Stumbling and Mumbling

Where's the production function?

Should small schools be closed? Is the government right to build giant prisons? These two questions share a common theme - the theme being that the government's pretensions to technocratic expertise are a sham.
In principle, the question of what size schools or prisons should be is a narrow technical - mere mathematical -  one. The question is simply: what is the production function?
Take prisons. A desirable output might be a low recidivism rate. Our independent variables might include prison size, guard-prisoner ratio, money spent on therapies and education, diet, controls for the type of prisoner and so on. We then regress the recidivism rate upon these variables across all prisons, and estimate the coefficients. We can then see the impact prison size has upon recidivism.
Prison size, then, is not a political question, but a mathematical one.
So, why are questions of school or prison size political ones, rather than mere maths? Here are some possibilities:
1. There are multiple desirable products of schools or prisons. For example, in schooling the desire to maintain community cohesion might argue for small schools, whilst exam grades might argue for larger ones. The production function approach still has a role here, however, as it allows us to better estimate the trade-offs between these objectives. It allows us to say (for example): "keeping a village school open costs one GCSE per student."
2. The notion of a production function might be incoherent. There might be more than one optimum technique. For example, big prisons might work well with big spending on education, but worse than small prisons if there's little such spending. 
3. There might be insufficient variation in inputs to allow us to estimate coefficients well - for example, if spending per pupil varies little from school to school.
4. Politicians - and unions - might not like what production functions say. For example, there's evidence (pdf) that spending on schools on its own doesn't (pdf) do much to improve outcomes.
5. Drawing attention to production functions forces politicians to look at outputs, not inputs. Their boasts of spending billions more on education (or health) become vacuous. And they have to do the grunt work of actually making things work.
Whatever the reason, the fact's the same. In not thinking about the maths of production functions, politicians  - unless they have Sraffian doubts about their efficiacy, which I doubt - are like cargo cults. They practice a sham ritualistic imitiation of rational management, not the substance of it.

What do mortgage approvals mean?

The Bank of England reports that mortgage approvals fell to a 12-year low, of 73,000, in December.
This chart shows one implication of this - house price inflation will fall further. Mortgage approvals have been a good lead indicator of house price inflation in the following 12 months; the correlation has been 0.61 since January 1992, with an R-squared of 37.8%. This relationship points to prices rising just over 1% during this year, with a highish probability of a fall.
The question is: so what? As Willem Buiter argued, house prices are not, in aggregate net wealth.
But there's another point that should add to our indifference. It's that the macroeconomic picture tells us nothing about individual experience.
By rights, I should be more conscious than most about the housing market, as I'm hoping to sell my flat  very soon.
But the aggregate housing market doesn't interest me. What matters is the market for flats in Belsize Park (top location, bags of potential, low outgoings, extensive views over Helena Bonham Carter) and for houses in Oakham. Macro news tells me little about either.
I'll say it again. The main function of macroeconomic data is to make work for macroeconomists. And this is a very valuable function; how d'you think I got that flat?

Conway and class war

If you're looking for someone who defrauds the taxpayer, you are four times more likely to be successful if you look amongst MPs than amongst benefit claimants. And you're 13 times more likely to find a fraudster on the Tory benches than you are among benefit claimants.
The National Audit Office reported (pdf) recently that there 6964 people were successfully prosecuted or cautioned for benefit fraud in 2006-07. That's one in 2585 claimants.
However, thanks to Conway, there are one in 646 MPs who are fraudsters and one in 198 Tory MPs who are.
Given that there are more fraudsters per person among MPs than among benefit claimants, the questions arise. Why do the main parties make so much fuss about benefit fraud? And why are the penalties for it greater than the penalties MPs suffer for fraud? For cheating the taxpayer of £77,000 Conway is merely suspended from the Commons for 10 days; for most people, being free of the company for MPs for a fortnight would be a blessed relief. By contrast, benefit fraud of such magnitude is highly likely to land you in chokey.
So, why the difference? There are four possibilities:
1. I'm just exploiting sampling error. If you take a small sample (Tory MPs say), the proportions you find of any behaviour - good or bad - are likely to be biased.
2. The sums involved in benefit fraud are larger. True, they are large relative to MPs outright fraud. But they are small relative to MPs incompetence. The DWP  estimates that in the 12 months to September 2006, fraud in Income Support and Jobseekers Allowance cost just £260m (figure 2C of this pdf). That's 26p  for every £10 we'll spend on the Olympics. And it's less than a penny for every £20 of overall government spending. MPs seem to expend a lot of wind on a small problem.
3. Freudian Projection. In smearing benefit claimants, perhaps MPs are doing what queer-bashers notoriously do; they project their own "sins" onto others. Knowing subconsciously that they themselves are frauds - if only in the non-criminal sense of falsely claiming to be able to control society for the better -  they imagine that others are also fraudsters.
4. Class hatred.  Politicians' hatred of benefit claimants is based not upon the damage they do but upon the fact that claimants are low class and unskilled. Conway, by contrast, despite his modest background, is "one of us."

"Contacts, contacts, contacts"

Commenting on Blair's latest job - advising Zurich Insurance on climate change - Bryan Appleyard says:

I know more about climate change than Tony Blair; if I had a dog and had taken it on my last trip to see Jim Lovelock it would know more about climate change than Tony Blair.

He's right, but misses the point. What Blair's selling is not expertise but contacts - not human capital but network capital.
And there's tons of evidence that network capital matters. Knowing other researchers allows scientists to collaborate on generating ideas. Social capital is often crucial in helping scientists turn inventions into companies. Non-executive directors are often useful not so much for their expertise as their contacts with suppliers or customers. And it's a knowledge of other people that helps entrepreneurs make money; any good salesman will tell you that it's at least as important to know your customer as your product.
It's for reasons like this that the internet and mobile phones can boost economic growth, particularly in Africa, and this may be why the privately educated earn more than the state educated, even controlling for qualifications.
Indeed, it could be that network capital is actually more important in raising incomes than human capital. I mean four things:
1. What looks like a return to experience is in fact a return  to knowing more people. If you've done a job for a few years, you'll have met likely employers and your reputation will have grown. Beyond a certain point, it may be these that raise your earnings more than the technical skills you've acquired from an extra year of work. This, surely, accounts for Blair's high earnings now.
2.  Returns to human capital might be over-estimated, if a university education serves to signal ability rather than merely develop it.
3. This lovely paper (pdf) by Marcel Fafchamps estimates that, among agricultural traders, network capital gives a higher return than human capital.
4. Macroeconomic evidence for a strong link between human capital and subsequent growth is surprisingly weak; see, for example, the critiques here (pdf) here (pdf) and here (pdf). 
Which brings us to a nice irony. Blair famously said the key to economic success, for individuals and economies, was "education, education, education." But his subsequent career demonstrates that what might be more important are "contacts, contacts, contacts."

Cheryl & Cashley: the economics

Should Cheryl divorce Cashley? The economics of the decision are more balanced than you might think.
The case for her doing so, other than that he's a lying cheating no-good shortarsed drunk with a dick the size of a bookie's biro,  is two-fold:
1. She has good outside options. Research shows (pdf) that the probability  of divorce rises, the higher is the woman's  share of family income. What's more - and in this she differs from Ms Beckham - she has a good probability of getting herself a better bloke than Cashley. (Not me though - her bandmate Sarah has first refusal.)
2. Maslow's hierarchy of needs (and indeed diminishing returns) predicts that, as income rises, other personal goals become more important relative to income. Among these is self-esteem, and this points in only one direction.
Against this, two points argue for her sticking with Cashley:
1. Income hedging. Although Girls Aloud are one of the greatest pop bands in British history, public tastes are fickle, so Cheryl faces income risk. As hers and Cashley's incomes are probably uncorrelated, Cashley is a good hedge.      
2. Joint incomes matter. The same research that showed that divorce rises with the wife's relative income also shows that divorce is less likely, the richer is the couple. This could be because spouses get utility from each other's income, independently of their own. Or it could be because a high joint income complicates the finances of the divorce decision. This effect, however, is weaker than the outside options effect.
The decision, then, is fimely balanced. But the 10-1 you could have got on them splitting at the start of the year now looks a good price.

Is the FA Cup draw fixed?

Another Cup draw, another easy match for Chelsea. There was only a one-in-15 chance that they would have been drawn at home to a League One side today. And it came up.
Of course, rare events do occur - that's the inevitable nature of the draw. But it's amazing how often they come up to benefit the money launderers. Last year, I calculated that there was only a one in 500 chance of Chelsea's easy matches in the first three rounds being drawn by true randomness.
Barring Saturday's away match at Wigan, this year has continued the sequence of easy draws in the early rounds; they were home to QPR in the third round.
The question raised by this isn't just the footballing one. It's about the nature of statistical inference.  The fact that an event is prodigiously unlikely to have happened by chance is not proof that it didn't so arise. Which raises the question: what, if anything, can probabilistic statistics prove?

Big government & growth

Big government is bad for economic growth. That's the finding of this new paper (pdf) from economists at the ECB.
The effect is large. Controlling for a few obvious things, they estimate that, within OECD countries a one percentage point  rise in the share of government spending in GDP cuts growth by 0.13 percentage points a year. This implies that the rise in government spending we've had in the UK since 2000 (from 37.2% of GDP to 42%) would, if sustained take half a point off GDP growth, making us more than 5% worse off in 10 years' time than we would have been had spending stayed at 2000's levels.
The cross country growth equations that this finding is based upon are, of course, subject to many problems. Not least is omitted variables bias. Could it be that there's something that's correlated with big government which itself depresses subsequent growth?
However, the paper doesn't altogether uphold the orthodox neoliberal case for smaller government. It estimates that it is indirect taxes, and not income taxes, that depress growth most.
This implies that a a cut in excise duties and VAT and rise in income tax could actually increase growth. As indirect taxes are regressive, it would also increase equality.

Livingstone & selection bias

Why is it that prominent left-wing politicians are such arsewipes? I don't just mean Galloway and Livingstone, but Derek Hatton before them, among others.
It's a paradox because ordinary left-wingers are (mostly) wholly decent people.
The answer, I reckon, lies in selection biases. It's in the nature of far left politics that only the worst specimens become public figures.
There are (at least) two reasons for this.
One is that if you're on the far left, the standard way for a decent hard-working but dull person to have a career in politics - think tank, special advisor, MP etc - is closed off. Instead, you make your name first in local politics. And this requires sharp elbows, "charisma", coalitions with shady characters and raw ambition; victory often goes just to whoever attends the most meetings.
The other is that there's little demand in the media for nuanced, lengthy argument. Even in the main parties, intellect is a drawback; just look how David Willetts or Ruth Kelly get treated. There's a premium instead upon soundbites and publicity gimmicks. This premium is larger for the far left, which is easily ignored when it behaves intelligently. So it's only demagogues who get attention.
These biases mean that, on the far left, it is the worst that rises to prominence. And in turn, good leftist ideas are apt to be discredited by association.
But this just shows how the status quo defends itself. Ordinary cognitive errors - selection bias and the association fallacy - combine to diminish support for alternatives. It's surprising how many sources of pro-capitalist ideology there are.

A liberal-left manifesto

Sunny's asking for a liberal-left manifesto for 2008. We need one, if only for one purpose - to avoid being discredited by association with this government.
In principle, of course, there's no should be no such link. The fact that it wants to bang people up without charge for 42 days shows the government isn't liberal. And the fact it prefers to cut inheritance tax than child poverty shows it isn't left.
A liberal-left manifesto for 2008 should make clear the yawning gap between liberal-left principles and what the government does. This doesn't mean policy-wonking (yet), but rather influencing the intellectual climate. I mean five things:
1. Reclaim the concept of freedom. There's more to freedom than low taxes and a formal, legalistic absence of state intevention, important as these can be. Real freedom means the ability to control one's own life. And this sometimes requires state intervention, as Gracchi points out.
2. Argue intelligently for equality. One of the most damning indictments of New Labour is its failure to do just this. We can and should do better. We can point out that greater economic equality might actually be better for the economy than low taxes on the rich, and that there's a moral case for equality, partly as a form of pooling risk.
3. Exploit the economic slowdown sensibly. This doesn't mean calling for protectionism, immigration controls or old-style Keynesianism. It means pointing out that the big lesson of the sub-prime crisis is that the vulgar free market cheerleaders were wrong. Unfettered markets don't pool risks anything like as well as theory predicts they should. Financial innovation has taken a wrong turning. It's been a way for egomaniacs to gamble, not a way for real people to insure themselves against economic crises. There's perhaps a case for state intervention to encourage the development of insurance markets against recession or industrial or occupational shocks, as Robert Shiller has shown.
4. Challenge authority. The really big fraud uncovered this week at SocGen wasn't Jerome Kerviel's trading. It's the pretence of every boss everywhere that they are in control of their organization. They're not. Managerial effectiveness is a fiction. What looks like good management is either an illusion or the goodwill and competence of workers.
5. Lose faith in big government. A lot of the right's objections to the welfare state are based not so much upon hostility to redistribution as upon the belief that the state is too big, unwieldy and incompetent. They're right. The liberal-left should think how state services can be provided with less red tape. This doesn't mean blind privatization, not least because this can crowd out the altruistic motiviations of workers that keep schools and hospitals going. It just means thinking about organizational design.
Points 3, 4 and 5 are of course connected. The job of the liberal left should be to argue that markets are often in some form better than state provision. And the job of the liberal left should be to argue that top-down management is as wrong in the private sector as public.

Exploitation and freedom

Russell Roberts argues that Wal-Mart can’t exploit workers because so many people apply for jobs there. Is this valid, or is it the sort of error that Norm deplores?
To see the problem, imagine a sort of Gulag Archipelago, in which some prison labour camps are more oppressive than others. Prisoners are free to apply to enter particular camps, but not free to leave the archipelago. There will then be thousands of applicants to enter the less onerous camp, and successful applicants will be happy to get there.
But it doesn’t follow that these are not exploited. And their freedom is horribly limited. It‘s a choice amongst oppressive exploiters; the option of leaving the system altogether doesn’t exist, at least for the individual*.
To Marxists, capitalism is like this archipelago. It offers choices of degrees of oppression and therefore some freedom. But the choice that matters most doesn’t exist. In capitalism, workers lack an outside option. 
The question is, are Marxists right? Defenders of capitalism would reject my analogy in two ways:
1. They would define coercion (and hence unfreedom) as only the conscious actions of state agents. Men in the archipelago are coerced. But men in capitalism are not.
2. Under capitalism, the outside option is a natural state of rural idiocy - subsistence farming and poverty.
Marxists reply that point 2 is wrong; outside options should be richer than a state of nature, with various forms of socialism among them.
And, they say, point  1 is a narrow etiolated conception of freedom that drains the notion of value. What’s more, people have often been deprived of an outside choice because state-sanctioned primitive accumulation has robbed them of their land**. As the man said:

Nowhere and at no time has the large scale ownership of land come into being through the workings of economic forces in the market. It is the result of military and political effort. Founded by violence, it has been upheld by violence and that alone.

(That’s not Marx, its von Mises.)
As you know, I side with the Marxists on these points. But that’s not my main point. The point is merely that long queues for jobs at Wal-Mart, or at any factory in the third world, is no proof at all that workers are free and unexploited.
* It’s possible that the prisoners could collectively revolt and overthrow the Gulag system. Does the fact that they don’t do so show tacit consent? This raises issues for another day.
** If you think it relevant that the primitive accumulation described in these links happened under a "Communist" government, you misunderstand Marxism as badly as the Chinese government does.