame. Michèle Belot of Edinburgh University presented a test of different employee monitoring schemes.⁴ Ms Belot gave student volunteers 780 euro area coins and asked them to separate the coins into their different types (there are 160) in exchange for a payment of €20. The job allowed the researchers to measure various forms of counterproductive behaviour: inaccuracy (the coins could be badly sorted), tardiness (the task could be completed late) and theft. They then tested different configurations of monitoring and rewards. A control group was not supervised at all, being paid immediately regardless of performance. Two other groups were monitored and received performance-related pay. The first monitoring scheme was lax: workers lost just €1 for every ten mistakes. The second was much harsher: the payment was cut by €15 if more than two coins were wrongly identified. The results are interesting, with the middle option-lax monitoring-unambiguously the worst: mistakes and lateness rose relative to the control group. Stricter monitoring does offer some benefits. Accuracy improved, with only 16 per cent of volunteers making more than ten mistakes. But workers slowed down, which Ms Belot interpreted as a sort of payback or ‘reciprocity’: the workers punished the monitor by cutting effort along a non-monitored variable (they were not fined for lateness, just inaccuracy).

Other papers also showed slack rules might be a bad idea. Maria Navarro Paniagua of Lancaster University found a link between pub opening hours and employee absence: with more time to drink, more employees went sick.⁵ If this hangover economics paints a worrying picture of British willpower, a paper by Elaine Kelly and Imran Rasul (conference chair) which tested the impact of cannabis decriminalisation, was even more concerning.⁶

Creating a quarterly panel data set covering hospital-admissions in the London boroughs between 1997 and 2009 allowed the impact of a de-criminalisation policy by Lambeth (in 2001-02) to be tested. The idea with the policy, popular in the Netherlands and Portugal, is that decriminalising ‘softer’ drugs frees up police time, so that resources can be spent on prevention (often education) rather than enforcement. But the case for decriminalisation is not clear-cut: some argue that lowering the expected cost associated with illegal activity will increase its prevalence, and that soft drugs can for a pathway to harder narcotics.

It was a great example of using data to answer a genuine policy question. The results of the difference-in-differences test were worrying given the current policy trend towards decriminalisation. Focusing on hospital admissions for class-A drugs like opium and cocaine, they found that 15-34 year old men were 40-100 per cent more likely to be admitted to hospital during the policy trial. It suggests that, in London at least, the threat of detection and punishment can be have an important deterrent effect.

A call for modest economics
Picking the right policy when faced by uncertainty was the theme of the Sargan lecture, delivered by Charles Manski.⁷ Mr Manski’s presentation was one of the most thought provoking of the week. Economists and policy advisers, he argued, often make decisions based on optimisation problems that cannot reasonably be solved. Partly this is down to demand: investors and policymakers want simple answers — a point estimate or a single optimal policy. Economists react by making stronger assumptions in order to produce these narrow predictions. But economists’ knowledge about assumptions is often quite poor, so that both the assumptions and the predictions they produce are not justified.

It is a critique that goes to the heart of economics. But Manski thinks it is solvable. His work offers two high-level ideas. One way to deal with uncertainty, familiar to investors, is to diversify. The idea would be that if the optimal ‘treatment’ — a policy or a drug, for example — is not known, the policymaker or doctor should randomly apply different treatment to different groups.⁸ They would then adapt the treatment as the results become clear. Another of Manski’s ideas is to loosen assumptions, accepting that this might mean reporting wider ranges when estimating the potential impact of a policy. It is a simple point, yet widely ignored. If credit-rating agencies reported expected loss ranges rather than simply ratings (a form of point estimate) it would have been clear in 2007 that not all the ‘triple A’ ratings they had stamped on assets were alike.⁹

Searching for growth: from Africa to Britain
It was a bitterly cold week, with April snow rather than showers. Friday morning was much needed: some sunshine and a discussion of how to get Britain growing at a decent rate again. Tim Besley and John Van Reenen presented the findings of the LSE Growth Commission.¹⁰ The report has an optimistic tone. For a start, stylised facts about Britain’s growth slowdown give a sense of hope. Britain's GDP per capita actually rebounded after 1980, having fallen back relative to America and Germany in previous decades. And the recent expansion was not all banking boom: many other sectors, including ‘business services’ (things like law and consultancy), and distribution, were more important than finance, even in the bubbly years between 1997 and 2007.

But more recently, British productivity has ground to a halt. Here the report offers a host of ideas. A couple of themes stand out. The first is the importance of promoting exit as well as entry. Education is one example: Britain’s universities are good, but its schools fall short of the standards in comparator countries. The Commission’s research review found that teacher quality is the single most important factor. But the market for teachers is not a fluid one: qualifying is hard, and it is tricky to get rid of bad teachers. A market with more flexible entry combined with stronger performance assessment and tough sanctions for poor performers would be better. The same principle — simultaneously lowering the barriers to entry and exit — can be applied elsewhere, for example to setting up new schools and winding down poor ones.

The second theme is how to get politics and economics to work well together. At present, Britain’s infrastructure decisions can be slow, and not evidence based. The Commission found that it takes too long to start projects in Britain: a new energy bill took 12 years, for example. A less political decision-making process would help. The group recommends building on the model used by the Bank of England’s MPC or the National Institute for Health and Clinical Excellence. Its proposal — a new ‘Infrastructure Strategy Board’ — would be politically accountable, but have independent experts as decision makers. It would use economic analysis to test for the best policy and take decisions quickly, compensate those made worse off.¹¹

If getting infrastructure right is important in Britain, it is even more so in Sub-Saharan Africa where close to 400,000 people living on less than $1.25 a day. The optimistic view is that the potential gains in Africa are huge, not least since road and rail densities are less than 10 per cent of those in Europe. The idea is that with better transport infrastructure, for example, the cost of trade should fall, leading to regional specialisation and higher trend growth. Alexander Morandi presented a historical test of this idea by looking at railway investment in Ghana.¹²   Between 1901 and 1903 a first railway was built to connect a port (Sekondi) to an inland mining town (Kumasi). Another followed, connecting Accra to Kumasi by 1923. The port-to-mine investments had another effect: opening up large tracts of fertile land. Fine spatial data on location and production reveals that the railways led to higher cocoa production and export. The impact was huge: around 30 per cent of cocoa production, or 4.5 per cent of GDP. The paper supports the idea that economic activity is path dependent: today, despite the fact that roads are more important, the areas around the railways have the largest manufacturing and service sectors.

With infrastructure so important in Africa, getting the politics right is vital too, as a paper on Kenyan road investment presented by Robin Burgess showed.¹³ Roads are the single largest public investment item, so the authors used historical road maps to construct a data set on the location of roads between 1963 and 2002. By combining this with data on the ethnicity and local origin of Kenya’s leading politicians, and with census data, the authors were able to test for ‘ethno-favouritism’. They found it: politicians tend to funnel cash for infrastructure back to their own districts, or to districts where their tribe (often Kikuyu, Luos or Kalenjin) is dominant, rather than to areas where the economic gains would be greatest. The effects this ethnic bias are big — favoured regions get close to five times more roads than their predicted share. But it is not all gloomy. Kenya has experienced lots of regime changes — flipping between single party rule and coalitions. And when a multi-party democracy is in charge, leaders do not just favour their own ethnic groups, but direct investment towards other unrepresented groups too. It is a rare call for more coalition politics.

The social network
If a conference is as a platform for making connections then the Royal Holloway provided a great one: a striking campus, generous food, drink and entertainment. On the final morning Matthew Jackson of Stanford University presented a paper on social networks. Since networks — the webs of relationships humans are embedded in — act as information conduits a person’s network determines whether they know things like the location of job opportunities, or about new technology. By transmitting this information, networks influence economic decisions.

But the quality (how quickly and how far) of transmission depends on the connectivity of the network. Mr Jackson showed nodes and link diagrams and outlined how different ‘centrality’ statistics — essentially measures of power or influence — work. Degree centrality is like a popularity measure: a simple sum of the connections a node has. Another measure — eigenvalue centrality — is more like a measure of how powerful your friends are (it takes into account your friends’ centrality, the centrality of their friends, and so on). Jackson proposes a new measure ‘diffusion centrality’: here a node's influence is measured by how far information travels across a network in a given amount of time after it hits the initial node.

Jackson examined the take up of microfinance in 75 villages in Karnataka, India. The basic puzzle is that adoption varies widely, from 7 per cent in some villages to 44 per cent in others that are otherwise (in terms of demographics, and other explanatory variables) very similar. After mapping the villages’ networks Jackson examined the relative importance of two types of transmission through each of the village networks. They found that basic information passing (i.e. knowing that there are going to be loans available in the village) is much more important than peer influence (taking loans because friends are). Word of mouth is particularly strong when villagers were users: people are ten times more likely to tell their friends about microfinance if they participated in it rather than just knowing about it. In addition, because there was random variation between the connectivity of the ‘injection points’ — the teachers and shopkeepers that were first told about the loans by the bank providing them — different diffusion models could be tested. The results suggested that injection points matter quite a lot. In finding the best person to tell about a new product or policy it is not enough to find the most ‘popular’ person: simple degree centrality does not explain diffusion well, the eigenvalue and diffusion measures do much better.

Productivity and policy
Nowhere is Britain’s economy odder than in the combination of slumping GDP and strong employment seen in the last five years. The combination-more workers, less output-means a big drop in labour productivity. Claire Crawford of the IFS started the last session of the conference with a detailed review of the data, focusing on wages.¹⁴ The paper shows the average wage is down 3.5 per cent since the start of 2008 (by contrast real wages continued to grow in previous recessions). Ms Crawford showed that compositional effects (replacing expensive workers with cheaper ones) do not explain the drop. Rather, unprecedented numbers — 70 per cent — of workers have taken real wage cuts.

This could be due to welfare reforms (rules that tie benefit payments to job search, boosting the potential workforce and keeping wages down) or weaker unions and wage bargaining. It is a view that sat well with Paul Gregg’s presentation. He put low real wage growth into a longer run perspective: it had started pre-crisis. He argued that this would be the norm, with weak wages lasting thing, even when the economy recovers. Both papers were interesting positive analyses, essentially recasting the productivity puzzle as a weak wage puzzle.

But Britain’s productivity puzzle is a question that requires normative analysis too. For example, if it is due to a permanent loss of supply capacity, then demand boosting policies will be self-defeating, leading to higher inflation. John Van Reenen’s paper grasped the policy nettle. Van Reenen started by observing that firms had faced a spike in the cost of capital (despite Bank Rate being low, the rates that firms pay have been high). That had reduced the quantity of investment. The quality of capital allocation was lower too, with many underperforming loss-making firms kept alive by banks willing to forbear on non-performing loans. This meant a big cut in the effective capital per worker. It also implied that there had been no large fall in efficiency (total factor productivity) in this recession compared to earlier severe downturns.

If Van Reenen is right, the argument that expansionary monetary or fiscal policy will simply result in higher inflation is wrong. It means schemes like the Bank of England's Funding for Lending scheme, which now aims to channel bank credit to the SME sector should potentially be even bigger, and public investment prioritised by HM Treasury. It is an area where more research, papers and debate from the RES’s economists are badly needed.

Changes at the top
Outgoing RES president Richard Blundell gave the presidential address on inequality. He had been warmly introduced by Imran Rasul who thanked Richard for his vigorous support for research, his colleagues and students. The RES presidency passes into safe hands when Charlie Bean takes over. Are there any changes he should make to ensure that the rush of new blood economics is experiencing is well used? Having mulled over all the papers, I was left with three ideas.

Academic economists are working on the right questions, but sometimes duck giving a full answer. I was impressed by the range of studies on current policy problems, many mentioned in this write up. But Manski’s call for a practical economics — prescriptive as well as descriptive — is important. At times, economists slip into writing clever empirical papers and leaving out the policy implication. At its worst, academia can slip into pointless self-referential debates as the famous ‘two Cambridges’ spat shows. Of course, papers developing technical methods are vital. But perhaps an explicit bias towards using theory and econometrics to practical ends wherever possible is justified. It is the policy questions that attract students to economics in the first place, and it is the policymaker’s question to the professional economist — ‘So what?’ — that matters.

Finding the right answers requires joint work with non-economists. This is a common observation, as ‘interdisciplinary’ papers become more common. Matthew Jackson showed why it is an important development: he worked with computer scientists and specialists in graph theory to crack social networks. Elsewhere, Andrew Haldane and Robert May have combined finance and biology to understand financial networks,¹⁵ Hal Varian is combining microeconomics with machine learning.¹⁶But these are the exceptions: most economists still operate in silos. And the structures in economics (including the conferences) often don’t help. I was sad to see that economic history was confined to a separate conference, in a different location, a few days after the RES. Capacity constraints may explain this, but some innovation is needed to bring researchers together.

Once economic research is complete and findings written up, getting the message out is a separate task. It requires yet more work, different skills and infrastructure. I'm not sure that European economics departments realise this. It is a simple fact that American departments are far quicker at dealing with enquiries about their findings and setting up meetings with their academics. Some European departments make an effort to be contactable-Warwick provides a useful ‘experts directory’. But getting through to others can be impossible. For those that are too busy to talk, shorter opinion pieces based on journal-length papers are helpful, as are portals like VoxEU. But there are still very few good European academic blogs. Anything that could change any of this would be welcome.

These are practical problems. They are simple to rectify. The RES conference was fascinating. It is a great time to be an economist.

Notes:

1. See ‘The disposable academic’ The Economist 10th December 2012 http://www.economist.com/node/17723223 and Noah Smith, ‘If you get a PhD, get an economics PhD’, May 2013 http://noahpinionblog.blogspot.co.uk/2013/05/if-you-get-phd-get-economics-phd.html

2. Mr Haldane explained that this is why the Bank of England had recently set out a four-pronged loosening package. This includes bank funding (extended term liquidity), lending (the Funding for Lending scheme), liquidity (lower liquidity requirements) and capital (zero capital charges on lending to support the Funding for Lending scheme).

3. Paolo Buonanno, Ruben Durante, Giovanni Prarolo and Paolo Vanin, ‘Poor Institutions, Rich Mines: Resource Curse and the Origins of the Sicilian Mafia’. http://bit.ly/10Nb1Il

4. Michèle Belot & Marina Schröder, 2013. ‘Does Monitoring Work? A Field Experiment with Multiple Forms of Counterproductive Behaviour’, FEMM Working Papers 130006, Otto-von-Guericke University Magdeburg, Faculty of Economics and Management. http://bit.ly/14sxNCy

5. Colin P Green and María Navarro Paniagua, 2013. ‘Play Hard, Shirk Hard? The Effect of Bar Hours Regulation on Absence’. http://bit.ly/134Vt4c

6. Elaine Kelly and Imran Rasul (2013). ‘Policing Cannabis and Drug Related Hospital Admissions: Evidence from Administrative Records’. http://www.ifs.org.uk/publications/6121

7. Manski (2013), ‘Partial identification and policy choice under ambiguity’, RES Sargan lecture 2013. 

8. See Manski (2013), "Diagnostic testing and treatment under ambiguity: using decision analysis to inform clinical practice", Proc Natl Acad Sci, 110(6) pp2064-9.

9. This is something some British policy institutions do already: the Bank of England uses interval forecasts, and HM Treasury uses upper and lower bounds for the impact of policies.

10. The Commission's members were Philippe Aghion, Tim Besley, John Browne, Francesco Caselli, Richard Lambert, Rachel Lomax, Chris Pissarides, Nick Stern and John Van Reenen.

11. The IMF mentioned the LSE Growth Commission report when they issued their Article IV statement about the British economy on 22nd May. http://www.imf.org/external/np/ms/2013/052213.htm

12. Jedwab, Remi and Moradi, Alexander (2012), ‘Revolutionizing transport: modern infrastructure, agriculture and development in Ghana’. http://eprints.lse.ac.uk/42263/

13. Robin Burgess, Remi Jedwab, Edward Miguel, Ameet Morjaria, Gerard Padro i Miquel (2013), ""The Value of Democracy: Evidence from Road Building in Kenya". http://bit.ly/14kEer4

14. Richard Blundell, Claire Crawford and Wenchao Jin, ‘What can wages and employment tell us about the UK's productivity puzzle?’ http://www.ifs.org.uk/wps/wp201311.pdf

15. Andrew Haldane and Robert May (2011) "Systemic risk in banking ecosystems", Nature 469, 351-355. http://www.nature.com/nature/journal/v469/n7330/full/nature09659.html

16. See The Economist, ‘Micro stars, macro effects’, November 2012. http://econ.st/UrovUv