Sunday, 30 May 2010

The economics zeitgeist, 30 May 2010

This week's word cloud from the economics blogs. I generate a new cloud every Sunday, so please subscribe using the RSS or email box on the right and you'll get a message every week with the new cloud.


I summarise around four hundred blogs through their RSS feeds. Thanks in particular to the Palgrave Econolog who have an excellent database of economics blogs; I have also added a number of blogs that are not on their list. Contact me if you'd like to make sure yours is included too.

I use Wordle to generate the image, the ROME RSS reader to download the RSS feeds, and Java software from Inon to process the data.

You can also see the Java version in the Wordle gallery.

If anyone would like a copy of the underlying data used to generate these clouds, or if you would like to see a version with consistent colour and typeface to make week-to-week comparison easier, please get in touch.

Friday, 28 May 2010

Google GDP

There's no specific point to this picture - feel free to draw your own conclusions.

It comes from a neat feature from Google, which I suspect has been around for a while if I'd looked for it earlier. Here's the link if you want to play.

Wednesday, 26 May 2010

HIPerbolic discounting

Latest in my series of bad puns on the word "hyperbolic".

HIPs are being scrapped by the new government. This will provide relief to millions of annoyed estate agents ad house sellers everywhere. But why?

HIPs should be a gift from the economics profession which is joyfully accepted by buyers, sellers and agents. The reason? Because houses are a classic "market for lemons".

A simple explanation of what this means and what the consequences are: The seller of a house knows much more about it than the buyer. Therefore, if the seller thinks it is worth £200,000, a buyer should apply a discount to take account of the risk that something is wrong. Let's say they apply a 10% discount and offer £180,000. After the discount, some sellers who do not have anything wrong with their house will prefer to stay rather than sell for less than it's worth. Other sellers, who do have a skeleton in the cupboard, will keep their house on the market. The average quality of the available houses declines. Therefore, buyers look at this situation and decide to apply a larger discount. They adjust their bids down even further, even more sellers withdraw, and so on. The market enters a death spiral.

To combat this problem, buyers would traditionally commission a survey of the property to make sure it is sound. Sensible, but expensive. A structural survey costs perhaps £700, and only after it is complete would the buyer decide whether to bid on the property. Even if they do bid, they might still not win. So a conscientious buyer might end up paying for three or four surveys before successfully acquiring a house. This adds a big risk - and transaction cost - to the process of buying, and must reduce the number of available bidders for each property.

HIPs were designed to solve this problem. Instead of multiple buyers commissioning a survey, the seller would do it once. Only one surveyor would be needed instead of several. And each buyer would know, in advance, the state of the property and could decide whether to bid without incurring a entry cost to do so.

Beautifully logical. It doesn't even impose any net cost on the seller, because they can simply factor the cost of the report into their asking price. And it must surely increase the number of bidders on each property, which would increase the sale price achieved. So why did sellers and estate agents scream blue murder as soon as the subject was raised?

First, the superficial claims:

  1. The buyer won't trust the report anyway, because it was commissioned by the seller. Only true if the seller does not use a respected firm of surveyors. There are plenty of big surveyor firms which have enough financial strength to be sued by a buyer if the survey turns out to be false. And there are plenty of insurance companies willing to participate in this market if people don't think the surveyor will have enough money.
  2. The buyer can't sue the surveyor, because the surveyor's client is the seller. Do these people really think it's beyond the wit of the UK's legal profession to design a contract structure which transfers this responsibility? This is a nation with more property lawyers per capita than any other in the world.
  3. The HIP is no use because it doesn't include a structural survey. True, but only because the estate agents lobbied for it not to! The original proposal was to make the structural survey mandatory. So shall we go back to that instead of dropping the scheme altogether?
None of those reasons are convincing. Why then do people really dislike this idea so much?

There are a few different reasons for these objections. (The presence of multiple causes is an interesting theme, which I'll analyse in a posting soon).


  1. Hyperbolic discounting. Although there are direct benefits to the seller in commissioning their own survey, the benefits come later - after the house is sold - while the cost comes up front. Therefore the cost gets much more weight.
  2. Availability heuristic. Even though sellers know, when they think it through, that they will gain from the scheme, the benefits are never directly visible to them. No seller will ever know that they achieved a price of £208,000 when they would otherwise have only got £202,000. They have to rely on economic theory to tell them. But that £700 is coming out of their pockets right now and is painfully concrete.
  3. Slippery slope. This episode reveals something that most lay people do not know. Selling a house is, often, a surprisingly impulsive decision. People will walk into an estate agent off the street and ask for a valuation "just out of curiosity". The estate agent knows this very well, and knows that once the valuation is done, they are more than halfway to persuading the homeowner to put the house on the market. Each step is a small one, and anything - like a £700 outlay - which might interrupt this process will put a damper on the market.
  4. Badly designed agency incentives. Estate agents are primarily rewarded just for selling a house. In fact, though, it's incredibly easy to sell a house. Just put your £200,000 property on the market for £150,000 - it will be gone in no time. The difficult part is getting someone to pay £220,000 for a £200,000 house. So really, that's what the agent should be paid for doing. Instead of charging 1.5% of the purchase price, they should be paid nothing for achieving the first 80% of the valuation, 5% of the gap between that and the valuation, and 10% of anything above the expected value. That would encourage the agent to actually care about the value the seller achieves, and not just to pump through as many transactions at as high a speed as possible. From the buyer's point of view, they want to find a property offering the maximum consumer surplus - the biggest gap between purchase price and happiness obtained. I'm not sure how to reward an agent for that, but a loyalty bonus based on how long the buyer stays in the house might be a good mechanism.

    All of this means that estate agents have different incentives to sellers, and made it their business to persuade sellers that this would kill the market. There is a grain of truth in that claim, because of item 3 in this list, but agents, in this campaign, are primarily serving their own interests and not that of their customers.
All of these things combine to make the market fail. And together, they have killed a proposal which is an economic no-brainer - a policy which would reduce waste, and help buyers and sellers to get what they both want at the right price.

Cognitive biases suck, don't they?

Tuesday, 25 May 2010

Do we ever make ANY difference at all?

This fascinating analysis from Tino Sanandaji (via Marginal Revolution) reveals an unexpected (to me) fact about the Japanese "lost decade(s)".

Japan's growth from 1990 to 2007, adjusting for changes in demographic composition of the population, was exactly the same as the US and Europe's!

So for all the attention on the US technology boom, European integration, Japan's property bust and vast stimulus programmes - none of it mattered. Growth per working-age individual seems to proceed at a steady 1.6% per annum regardless of the different policy regimes, tax rates, interest rates etc etc etc.

This sheds interesting light on the claim of Adam Posen (Bank of England monetary policy committee member) that Britain faces a Japan-style decade. So does the surprising fact that the UK was the leader among all rich countries in total factor productivity growth from 1990 to 2007!

So is all of economic growth explained by some common underlying factor, unrelated to economic policy? Let's say a bit of technological progress, plus the bankruptcy or shrinking of the least competitive firms each year. Maybe that is enough to ensure our society and wealth continue to progress.

And talking of technological progress - maybe it's relevant to point out this statement from Scott Sumner:
In 1927 Lindburgh flew a primitive airplane over the Atlantic. By 1968 we had the Boeing 747. (In 1969 the moon landing.) In the next 42 years commercial airplanes hardly advanced at all.
And yet, in 1970 the airline industry sold 550 billion passenger-kilometres. In 1995, 2.5 trillion. In 2010, about 5 trillion. So in this 42 years of making no advances, the airline industry somehow managed to provide ten times as much air travel and, presumably, ten times as much positive impact on people's lives. Maybe inventing new kinds of plane isn't what drives growth after all.

More on that in another article.

Monday, 24 May 2010

Economic flexibility or urban blight?

Having questioned the flexibility of the British economy in my last post, I came across an article which made me wonder the same about America:
...the city took a marked turn for the worse. We passed vacant lots surrounded by listing barbed-wire fences, the shells of abandoned mini-malls and the cement outlines of what might once have been petrol stations, now being reclaimed by bushes and grass. There were lots of these ‘urban fields’, where buildings had been torn down and nothing built in their place, and it was curiously unsettling to see how easily sizable chunks of a city could be erased and swallowed up once humans left.
...the neighbourhood reached its nadir – no small feat – with boarded-up houses, a couple of burned-out shops being used as squats, a low, forbidding bar called ‘Club Rolex’ and, saddest of all, the brick skeleton of what had been George Rogers Clark Junior High School. It stood like the shell of a dissolved monastery behind a high wire fence in an asphalt yard grown into a meadow of waist-high weeds, its roof gone, every window smashed and its walls slowly collapsing.
This - on the face of it - sad-sounding story provokes a question: if demand for services and facilities changes over time, what should happen? In a flexible economy, presumably the old infrastructure may well shut down. Presumably the kids who would have gone to that school are now attending some other school instead. People are buying their groceries from another store in another neighbourhood.

But this example does show the downside of flexibility. Resources that could have been adapted for a different use are lying idle instead. The people who do still live on those broken-down blocks have fewer local services and may not be able to participate meaningfully - either as employees or as consumers - in a local economy any more, due to the challenge of physically travelling to where the businesses are.

Presumably the benefits to the businesses - and to their present customers, if not those from their past - of locating in a different place outweigh the savings they'd have achieved by reusing the old buildings and land. Probably, it's economically efficient for these places to be abandoned. But the empty shops also show that there's a certain lack of flexibility in the local economy.

This is called granularity - the fact that economic activity can't, in practice, be divided up into arbitrarily small units. High thresholds need to be achieved before some activities can take place. For instance, to start a business using one of those shops, someone probably needs to be able to sign a medium or long-term commercial lease. A certain critical mass is required to occupy a full time employee running a shop, a hairdresser or a pool hall. Security and other concerns make it difficult for someone to come up with a casual, small-scale, experimental use for the old school building.

The outcome: there are probably some people living in this neighbourhood with no work to do, and others who can't find things they can afford to buy, or leisure activities they can afford to participate in - and all the while, buildings and capital sit idle. This is the very definition of a market failure, and it shows, on a smaller scale, why a Keynesian high-unemployment equilibrium can exist.

So one kind of flexibility - the dynamic American economy which shifts people and commerce out of one region and into another - exposes the lack of another kind. What's missing from this loop of East St. Louis is the stable community that would allow small-scale entrepreneurship to pop up in the gaps between large organisations. In other places - the island or village communities of rural Britain, for instance - it's easier for small-scale activities to sustain themselves, when the spare time of the milkman combines with the early evenings when the village school is empty, and the pensioners who want to contest a game of bridge.

Is one system better than the other? Not obviously. Economic production as measured by money is higher in the American system, but perhaps quality of life is better in the country village. Both systems work in their own way, and humanity - with the help of niche-promoting technology like the Internet - is gradually pulling itself up the hill to a better combination of the two.

How's the UK economy doing?

Some news has been creeping in which gives us a picture of the British economy that's gradually becoming clearer. First, from the ONS via David Smith:
Business investment for the first quarter of 2010 is estimated to be 6.0 per cent higher than the previous quarter.
From the same source, some news on public borrowing:
...in December, the Treasury preducted public sector net borrowing for 2009-10 of £178 billion on the definition it uses. The latest figures from the Office for National Statistics show a downward revision to the outturn, taking it to £156 billion...At one time people were predicting figures as high as £220 billion.
Recent growth estimates are decent (though not outstanding) too:
Gross Domestic Product (GDP) increased 0.2 per cent in the first quarter of 2010, compared with an increase of 0.4 per cent in the previous quarter.
There was a concern prior to the release that growth might be negative, due to heavy snow in January and February. So growth of 0.2%, though hardly vibrant, was a relief. And growth figures for the previous two quarters have been revised upwards by 0.2-0.3% after the initial release - we can't be sure that will happen this time, but in any case I think we can expect strong figures in the second quarter.

Finally, there's a surprising leap in inflation over the last few months:
CPI has risen from 1.1% in September to 3.7% now and the Bank will be hoping its reversal is just as dramatic. RPI inflation has gone from minus 1.4% to 5.3%. RPIX, the retail prices index excluding mortgage interest payments, is now at 5.4%, more than double its old target of 2.5%. CPI is a long way from its 2% target.
So what is happening? Two things, it seems to me.

First, demand is returning. The property market has recovered well from the depths of the recession, business investment is increasing and consumer demand has held up well because unemployment did not rise as much as feared. This is helping to fix the public deficit - in addition to which, that 3.7% inflation figure reduces it by the equivalent of another £6 billion (exactly as much, coincidentally, as the public spending cuts announced today).

Second, supply is constrained. Because unemployment only rose a bit, there is not a lot of spare capacity (apart from some marginally productive ex-employees) to keep prices down. Hence the inflation figures. This does not speak well for the flexibility of the British economy, though it has helped to sustain the momentum of aggregate demand. If UK companies don't find some hidden reserves of productivity, we won't be able to get the surge of growth that I hope is coming in the recovery. (Note, however, that the figure is not quite as bad as it seems - over a third of the inflation figure comes from January's VAT rise, and the total without it is only about 2%).

I am not sure whether I'd rather have a US-style economy, where a jump in unemployment in the recession is followed by fast growth in the recovery, or a French-style one, where employment and demand are sustained at the expense of future growth. I hope that the inflation is only an outcome of quantitative easing and the weakness of the pound, and that we'll be able to get both decent growth and a recovery in jobs over the next year.

Posting partly inspired by Nick Rowe's question on TheMoneyIllusion.

Sunday, 23 May 2010

The economics zeitgeist, 23 May 2010

This week's word cloud from the economics blogs. I generate a new cloud every Sunday, so please subscribe using the RSS or email box on the right and you'll get a message every week with the new cloud.


I summarise around four hundred blogs through their RSS feeds. Thanks in particular to the Palgrave Econolog who have an excellent database of economics blogs; I have also added a number of blogs that are not on their list. Contact me if you'd like to make sure yours is included too.

I use Wordle to generate the image, the ROME RSS reader to download the RSS feeds, and Java software from Inon to process the data.

You can also see the Java version in the Wordle gallery.

If anyone would like a copy of the underlying data used to generate these clouds, or if you would like to see a version with consistent colour and typeface to make week-to-week comparison easier, please get in touch.

Friday, 21 May 2010

Dublin up

I'll be on the panel next Friday afternoon at Behavioural economics, policy and business at the Geary Institute in University College Dublin. I'm not sure whether they still have room in the audience, but if you're interested in coming please contact me or Emma Barron.

The Geary Institute has a very good behavioural economics blog which you'll see in the right-hand column of this site. I'd recommend you subscribe to it if you haven't already (there is, naturally, a focus on Ireland-specific issues, some of which are distinct from the behavioural research they also cover).

Also on the panel are Pete Lunn, author of Basic Instincts: Human Nature and the New Economics and Gerard O'Neill of Amarach Research, an interesting market research agency (see for example this recent blog post about pricing), and Liam Delaney and Colm Harmon whose work you can see on the Geary blog.

Hope to see some readers (and pick up some new ones) there.

Wednesday, 19 May 2010

Easterly's too weak

When economists say they don't know much, I'm always sympathetic. But when they say we will never know anything, that's going too far.

William Easterly presented an idea at the LSE tonight: "We don't know how to solve global poverty - and that's a good thing" (thanks to Andrea James for alerting me to this).

Part of his logic is impeccable. He argues that democracy, freedom and self-determination are normative goods - and thus, even if we could solve development problems by imposing top-down, autocratic solutions, we should not. Later, he shows some tentative evidence that authoritarianism doesn't even provide economic benefits - at least there's no convincing evidence for it, and there's some to suggest the opposite.

So far so good - it's nice to have an argument against the position that countries need (at least temporarily) an autocratic system to kick-start economic growth. The four examples that are often used to support this argument - Singapore, Taiwan, South Korea and China - do not provide clear evidence at all, according to Easterly.

But he jumps too far from here to a dangerous conclusion. Because autocracy does not work, he says, this means there are no universal solutions in development. We can only rely on local, patchwork approaches.

I don't see how this follows. It's likely that there are still some universal principles - for example clear property rights, free primary education, certain kinds of infrastructure, or openness to foreign investment (or indeed the reverse) - which work everywhere. And while it's fair to say we should not impose these principles from the top, this doesn't mean we shouldn't argue for them, provide evidence and try to persuade people to adopt them.

Just as there are universal discoveries in physics or biology - and indeed economic laws like supply and demand - which still seem to work even with no legal authority to enforce them, there may well be principles of growth and development which we can discover and recommend that poor countries adopt. If we can make a discovery in one place that will help people in another, it would be a dereliction of duty for us not to do so.

The right answer is a blend. Any farmer knows that their strategy must be governed by local conditions - the flatness of the ground, the local weather, the indigenous animals and insects will all influence their choice of crop and how, where and when they plant it. But they can still take advantage of scientific discoveries - about fertilisers, temperature or crop rotation - which are made on the other side of the world and apply everywhere. Development is just the same. A mix of local and global discoveries will work.

It's great that development economists recognise that paternalistic attitudes don't work. But we mustn't use this as an excuse to walk away from the responsibility we all share as humans to make the world a better place.

Update: Easterly comments here on Dani Rodrik's argument: "democracy, globalization, the nation state: pick any two". Not sure that Rodrik is suggesting democracy and development are not compatible, but maybe I'll get a copy of the book when it's out.

Sunday, 16 May 2010

The economics zeitgeist, 16 May 2010


This week's word cloud from the economics blogs. I generate a new cloud every Sunday, so please subscribe using the RSS or email box on the right and you'll get a message every week with the new cloud.


I summarise around four hundred blogs through their RSS feeds. Thanks in particular to the Palgrave Econolog who have an excellent database of economics blogs; I have also added a number of blogs that are not on their list. Contact me if you'd like to make sure yours is included too.

I use Wordle to generate the image, the ROME RSS reader to download the RSS feeds, and Java software from Inon to process the data.

You can also see the Java version in the Wordle gallery.

If anyone would like a copy of the underlying data used to generate these clouds, or if you would like to see a version with consistent colour and typeface to make week-to-week comparison easier, please get in touch.

Friday, 14 May 2010

Not forecasting the coalition

I recently re-read a remarkably prescient article from February claiming the Liberal Democrats would not enter a coalition with the Tories.

It gets exactly right the entire set of conditions the Lib Dems would demand for a coalition. The only error - thinking that the Tories would say no.

The key reason that the authors think a coalition could not work?
Clegg is opposed to forming a coalition because aides and senior MPs argue it would be highly dangerous for the Liberal Democrats to become minority partners in a coalition government on the grounds that the majority party could manipulate the timing of the next election to suit it. The Lib Dems have long campaigned for fixed terms at Westminster to deprive the prime minister of the initiative on election timing.
And so we ought to look quite carefully at the 55% threshold for dissolution of Parliament, and instead of assuming it's a Conservative attempt to entrench power, think about the Lib Dems' motivations. In politics, as in economics - and murder mysteries - the key question is often: cui bono?

Tuesday, 11 May 2010

So who won? Resolving the democratic dilemma

Regular readers may be relieved to hear that I'll be back onto regular economics soon. If nothing else, even I will be bored of the election before long.

But I have a couple more political articles to come first.

The debate over forming a coalition to govern the UK seems to be mainly focused on the question of "who won?" Was it the Tories, because they are the largest single party with 36% of the vote (and 47% of the seats)? Was it a combination of Labour and Liberal Democrats, because together they represent over 50% of people? Or, similarly, the Conservatives and Liberal Democrats, who have about 59% between them? Should we reform the voting system so that the majority of votes always reflects a majority of parliamentary seats?

All this debate ignores a deeper point. This is not a football match. The electoral system is not intended as a talent contest to award a prize to whoever can perform better and impress more people. It's meant to provide a way for the preferences and needs of the population to best influence the decisions of society as a whole.

Even had the election been decisive, and were there no third party to confuse the issue, we'd have (say) 55% of votes and 65% of seats held by one party. Sure, a clear victory. But does this mean the wishes of 45% of the population should be ignored?

The Tories have certainly "won" in some sense, with 36% of votes and a clear swing in the nation's preferences towards them. Of course, 30% would have voted Tory no matter what, but 6% of people clearly switched their vote from Labour to the Conservatives. That 6% have effectively decided the election (or not quite decided it, as is becoming clear).

So if Labour and the Liberal Democrats do form a coalition (though this now looks unlikely now due to an improbable alliance between Labour backbenchers and Conservative newspaper owners) they have to acknowledge that a big chunk of the country has voted Tory. The choices of these people should be listened to.

So, of course, should the choices of the 64% who voted another way. But how can we listen to both sets of people, when their policy preferences are often opposed?

What is clear that any government which follows the will of 55% of the people and ignores the preferences of the rest is not maximising societal utility. Of course, the 45% deserve less influence than the 55%, but not none at all.

There is no obvious way to achieve this with a purely representative system, and especially not a geographical one. Compromises are possible - a Lab-Lib coalition with Ken Clarke as business secretary, say? Not very practical, I fear. But there could be another way.

In economics, we don't expect the provision of shops, products or services to be governed by majority rule. Just because more people use Tesco than any other supermarket does not mean Sainsbury's, Asda and Morrison's - and that deli round the corner - can't exist too. Even if Tesco and Sainsbury's together could get over 50% of the market, they would not get to shut the others down.

Coca-Cola might have over 50% of the UK cola market, and your local pub is more likely to sell you Coke as a result, but Pepsi, Tango and milk still exist.

Instead, people choose where to allocate their resources and any conflicts or competition are resolved this way. If you and I both want to live on a particular piece of land, we simply bid for it and whoever's willing to pay more gets it. If I win because you'd rather spend your money on Pepsi instead of on a bigger house, you can do that. Ultimately, somewhere in the world, because of our choices, a piece of land will be used to grow sugar cane or provide accommodation for a family.

In this way, whatever's important to you gets your attention and your resources. And we all get a better deal as a result. I can trade what matters to me against what matters to you, and we will both end up with more of what we want. The world's limited resources are allocated efficiently.

Now, what if we could do the same with policies?

Personally, I don't feel strongly about whether Trident is renewed or there's a tax on £2 million houses. On the other hand, I'd really like a permissive policy on immigration, closer integration with Europe and a bit more monetary stimulus. I don't mind taxes staying where they are, or going up a bit. But I think money in some parts of the health service could be spent a bit more efficiently through market mechanisms, freeing up some resources for better education.

There's no party I can vote for which expresses these views. But I'd be quite happy to trade with someone who feels strongly about different things, so we can both get a bit of what we want.

I'm not calling for a Swiss-style referendum system - that's too simplistic, and leads to policies being considered in isolation without taking into account the trade-offs between them. Indeed the strength of the party system is that each party has put forward in its manifesto a particular combination of trade-offs which they feel will lead to a better society.

But it's clear from the result that none of the packages on offer are widely supported. And this will always be the case - until we put some kind of scoring mechanism on policies, so that people can show how much they really care about different things.

So what I propose is a points system. We each get 100 tokens to allocate to different policies. Each of these policies has a price, and we choose which ones matter to us.

How is the price set? By the tokens of those willing to bid against it. If I pay 50 to stop Trident being renewed, you need to pay at least 51 to get it.

Of course these things bring financial costs too, and if Trident is approved then about £3 billion a year in taxes will have to be found to pay for it. These are intrinsic parts of the policies, so you can't bid for Trident without also being willing to accept the cost of it. If you've spent all your tokens on Trident, you may find that someone else's tokens get to decide whose income tax is going to pay for it.

A system like this would need lots of technology to implement it - an eBay-like bidding system is an example of a similar, though simpler, algorithm. And perhaps most voters, at least today, would not be ready for this level of complexity. So I'd still be happy to see a representative democracy where our elected MPs cast these votes on our behalf - especially if they're given a formal duty to represent the broad interests of their constituents and not of their parties.

A simple version of this system could even be used now, to produce a policy programme for a coalition government. It seems reasonable to believe that:

  • 36% of people want a smaller government, a cap on immigration and disengagement from Europe, and more local control over schools and hospitals
  • 29% of people want roughly the status quo, with an emphasis on equality and help for the poorest, and a fairly strict defence and justice policy
  • 23% of people want an immigration amnesty, pro-European policies, substantial redistribution in the tax system and more civil liberties
With this knowledge, it shouldn't be impossible to bash out a policy programme that roughly balances the priorities of the electorate.

In return for more redistribution, Labour voters would probably be willing to give up their party's security policies and tight central control of public services. Liberal Democrats would have little choice but to give up on Europe and immigration, but could get their way on civil liberties and some tax reforms. And Tory voters would be able to cut public spending and enhance local control through private involvement in some aspects of education and health, but might not be able to get their way on the form of tax cuts.

European policy would probably remain centred around the centrist position that Labour has implicitly pursued, with a bias towards the Tories' preferences as they won more votes than the pro-European Liberal Democrats.

Sure, the complexity of the horse-trading involved in this kind of compromise would make the last few days look like a game of Countdown. But compromise is what we elect politicians for.

A proportional representation system would still be an improvement on the current situation - it would help to make these negotiations more balanced, as the number of MPs putting forward each position would better reflect the electorate's choices.

But the core problem with any winner-takes-all system - FPTP or coalition - is that there is no single identity for the country, and nobody has made a single decision. There is no one thing that the population "wants". Indeed, Kenneth Arrow's famous Impossibility Theorem says that there is no way to perfectly reflect the preferences of a population in the decisions of a government. But a system that accurately transmits voters' wishes into a policy compromise would surely be a step closer to the ideal.

Of course nothing like this is going to happen in the near future. But I hope that whoever ends up in power will have the wisdom and restraint to govern in the interests of the whole country - not just the minority, or even majority, who has chosen them.

Sunday, 9 May 2010

The economics zeitgeist, 9 May 2010



This week's word cloud from the economics blogs. I generate a new cloud every Sunday, so please subscribe using the RSS or email box on the right and you'll get a message every week with the new cloud.


I summarise around four hundred blogs through their RSS feeds. Thanks in particular to the Palgrave Econolog who have an excellent database of economics blogs; I have also added a number of blogs that are not on their list. Contact me if you'd like to make sure yours is included too.

I use Wordle to generate the image, the ROME RSS reader to download the RSS feeds, and Java software from Inon to process the data.

You can also see the Java version in the Wordle gallery.

If anyone would like a copy of the underlying data used to generate these clouds, or if you would like to see a version with consistent colour and typeface to make week-to-week comparison easier, please get in touch.

Friday, 7 May 2010

Behavioural politics: the story so far

The election results are almost in, and Nick Clegg has just made a brilliant tactical intervention. Peter Mandelson must be jealous.

But more of that later. First, a look back at the days of the campaign. Click each of the links below to see what happened on that day.

You'll remember that I rated the three parties each day on their behavioural nous. But who came off best throughout the campaign? The ratings are listed against each day below, and at the end you can find out the totals (ratings are given in order Conservative, Labour, Liberal Democrat).

Day 1: 7, 7, 5
Day 2: 8, 5, 4
Day 3: 6, 7, 4
Day 4: 6, 6, 6
Day 5: 8, 7, 5
Day 6: 6, 8, 7
Day 7: 6, 7, 6
Day 8: 6, 6, 5
Day 9: 5, 6, 8
Day 10: 3, 5, 9
Day 11: 5, 6, 8
Day 12: 5, 6, 8
Day 13: 6, 6, 5
Day 14: 5, 6, 7
Day 15: 6, 5, 6
Day 16: 7, 4, 6
Day 17: 6, 4, 7
Day 18: 7, 5, 6
Day 19: 7, 5, 8
Day 20: 7, 6, 5
Day 21: 8, 5, 6
Day 22: 4, 5, 6
Day 23: 6, 3, 7
Day 24: 6, 7, 6
Day 25: 8, 6, 4
Day 26: 8, 5, 7
Day 27: 8, 6, 6
Day 28: 6, 9, 7
Day 29: 7, 6, 8
Day 30: 7, 7, 7

The average ratings throughout the campaign are, therefore...

  • In third place, Labour with 5.87
  • In second place, the Liberal Democrats with 6.30
  • And just making it into first place, the Conservatives with 6.33
This is nothing to do with policies or the genuine merits of each party: it is simply an evaluation of how well the parties have run the psychology of their campaigns.

And how has this manifested itself in the electoral outcome? Not in number of votes or even seats, but in terms of the power that's become apparent today. I think the three parties show up in exactly the order that the ratings indicate.

If you're new to this series, please start with Day 1 above and carry on through.

Thursday, 6 May 2010

Behavioural politics, days 25-30 of 30

Friday once again was a day for the aftermath of Thursday night's debate.

Unforuntately right now, I have no time to discuss it as I'm very busy with the actual election. This is also why updates have been behind for the last week.

However, as a precommitment device to combat my own tendency to post hoc rationalisation, here are my summaries and ratings for the last six days of the campaign.

Day 25 (Friday): Cameron builds momentum on the back of "winning" the debate (see day 24 for explanation via confirmation bias). Social proof from other voters even more important than from head of M&S. Tory: 8/10, Labour: 6/10, Lib Dem: 4/10 (Clegg had most to lose as the unofficial king of debates).

Day 26 (Saturday): Newspaper endorsement day. The Guardian switched from Labour to Liberal Democrat - a tiny bit sensationalist I think, but their reasoning was understandable. Most others picked the Tories. Social proof again, and the momentum effect continuing to work for Cameron. Tory: 8/10, Lib Dem: 7/10, Labour: 5/10.

Day 27 (Sunday): Sunday papers line up with the daily papers, endorsing Tories and Lib Dems. But a couple of polls show a Lib Dem waver. Cameron managing the momentum by pre-announcing what he'll do in his first days of government - a clever strategy both as it is believable, and offers voters a feeling of certainty about the future in a time of great uncertainty. Labour may benefit from Lib Dem drop-off. Tory: 8/10, Lib Dem: 6/10, Labour: 6/10.

Day 28 (Monday): Clegg gets some useful coverage criticising Cameron for "measuring the curtains in Number 10" - the downside of the momentum strategy. The Philippa Stroud story (Tory candidate used to run prayer sessions which some claim were intended to "exorcise" gay people) sneaks onto the BBC's website but is not noticed by most voters. However, undoubtedly the story of the day is Gordon Brown's amazing speech to Citizens UK - which leaves most people wondering where he's been for the last two years. Labour: 9/10, Lib Dem: 7/10, Tory: 6/10.

Day 29 (Tuesday): The polls have been steady for the last few days and the result will clearly come down to two things: individual voting behaviour in each constituency, and the unpredictable last-minute choices of undecided voters. Today, therefore, tactical voting is the topic. Some Labour ministers endorse tactical voting to keep the Tories from winning. Officially, Gordon Brown does not, but he has to say that. However the mixed message dilutes the potential last-gasp recovery message from Brown's Monday speech. Cameron gets some good coverage for the start of his 36-hour campaigning marathon. Liberal Democrats should benefit from tactical voting call, Tories may be hurt a little. It's not clear though whether tactical voting will be more deployed against Labour or the Conservatives, and whom Lib Dem votes will hurt in Lab-Con marginals. Lib Dem: 8/10, Tory: 7/10, Labour: 6/10.

Day 30 (Wednesday - try finding the 5 May link on the BBC website by the way. Thanks, Google!): All the final polls projections are out, and nearly universally point to a hung parliament with the Conservatives the largest party. My own projections from CountMyVote are Conservative 285, Labour 198, Liberal Democrats 132. If this happens, things will be very precarious - even a Lab-Lib coalition would barely have a majority. Most other projections show the Tories closer to a majority, Labour also with more seats, and the Lib Dems with fewer. Five Thirty-Eight, an excellent site which predicted the 2008 US election results very accurately, is interesting on this. Will these projections have a psychological effect on voters? They may motivate Lab-Lib tactical voters a bit, but against that they might also encourage some uncertain Tory voters to vote Conservative to try to ensure a clear majority. The best bet for Labour is the hope that people, their dislike of Brown expunged by an endless campaign, will revert to the incumbent government in the polling booth. It's too hard to predict the psychological impact of these numbers, so I'm going to call today's ratings: Conservative: 7/10, Labour: 7/10, Liberal Democrat: 7/10.

Time to go and vote now. After the fog has cleared I'm sure there will be plenty to add when hindsight shows up. And maybe I'll get back to a bit of that "economics" stuff. The bond markets are opening at 1am on Friday. I can't wait.

Monday, 3 May 2010

Behavioural politics, day 24 of 30

The leaders' debate tonight showed the immense power of the phenomenon of confirmation bias.

The debate was interesting but neutral in terms of campaign impact - in my judgement, the candidates performed about as well as each other. And yet watching it on twitter was very informative yet again, primarily because of the reactions when the post-debate polls were released.

Nearly all the polls showed Cameron as the "winner", with Clegg and Brown in second and third place or vice versa. A typical poll (see writeup here) gave:
Cameron 35; Clegg 33; Brown 26
But when you look at the voting intentions of the people who answered, what do you get?
Conservative 35; Lib Dem 36; Labour 24
Spot anything? The numbers are almost identical.

When there are no clear distinctions between the quality of the performances, people who already agree with David Cameron will think he's talking sense; Labour supporters will think Brown is the only one who isn't lying; and those who support Nick will agree with Nick.

So it's no wonder that Cameron, who's in the lead in most of the regular polls, will come out as having won the debates. This is more true the further we go into the campaign.

But there are still some undecided voters, and the Angus Reid poll splits them out. About 37% went for Clegg, 25% for Cameron and 22% for Brown. Even this doesn't tell us that much, because undecided voters are likely to have a specific profile, and I'm not surprised that they would be pro-third-party.

A revealing detail is that while Tory partisans clearly chose Cameron as top performer, Labour and Lib Dem supporters were more likely to "swap", picking Clegg or Brown respectively as the winner. This is likely to indicate that the parties are drawing from a common base of shared support. Tory and Lib Dem supporters have a smaller swap tendency, showing that there are some swing voters between those two parties; while only 4.5% of Labour supporters thought Cameron best, and only 3.5% of Tories rated Brown.

Labour and the Lib Dems were also more likely to think nobody won; I wonder if that might indicate weaker support of their party, or a lower likelihood of voting.

Given all this, it's surprising to hear today that a third of voters are undecided (I think what Doug Alexander means by this is that a third have said they might still change their minds). This debate hasn't changed much in itself, but the timing of it has managed three important things:

  • moved the news cycle on from yesterday's "bigot" news (the intensity of yesterday's coverage, and Brown's instant apology, also made it easier to move on today). This is a critical save for Labour, as they came close to annihilation yesterday, so they get the best ratings today.
  • sustained the Lib Dems' momentum for yet one more day, making it just that bit more believable.
  • helped the Tories firm up their own momentum, which will become more important over the next week.
Ratings: Labour 7/10, Conservative 6/10, Liberal Democrats 6/10.

Sunday, 2 May 2010

Behavioural politics, day 23 of 30

Oh, Gordon.

Today, in possibly the most embarrassing thing that could possibly have happened to a modern politician, Gordon Brown was caught on tape describing a voter as "a bigoted woman".

Most of you will know the details all too well by now, but those who don't can read them here and should watch Jon Stewart's take on it - if you can find a video clip that works in your country.

Although this was widely considered to have written off Labour's chances in the election, the focus on it has been so intense that Brown has not been harmed as much as expected. Attention is the currency of an election campaign, just as it is in the commercial world. And Gordon Brown has had more attention than he could possibly have bought.

The short-term polls bear out this surprising result - Labour has had no drop-off in support since this happened.

On the other hand, it creates such a strong narrative that it allows little room for any other story. That gives Labour in particular - but also the other parties - no opportunity to present a new message and change the pattern of their support. This event just further entrenches voting intentions that have been gradually solidifying for the last two weeks.

Really the other two parties have had no look-in today, so the ratings reflect Brown's squandering of Labour's chances of catching up but also the fact that they haven't been damaged fatally. At least the party will no longer have any reason to feel guilty about getting rid of him after the election. In the short term, the Liberal Democrats have survived another day as a front-runner, and that is all they really need to do right now.

Ratings: Liberal Democrat 7/10, Conservative 6/10, Labour 3/10.

The economics zeitgeist, 2 May 2010


This week's word cloud from the economics blogs. I generate a new cloud every Sunday, so please subscribe using the RSS or email box on the right and you'll get a message every week with the new cloud.


I summarise around four hundred blogs through their RSS feeds. Thanks in particular to the Palgrave Econolog who have an excellent database of economics blogs; I have also added a number of blogs that are not on their list. Contact me if you'd like to make sure yours is included too.

I use Wordle to generate the image, the ROME RSS reader to download the RSS feeds, and Java software from Inon to process the data.

You can also see the Java version in the Wordle gallery.

If anyone would like a copy of the underlying data used to generate these clouds, or if you would like to see a version with consistent colour and typeface to make week-to-week comparison easier, please get in touch.