Stumbling and Mumbling

The Kentish bias

England have dropped Geraint Jones. This is a surprise, because it flies in the face of a longstanding bias amongst the Test selectors in favour of players from the home counties in general, and Kent in particular.
This chart shows what I mean. It plots the number of Test caps won by players from each of the old counties (I'm excluding Durham) since 1946 against the number predicted by a simple two-variable model. The two variables are:
1. A home county dummy. This equals 1 for Essex, Kent, Middlesex and Surrey, and 0 for the others.
2. The number of county championship titles multiplied by a dummy if a county contains a Test ground. Logic tells us that stronger counties - as measured by the number of championships they've won - should produce more Test players. Oddly, though, this is true only of counties with a Test ground.
These two facts alone explain three-quarters of the variation in counties' test caps.
What's even more amazing  is the size of the home county effect. The home counties have, on average, 206 more Test caps than other counties. As the average county has only 345 caps, this is a lot.
However, Kent have produced even more caps than this effect would predict - 611, against a predicted 423.
This, I guess confirms what fans outside London have always suspected - there's a bias towards (spit) southern players.
Another thing: did I mention that I don't have a girlfriend?

Ignoring research and logic

An ounce of anecdote is worth a ton of research. The Hon. Jackie Ashley reports:

Joan Ryan, a Home Office minister, has reportedly warned her colleagues of anecdotal evidence from Southampton that the effect of the new migration has been to depress wages for low-paid workers.

This flies in the face of more rigorous research (all pdfs) which shows that immigration doesn't depress wages.
There are good reasons for this. First, as Tim says, immigrants spend their wages, thus increasing demand for labour.
Second, many immigrants  fill jobs which native workers don't do. The Polish plumber is a good thing for low-paid native workers. He doesn't drive their wages down (they're not plumbers) but he does buy beer and burgers, thus increasing demand for bar staff and fast-food workers.
But let's say - contrary to rigorous evidence and theory - that Ryan is right. So what? The Hon. Jackie says:

Migration is a class issue...Who benefits? Certainly, the service companies employing cheap cleaners, security staff, receptionists and the rest. Certainly, too, the middle-class households enjoying a boom in the servant sector - not butlers and maids, but cleaners and plumbers, nannies, painters and decorators.

But it's a howling fallacy to infer from this that the solution is immigration controls.
Put it this way. Pretty much any economic change - higher gas prices, Albanian immigrants, whatever - is a "class issue", insofar as it changes the distribution of income. But it would be absurd for the state to make ad hoc interventions in every case. Leaving aside (which we shouldn't) the effects on liberty, the government just doesn't have the expertise or administrative machinery to intevene.
Instead, the solution is a simple, general rule - a tax system that ensures that the benefits of immigration are spread more equitably.
But hey?  Groan writers are in favour of the exercise of state power, rather than equality, aren't they?

What skill?

Anthony Bolton's retirement as manager of Fidelity's special situations fund is inspiring many tributes like this (reg req). It's also inspiring some bad economics.
Jonathan Davis (who's usually excellent) says Bolton's great returns - his fund has turned £1000 into £130,000 over the last 27 years "defies the findings of a thousand academic research papers that no professional investor can beat the market by so much for so long."
Well, his performance might defy a thousand bad papers. But it's entirely consistent with this good one. Eugene Fama and Ken French say value stocks - those on low price-book ratios - have out-performed the market around the world because they compensate investors for taking distress risk. That is, the danger that such stocks will do badly in recessions.
Anthony Bolton has merely exploited this fact. My chart shows that the performance of Fidelity special situations, relative to the All-share, is closely correlated with consumer spending. In recessions, his fund under-performs. In buying "under-valued" stocks, Bolton has merely taken distress risk and been compensated in normal times. Just as French and Fama would expect.
Except for one thing. If we control for retail sales and market returns, Fidelity special situations has actually under-performed. A simple regression of annual returns since 1980 upon annual returns on the All-share and annual growth in retail sales gives us:
   Fidelity = -9.76 + 1.13 x All-share + 3.95 x retail sales.
R-squared = 67.7%.
The intercept of -9.76 implies that Bolton would actually lose almost 10% in a 12-month period when the All-share and retail sales volumes were flat.   Hardly a sign of skill.
Now, I'm not saying here that Fidelity's has been a bad fund. Quite the opposite. For many investors, it's entirely sensible to take cyclical risk. Personally, I'd take cyclical risk rather than trust someone's stock-picking "skill" every time.
I'm only saying that Bolton's performance is consistent with the efficient market theory, properly understood.

Lesbians' pay

Tim complains about yet more research on the gender pay gap. But here's a queer thing. The same issue of Centrepiece reports on the pay gap between lesbians and heterosexual women.
And guess what? Controlling for obvious things like education and occupation, lesbians earn 11% more than heterosexual women.  Most of the male-female pay gap, then, is a penalty for heterosexual women only, not for women in general.
There are four possible reasons for this:
1. Heterosexual women suffer a pay penalty because of the risk they'll leave to have children. Lesbians suffer this less.
2. Selection bias. The sort of people who admit to being lesbians in a survey are likely to be more than averagely confident. This could raise their pay.
3.  The pay penalty that women suffer is a penalty for psychological gender rather than physical gender. There's some academic evidence (not just old jokes) that some lesbians have more masculine psychological traits than straight women, as measured by (say) the Bem Sex Role Inventory. If it's masculine psychology, rather than male physicality, that gets a pay premium, this would benefit some lesbians.
This, though, is a mixed blessing. Other research has found that masculine psychology is a drawback when investing (pdf).
4. Maybe lesbians have unobserved characteristics that make them attractive to employers. Perhaps the fact that growing up knowing they are in a minority makes lesbians more adaptable or creative. Certainly, there's strong evidence that some are geniuses.

Assortive mating and social mobility

It's very difficult for governments to increase social mobility. This is one lesson I took from this paper in the latest Economic Journal (draft here):

For both men and women [in both the UK and Germany] 40-50% of the covariance between their own permanent family income and that of their parents can be attributed to sorting in the marriage market...Assortive mating appears to be a major factor in the intergenerational transmission of economic status.

In other words, the benefit of having rich parents doesn't just come from inheriting good genes, or from getting a better education. It comes from being more able to marry a high-earning spouse.
The authors show that selective education can exacerbate this problem. They show that assortive mating is even greater in Germany than in the UK, partly because the education system is highly selective, and so introduces smart boys to smart girls early on.
This finding will be a blow to anyone hoping that policy-makers can increase social mobility - because there's not much governments can do to reduce assortive mating, without appalling infringements of freedom.
This corroborates one of my priors - that governments should worry more about inequalities of wealth and power, and less about inequalities of opportunity.

Choosing money

This excellent piece by Jamie Whyte contains a passage that needs unbundling. It's this:

I know bankers who work very hard and earn a lot of money. They know that more wealth will not increase their happiness as much as more leisure would. Nevertheless, they keep working. They prefer money to happiness.

How can we explain this behaviour? I reckon there are five different possibilities.
1. Revealed preference. The fact that they work long hours shows that they prefer money to leisure. For them, money does make them happier. Sure, this isn't true of everyone. But people have different utility functions. It's a howling fallacy to beleive that what's true of the average is true of all.
2. Intertemporal substitution. In working long hours, bankers aren't maximizing happiness this week or this year, but rather lifetime happiness. They work hard in the hope of retiring early.
3. Youth. Younger people have reasons for preferring work to leisure. They have the energy to get up early, so the marginal disutility of work is lower. And the marginal utility of income is higher - say because they've got debts to pay, or because they get more of a buzz from fancy cars and clothes than older folk get. As Jamie's banking friends get older, many will downshift, just as I did.
4. Corner solution. For bankers, the choice of working hours is simple - it's 80 hours a week or none. They just don't have the choice of trading off a few hours lesiure for less income. They are forced to a corner solution.
5. The ambiguity of happiness. Bankers love the challenge of their job, and the sense of fulfillment it brings. This fulfillment is not "happiness" in the sense of easy pleasure, but it is - for them - a deeper happiness. For this reason, I'm not happy when Jamie says that some academics have "sacrificed happiness for discovery." Could it not be that, for them, the greatest happiness lies in discovering things?
You can argue the toss about the relative importance of these five factors. What's important, though, is that none entail any irrationality on the part of bankers. And only point 4 even hints at any case for state intervention.
The lesson for me is the one Jamie gives.  Give people freedom, and they'll get as much happiness - in the sense that they define it themselves - as they want.

Against CGT on housing

The Conservative Party Reptile asks what would be the consequences of imposing capital gains taxes on housing, as Ross Clark recommends. I think they'd be disastrous.
First, they'd cause a fall in consumer spending. This is because many people - around a quarter according to the Pensions Commission (pdf) - are using their housing wealth as a form of pensions saving. A tax on this wealth would cause a drop in anticipated retirement income. To make up for this shortfall, people would cut spending now.
And then the paradox of thrift would hit us. If we all try to save more, demand and output fall and so we all end up with lower incomes and less ability to save.
Second, such taxes could increase labour immobility. Let's say CGT is only applied when a house is sold - which I guess is the only feasible possibility.
This will make people even more reluctant than they are now to take jobs which requires them to move house. This will have two effects. Unemployed home-owners will prefer to stay unemployed (at least for a while), rather than move in search of work. And employers who want to lure workers to move will have to pay even higher wages. The upshot is that the natural rate of unemployment - the rate consistent with stable inflation - will rise.
This effect could be significant. Andrew Oswald gives us evidence here (pdf) that the housing market is an important influence upon the NRU.
So, I reckon CGT on housing is a bad idea. But I don't want to be too hard on Ross. There is an argument for taxing housing to ensure optimal ownership of the housing stock; a case for high council tax is to deter single people from occupying excessively large homes. And I share his discomfort at the British enthusiasm for higher house prices.
But there are better  solutions here than CGT. Better planning regulations are one possibility. A tougher inflation target (or one that takes account of house price inflation) is another. And ideally, more of us would work from home, thus allowing more people to sell up from expensive areas.
Another thing: I reckon Ross is spot on to want higher inheritance tax. It's surely absurd that you must pay tax if your dad gives you a job, but not if he gives you money in his will. But this is another story.

The defensive anomaly

Speaking of the brown stuff reminds me of fund managers. Here's another tale of their utter uselessness.
In my day job, I've been pointing out for years that low-risk stocks do better than they should. This chart shows my point for this year. It shows the performance of a basket of the 20 stocks with the lowest volatility of monthly returns in the previous five years; it's rebalanced every quarter.
Not only has this basket out-performed the market. In the last six months it's easily out-performed every fund in Trustnet's database of all companies unit trusts. In fact, similar baskets of defensives have out-performed almost all funds ever since 1999.
But here's the scandalous bit. The tendency for defensive stocks to do well is one of the oldest-attested quirks in the market. It was pointed out in this classic paper way back in 1972.
And yet, judging by performance, no unit trust manager seems to exploit this anomaly. Like our lame-brained politicians, they think their "judgment" is better than the evidence.

Colour and character

Clothing doesn't just reflect character - it determines it. For example, according to this paper (pdf), teams that wear black - the colour of evil - play more aggressively than other teams:

Many inherently evil characters choose to wear black. However, the present investigation makes it clear that, in certain contexts at least, some people become the bad guys because they wear black.

What, then, should we make of a "team" that wears brown? Iain Dale has the answer.

Compensating for disability

Many victims of terrorist attacks are grossly under-compensated for their injuries, according to this paper by Andrew Oswald and Nattavudh Powdthavee.
They've calculated the income disabled people need in order to have the same life satisfaction as the able-bodied. For the seriously disabled - those unable to climb stairs or dress themselves - this is £143,000 a year. For the less disabled, it's £58,000 a year.
However, Danny Biddle - the worst injured of the 7/7 victims got a lump-sum payment of less than £119,000 - equivalent to an income of under £6,000 a year.
Things do, though, change a little once we allow for the fact that the disabled get used to their condition, so their subjective well-being rises in the years after their injuries.
For the less seriously disabled, estimate Oswald and Powdthavee, this habituation is almost 100% after 3 years - which suggests compensation (expressed as an annual income) should fade away to almost nothing. But for the seriously disabled, adaptation is much slower, which means compensation should stay high.
The bottom line here is that compensation policy for victims of accidents and crime doesn't bear much relation to economic evidence, psychological research or justice. But I guess you knew that.