Stumbling and Mumbling

Forecasts vs mechanisms in economics

This discussion between Edmund Conway and Andrew Lilico on the Today programme on the alleged crisis in economics seems to me to rest upon a misunderstanding of what economics is.
Conway says the crisis has been “an earthquake for economic thought” and Lilico says we need “new theories.“  This, though, seems to regard economics as a settled but inadequate body of knowledge and theory. It’s not. It is instead a vast number of diverse insights. What’s more, all of the insights that help explain the current economic crisis were, in truth, well known to economists before 2007, for example:
1. Risk cannot be simply described by a bell curve. But we learnt about tail risk on October 19 1987. And we learnt from the collapse of LTCM in 1998 that correlation risk, liquidity risk and counterparty risk are all significant.
2. Assets can be mispriced. But we’ve known about bubbles for centuries - since at least 1637. Their existence does not disprove the efficient market hypothesis; as I’ve said, the EMH is not the rational investor hypothesis. Nor, contrary to Conway’s implicit claim, is the EMH inconsistent with the possibility that behaviour can be swayed by emotions; the EMH allows for the possibility of time-varying risk premia*.
3. Long periods of economic stability can lead to greater risk-taking. We’ve known this since (at least) Hyman Minsky.
4. Banks can suffer catastrophic losses - which are correlated across banks. We learnt this - not for the first time - in the Latin American debt crisis of the early 80s and in the crises in Japan and the Nordic countries in the early 90s. Banking crises are a regular feature of even developed economies.
5. Institutions, such as banks, can be undermined by badly designed incentives.  But there’s a huge literature on the principal-agent problem.
The current crisis, then, has not thrown up much that economists didn’t know.
Instead, our problem is a different one. It’s that what we have are lots of mechanisms, capable of explaining why things happen and the links between them. What we don’t have are laws which generate predictions. In his book, Nuts and Bolts for the Social Sciences, Jon Elster stressed this distinction. The social sciences, he said:

Can isolate tendencies, propensities and mechanisms and show that they have implications for behaviour that are often surprising and counter-intuitive. What they are more able to do is to state necessary and sufficient conditions under which the various mechanisms are switched on.

This is precisely the problem economists had in 2007. We knew that there were mechanisms capable of generating disaster. What we didn’t know is whether these were switched on. The upshot is that, although we didn’t predict the crisis, we can more or less explain it after the fact. As Elster wrote:

Sometimes we can explain without being able to predict, and sometimes predict without being able to explain. True, in many cases one and the same theory will enable us to do both, but I believe that in the social sciences this is the exception rather than rule.

The interesting question is: will it remain the exception? My hunch is that it will; economists will never be able to produce laws which yield systemically successful forecasts. 
What’s more, I am utterly untroubled by this. The desire for such laws is as barmy as the medieval search for the philosopher’s stone. If you need to foresee the future, you are doing something badly wrong.
* The basic insight of efficient market theory is that you cannot out-perform the market except by taking extra risk. I am sick and tired of hearing people who still have to work for a living trying to deny this.

Social mobility in the long run

In this new paper (pdf), Gregory Clark of A Farewell to Alms fame makes a gobsmacking claim:

Pre–modern England, all the way from 1250 to at least 1860, was a society without persistent social classes. It was a world of complete social mobility, with no permanent over-class and under-class. It was, despite all appearances, a world of complete equal opportunity…Instead of moving from a world of immobility and class rigidity to a world of complete mobility we have moved in the opposite direction.

His evidence for this comes from a studying the distribution of surnames. In the 13th and 14th centuries, the richest people were overwhelming those whose names derived from places, whilst the poor tended to have names derived from crafts: Smith, Wright and suchlike. However, by the 16th century, this link between wealth and surnames had vanished; the rich were as likely to have craft surnames as the general population.
This, says Clark, suggests that there was complete social mobility, if we look at a long enough time-span.
How can we reconcile this with the fact - reported in a new paper (pdf) from the OECD - that the UK now has low social mobility?
One possibility is that there’s no inconsistency, just different time periods. Let’s say that the coefficient on father’s earnings of a son’s earnings is 0.5; this is consistent with Miles Corak’s estimate. In other words, if a man’s income is 100% more than the average, his son’s income will be 50% more. Then the coefficient over five generations will be 0.5 to the power five. Which means that the man’s great great great grandson will get only 3% more than the average.  Over five generations, then, inherited advantage will, for all practical purposes, vanish.
There is, though, a more sinister possibility - and here I‘m speculating wildly. Today, a major determinant of someone’s economic success is their education. This, though, is highly heritable not only because rich parents get their kids into the best schools, but also because cognitive skills are partly inherited. However, in the past there were many other routes to wealth: predation, winning a lord’s favour, skill on the battlefield and so on. If these were less heritable than education, mobility might have been greater.
A further question is: why should we care if there is social mobility over four or five generations? Those who worry about a lack of social mobility do so because individuals’ life chances are blighted by the misfortune of who their parents happen to be. Is it really a consolation to these individuals that their as yet unborn great-grandchildren won’t suffer such disadvantage? Surely, we worry about injustice because of its effect upon existing individuals. If so, the claim that there is social mobility over the long-run is irrelevant.

Invariant inequality

The Office for National Statistics today published its latest figures (pdf) on income inequality. These show that, in 2007-08, the richest 10% received 28% of all post-tax income - that is, incomes including benefits after direct and indirect taxes - whilst the poorest 10% got just 2%. That means the best-off 10% have 14 times the income of the worst-off 10%.
What the figures also show is that this inequality is barely changed at all in recent years.
Indeed, what is striking is just how insensitive inequality is to either policy or to the economic “cycle.” Looking at this chart, you would struggle to identify either a change of government or any macroeconomic boom or recession; the exception here is that the 1990-92 recession increased the gap between the two Gini coefficients slightly, because increased numbers of people on benefit meant that the state did more redistribution.
One reason - of many - why inequality hasn’t changed much is that the tax system (as distinct from benefits) does nothing to reduce it; direct taxes have a small equalizing effect, reducing the Gini coefficient by four points, but this is exactly offset by the regressive effect of indirect taxes.
The only message I’d take from this is that if you want to reduce inequality, you need something much more radical than has been tried in recent years.  I suspect that one reason why James Purnell claims to be interested in equality of capabilities is that he’s realized that it’s just too damned hard to achieve much greater equality of income this side of revolution.

Cognitive biases & state intervention

Leigh Caldwell says that, thanks in part to people’s cognitive biases, markets can fail and so “Government regulation …may be a way of improving how the market works rather than replacing it.” His use of the word “may” is well judged. The fact that decision-making is subject to cognitive biases is not, in itself, an argument for government intervention, because governments can be as irrational as market participants. The notion that the state is rational is as fanciful as the notion that it is benevolent.
Take just five examples.
1. The “fighting the last war” error. Alistair Darling wants banks to hold higher capital assets ratios. This would have been the solution to excess lending. But it deepens our current problem, which is insufficient lending.
2. The availability heuristic. Governments can overweight the significance of rare but salient events, and so over-react. This is what led to the dangerous dogs act.
3. Misperceiving randomness. A related error is to fail to see that freak events are just freaks, and do not require a response. And example of this error was when a convicted rapist won the lottery and then-Home Secretary David Blunkett promised to find ways to ensure this didn’t happen again - oblivious to the fact that the law of probability was a sufficient safeguard.
4. Promoting over-confidence. Leigh gives a good example of this. If the government ensures that customers are better-informed, it might merely encourage them to take risks. I suspect this helped contribute to the mis-selling of financial products such as split cap investment trusts; some consumers, believing the products were regulated, over-estimated their security.
5. Ignorance of less obvious mechanisms. For example, the US’s tacit support for Iraq in its war against Iran in the 1980s helped strengthen and embolden Iraq to invade Kuwait. A less dramatic example is that higher alcohol taxes might promote binge-drinking, not reduce it.
I could go on. My point is merely that there is no guarantee that government rationality can act as a bulwark against private sector irrationality. Indeed, it may be that pressure from the mass media mitigates against the government acting rationally. The choice is not: irrational markets vs. rational government? It’s irrational markets vs. irrational government.

Relative poverty for the hard of thinking

Not all Tories are as thick as farmyard animals. But if I were in the habit of drawing inferences from anecdotal evidence, I'd probably think they were. Take Dizzy‘s complaint about the Child Poverty Bill, which aims to outlaw child poverty:

Can someone tell me how you can outlaw a moving target of a median? You can't. There will always be 50% above the median, and 50% below it, because that's what median means…. You can't eradicate relative poverty because by definition it's relative.

This is the purest arse-dribble. For the hard of thinking, take society A in my table., which comprises seven members. Median income is 50, so the poverty line - 60% of the median - is 30. Three people are then in relative poverty.
The government then levies a tax on incomes over 80. This raise 10 from the richest man and 8 from the next richest. This money is then used to lift the three people above poverty. In society B, no-one now lives in poverty.
This shows that it is conceptually trivial to eradicate relative poverty through the tax and benefit system alone.
The difficulties aren’t conceptual, but practical. Such a tax might cause the rich to emigrate, so the government would have to tax middler earners. Or maybe such a tax and benefit system would dampen work incentives and so reduce average incomes all round.
These points, though, are empirical ones requiring empirical support. Dizzy’s claim that “You can't eradicate relative poverty because by definition it's relative” is simply 100% false. Indeed, the opposite is true. Relative poverty is easy to eradicate precisely because it’s relative.
Which brings me to my question. A moment’s thought would tell anyone this. So why didn’t Dizzy put in that moment’s thought? Is it because, like some other Tories, he just doesn’t want to think, even minimally, about poverty? Or is it because he has such an inflated sense of his own entitlements that he thinks his opinion is worth reading even if he puts no thought into it? Or is it because, in political debate, grunts of prejudice are as good as thinking anyway?

Grammar schools, mobility & cognitive bias

Why do so many people think grammar schools are a way of improving social mobility? It’s not because the social research says they are. I suspect instead that plain error is involved.
One such error is the availability heuristic, mixed with survivorship bias. The handful of working class people for whom grammar schools were a means of upward mobility have high profiles, make lots of noise and get lots of attention. It’s easy, therefore, to over-estimate their numbers. This is especially true if you were one of these people yourself. By contrast, the many who failed the 11-plus and lost opportunities to “move up” get ignored. Anecdotal evidence can be systematically misleading.
This error might be magnified by a tendency to exaggerate the poverty of the backgrounds of people who did get through grammar school; doing this is a corollary of the middle England error, the tendency to exaggerate median incomes. Matthew D’Ancona (educ: St Dunstan’s, annual fee = £12,660, £2000 more than a full-time minimum wage job) probably thinks that anyone who doesn’t have punctuation in their surname is working class. The truth is, though, that the typical grammar school product is more like Oliver Kamm than me, to take two Wyggestonians.
There is, I suspect, a third motive: wishful thinking. The “right” want to believe that perceived injustices are the result of state actions, rather than impersonal societal forces. This leads them to over-play the role of policy in depressing social mobility. This is especially easily done if they can convince themselves, against the facts, that the policy was that of their opponents. 
Now, I don’t have a dog in this race: I don’t give a damn how much social mobility there is. I say all this not to take a stand on education and social mobility, but rather to point out just how political opinions are shaped by inferential error rather than hard evidence. This is, of course, much easier to spot in other people’s opinions.
Another thing: Let’s be clear. You cannot support both more grammar schools and more choice in education. If schools select pupils, pupils cannot - except by happy coincidence - select their schools.

Markets in public services

The latest row between Neal Lawson and Tim Worstall suggests to me that Bryan and Danny might be right - abstract ideas can be distracting delusions.
Neal writes:

Markets, the mechanism of choice, are designed to allocate the spoils to the winner. That's OK with supermarkets and car makers – it's not OK for the state. We can't have and can't allow schools and hospitals to fail and be replaced by the fittest.

Tim replies:

This is in fact what we actually want, a continual ratcheting up of quality in pursuit of the discerning consumer so that the spoils do indeed go to those who are providing the goods or services better than those others around them.

So who’s right? To a large extent, it’s a matter of empirics, not abstract theorizing about the virtues or not of markets.
To see what I mean, imagine that parents are allowed to choose freely, through some market-type mechanism, between a good school and a bad one.
What happens? A lot depends upon why the good school is good - and in particular upon whether its quality is scalable.
Let’s imagine school quality is not scalable - that the school is good because it has some good teachers, who only shine because they have smallish classes to work with.
In this event, there are two possibilities.
One is that as the school expands in response to parental choice, its quality worsens. If there are informational frictions - say if parents choose schools on the basis of out-of-date information - then the “good” school could (for a while at least) become worse than the “bad” school.
Alternatively, the school might recognize that it is not scalable and simply refuse to expand. There’d then be excess demand for its places. This poses the danger that places will be allocated according to class - say because house prices rise near (pdf) the school’s catchment area, or because posher parents are better able to wheedle their kids into the school.
In this world, Neal is right. Market allocation is a bad idea.
But what if school quality is scalable? Then - assuming parents have good information about quality - Tim would be right. Good schools could expand at the expense of bad ones, thus raising average quality. And because there’d be no need for the school to ration places, if it could expand without loss of quality, it would have no need to allocate places preferentially to posher parents. So there’d be no increase in inequality of opportunity.
The issue, then, is not so much one of principle as empirics: what is the educational production function? To what extent is a good school scalable?
The thing is, AFAIK (I’d welcome corrections) there’s not much hard evidence here; certainly, neither Tim nor Neal provide any.
Yet again, it seems, the strength of people’s opinion is disproportionate to the evidence base.

Grammar schools & social mobility

The Milburn report has caused the usual suspects to complain that the abolition of grammar schools is a big reason for lower social mobility. The evidence for this is, however, almost entirely anecdotal. It has almost no basis in hard research.
Granted, some researchers have found evidence that grammar schools improve academic performance. This paper (pdf) by Brendon McConnell, Adele Atkinson and Paul Gregg looked at those LEAs which still have selection and concluded:

Children educated in grammar schools do substantially…better than their peers [with similar results at age 11] in similar non-selective areas. This is equivalent to raising 3½ GCSEs from a grade ‘C’ to a ‘B’.

And Fernando Galindo-Rueda and Anna Vignoles studied (pdf) pupils performance before and after comprehensivization and found that: “the most able (top 20%), particularly the most able women, did do better in a selective school system.”
Not all researchers, though, have corroborated these findings. Eric Hanushek and Ludger Woessman studied the effect of early selective education around the world, and found that it tended to reduce average performance. And Alan Manning and Jorn-Steffen Pischke have estimated (pdf) that the pattern of students test scores at 16 is similar to that at 11, under comprehensives or selective education systems. This is consistent with the possibility that, on average, comprehensivization made no difference to educational outcomes. They say: “we probably do not know very much about the effect of comprehensive schooling in Britain, or elsewhere for that matter.”
Other researchers have found, from evidence around the world, that selective education can actually reduce social mobility. Giorgio Brunello and Daniele Checchi estimate that the correlation between parental background and individual earnings is increased by selective education. Even this finding, though, isn’t universally shared. Fabian Waldinger has estimated (pdf) that selection doesn’t causally affect the relationship between family background and educational outcomes - though this view is scant consolation for those who believe that grammar schools improve social mobility.
Indeed, even those who have found that grammar schools do improve education out-turns estimate that they do very little for social mobility. This is simply because few poor children get into them, even if they have high ability. Atkinson, Gregg and McConnell estimate that 11-year-olds of high ability who are so poor that they are eligible for free school meals are only half as likely to get into grammar schools as those who aren’t eligible for them.
Reading this research, I’m struck by the contrast between the vehemence of grammar school supporters and the ambiguity and complexity of the research.
But of course, everyone thinks that the fact they went to school means they are experts on education. Ego beats evidence every time.

Milburn on social mobility

There’s something that puzzles me about Milburn’s report on social mobility. A major claim in it is this:

There is likely to be a need for many more professionals in future. Some studies suggest that up to nine new jobs in ten created over the next decade will be in professional and managerial sectors. Once retirements are taken into account, around seven million new professionals may be needed (p21)

I’m sceptical of this. Will we really have more civil servants or journalists or bankers in future? Do we really need even more accountants or managers or lawyers?
But let’s assume this claim is true. If this is the case, then surely the professions will open themselves up to people from poorer backgrounds by necessity, without the need for any government intervention.
Put it this way. Milburn says:

Today’s younger professionals (born in 1970) typically grew up in a family with an income 27% above that of the average family (p19).

But let’s assume his futurology proves too pessimistic, and all future jobs will be professional. Then, by definition, the typical professional will have grown up in a family with average income. Simple maths will have solved the “problem” of social mobility on Milburn’s definition, without the need for intervention.
Of course, the futurology could well be wrong, and the professions will remain minority occupations. But if this is the case, then aren’t Milburn’s recommendations too weak? If there remains only little “room at the top” then social mobility requires - by definition - that people from rich families be downwardly mobile. And yet Milburn gives us no ways of achieving this - which is not surprising as rich families will naturally fight hard to give their children a good start.
My priors remain - that social mobility is a foul ideal, and is technically unattainable anyway. For serious egalitarians, the challenge is not: how can we get a few people from poor families to join the rich and powerful? It‘s: how can we overthrow the rich and powerful.

Being on the left

Open Left is asking people what it means to be on the left today. I’m struck by how much, as a Leftie, I’m unhappy with many answers.
For one thing, there seems to be an odd optimism about human nature. Jessica Asato says: “I believe that human beings are more co-operative than competitive.” Polly Toynbee says: “Life on the left means trusting that the better side of human nature can prevail against selfishness and greed.”
Relatedly, there’s faith in something called people‘s “potential”. Lewis Iwu says he is angered by “the under-utilisation of potential in this country.” Jessica says: “I do not believe that an individual’s character and station in life is pre-ordained at birth.”
This optimism about human nature, however, seems to coexist with a pessimism about the scope for institutional change. Mr Iwu complains about people “not going after the next promotion…because they feel alienated and that they do not belong to the ‘right’ club.” This seems to take for granted the existence of hierarchies. But why should we? Shouldn’t the left instead think about how to encourage the creation of fulfilling jobs, rather than tolerate “winner take all” workplaces?
Even more bizarrely, Francis O’Grady writes:

Because the relationship between employers and workers is a fundamentally unequal one, strong trade unions are an essential force for fairness in any healthy democracy.

Any healthy democracy? Isn’t it conceivable that one form of democracy would consist of if not abolishing then greatly mitigating class conflict, through the use of worker ownership and control, and policies (among them a citizens’ income?) that diminish the power of employers? Why should anyone on the left take for granted the inevitability of class conflict?
This “human nature-optimism, institution-pessimism” is, if anything (and to simplify), the mirror image of my view.
Not everyone has the “potential” to overcome the disadvantages they are given by the lottery of birth. This is why I believe redistribution is so important; my conception of justice requires that we compensate people for bad luck over which they had no control.
What’s more, people tend, given the opportunities, to be greedy and stupid. For this reason, giving power to rulers, be they politicians or bosses of companies, will lead to both unjustified inequality in incomes, as the powerful extract rents, and to potentially catastrophic failures - as we saw with banks - because organizations are just too complex to be run from above. The solution to these problems is democratic control.
In this sense,  I support radical institutional change - greater redistribution plus equality of power (a notion Open Left’s respondents seem to give little weight to*) precisely because I have no romantic illusions about “human nature.”
* James Purnell is a curious and I fear insincere exception here.