Stumbling and Mumbling

The price elasticity of marriage

Tory proposals to give tax breaks to married couples might cause a surprisingly large increase in the number of people getting married. That’s the implication of this new paper (pdf), which concludes that “even small changes in the cost of marriage can have significant effects.”
The authors established this rather cunningly. In 1980, 34 US states required people to take blood tests before they got married; these could cost up to $200, though usually less than $40. Between 1980 and 1995 most states abolished this requirement. By comparing marriage rates in states that abolished the test requirement to other states, the authors can then estimate the effect of a small reduction in the cost of getting married.
And they found that, on average, the blood test requirement reduced the number of marriages by 2.8 per cent - that’s a large number for a small cost. And the effect is bigger for the poorer or less educated.
This suggests that demand to get married is quite price-elastic. If so, a tax break for married couples might increase marriages quite a lot.
Does this mean such a policy would be an effective way of reducing “crime, educational failure and worklessness” which, says Iain Duncan Smith, are “all fuelled by breakdown of family life"?
I’m not sure. The reason why small price changes increase marriages a lot is probably that there are lots of happily cohabiting couples who are just about indifferent between getting married or not. Cutting the price of marriage tips these into wedlock. But it’s not these couples who are, for the most part, responsible for children being badly brought up. That responsibility lies instead with the couples who only met briefly for a quick knee-trembler in the doorway of JJB Sports. It’s less clear that these will marry for a tax break, and even less clear that they’ll become good parents by doing so.
So, even if marriage tax breaks increase  the number of marriages, they might not do much to mend our (allegedly) broken society. Maybe Laurie’s right - it is mere fetishism.

Public schools & charities: a paradox?

Should public schools have charitable status? There’s a paradox here - the people who should favour such status actually oppose it, whilst those who should oppose it actually support it.
To see what I mean, bear in mind that the case for charitable status is not that private schools save the tax-payer money. If this were the argument for treating public schools as charities, we’d also regard private health insurance, or private security guards, as charities. But we don’t.
Instead, the case for public schools being charities is that education might be a public good - its benefits spill over to society as a whole. This is why private healthcare doesn‘t have charitable status, despite it saving tax-payers‘ money; its benefit are purely private.
But is education a public good? Herein lies my paradox. 
For education to be a public good, its social returns must exceed its private returns.
But this claim is typically associated with New Labour, at least in its early years. Blair’s claim that his priorities were “education, education, education”, and the target of having 50% of youngsters in higher education, were founded on the notion that education is a social good. In some post-neoclassical endogenous growth theories, to use Brown’s phrase, there can be external benefits to education which cause the economy to grow quicker if people are better educated.
So, the first leg of my paradox is that those New Labour characters who have been most enthusiastic about the economic role of education should be most supportive of public schools’ charitable status - as they are the people who think education is a public good. But it fact, many of these seem to oppose such a status.
Equally, though, there’s an inconsistency in those who defend public schools’ charitable status. Many of these - such as Tim? - are those who, in other contexts, argue that unregulated labour markets work reasonably well. But if this is the case, the benefits to education are largely private. If workers are paid their marginal product, then the gains from education - better human capital - are captured by private individuals, in which case schooling is not a social good, so it shouldn’t have charitable status.
So, many people’s positions on the charitable status of public schools seems inconsistent with their attitude to labour markets.
So, who’s right? Is education a public good or not? The evidence is mixed. Some researchers have found (pdf) spill-overs (pdf), but others have been very (pdf)  sceptical.
There is, however, a third possibility. Maybe education is a positional good. If so, it has negative spillovers - better education for me makes you worse off. One survey (pdf) has found that over a third of people do indeed regard schooling as a positional good. 
This might be because education raises people’s self-confidence by more than their ability - which makes them a pain in the neck for the rest of us. Or it might be because their education puts them into positions of power, from which they are better able to exploit others.
If this is the case, then there’s a case not only for withdrawing public schools’ charitable status, but for taxing them.
Now, you might object here that I’ve taken a narrowly philistinistic economists’ perspective. I have. But it’s not obvious that a wider view alters things. After all, for every public schoolkid who is an adornment to the nation, such as Helena Bonham Carter or Hugh Laurie, there’s a Piers Morgan or Ed Balls.

Macroeconomics & the black box

The Economist’s article on the state of economics is rather good. But there’s one good point it makes which needs developing, namely its criticism of modern macroeconomics for having “too little interest in the inner workings of the financial system.”
This failure, though, is part of a wider and longer omission - it is not a mere quirk of DSGE models. Macroeconomics has traditionally ignored the inner workings of all companies. It has seen non-financial firms as mere production functions, transforming inputs into output in a stable (and smoothly differentiable) way. Regarding banks as black boxes for converting savings into loans is a natural consequence of paying insufficient heed to institutions.
However, we can only understand this crisis by looking inside the black boxes. We can do so from two perspectives.
One is the Austrian angle, which stresses the role of entrepreneurship, of discontinuous innovation. But this played a crucial role in the crisis. Banks used new mortgage derivatives as collateral to raise wholesale funds. This had the effect - we now know! - of increasing the riskiness of that funding. Macroeconomists’ natural tendency to assume stable production functions distracted them (us) from the consequences of this innovation.
The second perspective is that of principal-agent problems.
It is only in this context that criticisms about banks’ handling of risk makes sense. Intellectually, everyone has known about tail risk, correlation risk, liquidity risk and counterparty risk for years; the crash of 1987 and the collapse of LTCM in 1998 taught us all about them.
The problem is, though, that agency problems within banks militated against understanding these risks.
For example, who is most likely to keep his job as risk manager - the guy who tells his board “our value at risk today is £112.5m” or the guy who says “we dunno the default risk on our CDOs because we haven’t enough data. It’s really hard to quantify correlation risk or counterparty risk and the maths of non-Gaussian copulas is really tricky”?
The fact is that people in power very often want precise answers, not true ones - especially when it‘s only other people‘s money they are managing. Offering precision rather than truth is therefore a trick one can use to climb up the greasy pole.
Or take another example. Which trader is going to get more capital and more autonomy from his bosses - the one who says “these assets are mis-priced: we can make a fortune here”, or the one who says: “these look cheap, but they carry all sorts of liquidity, tail, correlation and counterparty risks”?
It’s agency problems, then, that help us understand why banks behaved as if they were really moronic.
Again, mainstream macro’s “black box” mentality led it astray. Economists listened to Taleb railing against the risk management  and said - rightly - “no-one’s that stupid.” But they (we) failed to consider that institutions can be structured so as to filter out intelligence.
In other words, macroeconomics failed in this crisis because it has for ages under-rated the importance of the stuff we (might) learn in industrial organization classes.
Herein, though, lies a quirk of history. There’s one significant figure who cautioned against this error - Marx. In Capital (ch 6) he criticized orthodox economists for ignoring “the hidden abode of production.”
But mainstream economists - including Keynes - ignored him. They did so, in large part, because they were not especially interested in Marx’s question of how income was distributed between profits and wages.
So, could it be that a by-product of a lack of interest in this question has been “black box” thinking that’s proven very costly for mainstream economics? The amazing thing is that macroeconomics managed so well for so long with this omission.

Tom Watson, age and mental strength

Tom Watson’s brilliant round at Turnberry yesterday, at the age of 59, highlights again the question of the relationship between age and performance. A new study of PGA Tour golfers by Harold Fried and Loren Tauer sheds fascinating light on this.
They looked at the efficiency of golfers - the way in which they convert ability (as measured by factors such as the number of greens made in regulation or driving length) into earnings. This is a measure of golfers’ mental toughness. A player who shoots two great rounds but then falls apart is inefficient, whereas a less able one who keeps his game together will finish better, earn more and be more efficient.
They found that there’s an n-shaped relationship between efficiency and age. Players’ mental strength increases up to age 36, but then falls off.
This is because there are offsetting effects of age. The good news is that age brings wisdom and a sense of perspective and hence ability to cope with pressure.  But the ability to concentrate wanes as folk get older.
You might think that older men, with more secure wealth and reputations, also have lower motivation - but the Yerkes-Dodson effect tells us this has ambiguous effects.
This relationship suggests we shouldn’t expect Watson to keep up yesterday’s great round - a fact which is reflected in his odds of 100-1 to win the championship.
Except for one thing. The n-shape is only a weak relationship - there’s a lot of statistical noise around it (though there is a selection bias in the research which might exaggerate this noise).
This is, of course, not just a story about golf. It corroborates other evidence suggesting that the fall-off in productivity as people get older is weak.

Can governments increase growth?

Hopi says politicians should do more to promote economic growth.
Of course, it’s easy enough to boost short-run growth - just throw enough money around. Promoting long-run sustained growth is, however, much trickier. It might even be impossible. This paper concludes:

Except for the decades around WWII, there is no evidence of country specific effects on long run growth rates…Growth rates are determined by international factors, and are insensitive to national policies, especially for small countries. This implies severe restrictions on the ability of most governments to increase national long run growth rates.

To get an idea of what they mean, here are some annualized real GDP growth rates for some significant countries between 1980 and 2007. I present the figures in ranges, such that we can be 95% confident that true growth is within this range. I do this because, even over a period as long as 27 years, it’s possible for two countries with identical true growth to differ if one has good luck and the other bad.
France: 1.7-2.5%.
Italy: 1.3-2.2%
Spain: 2.4-3.6%
Sweden 1.6-3.0%
UK: 2.0-3.2%
US 2.4-3.7%
These ranges suggest we can be pretty confident that Italy has done worse than the US or UK. But we cannot be at all confident that the  US has out-performed Sweden, or vice versa. The opposing poles of mixed capitalism - social democratic Sweden and freer market US - are consistent with similar, maybe indistinguishable, growth rates.
Big differences in institutions and policies, then, seem to generate similar growth rates. Which suggests that - at least within the wide parameters set by actually-existing mixed capitalisms - policies (or at least those that have been tried) might not make much difference to trend growth.
Perhaps, therefore, we shouldn’t look to national governments to promote long-run growth.
Equally, GDP growth should not be the test of the success or failure of policies. 

How many are unemployed?

Today’s figures (pdf) show that there are 2.38 million unemployed. This, though, is only part of the story. The figures also show that there are 927,000 part-time workers who’d like a full-time job (table 3) and 2.1 million economically inactive people (students, the ill, discouraged workers, home-makers, table 12) who’d like a job.
If we add these together, there are 5.4 million frustrated workers - those who’d like to work more than they are actually doing. This is equivalent to 17.3% - more than one-in-six - of the labour force*.
What’s striking here is that this measure has always been high. Even at the height of the boom, there were almost 4 million frustrated workers - equivalent to 13.5% of the workforce.
This suggests that Marxists are correct. Capitalism, at least in its actually existing form, is incapable of providing full employment.
In this context, Don Paskini is surely right. Trying to cajole the long-term sick into work by cutting their benefits is fatuous, as there’s little work to be had. Such a policy is much like Edward III’s Statute of Labourers - an act of class warfare lacking any decent economic logic.
Our biggest economic problem is a lack of jobs for those who want them, not the fact that some people don’t want to work.
* I say “equivalent” because the economically inactive are not counted as part of the workforce.

Age & productivity

The CBI has said it is disappointed by the government’s apparent decision to allow workers to continue working past the age of 65. Some recent research, however, suggests it shouldn’t complain.
The CBI would have a point if older workers were less productive than younger ones. If this were the case then a mandatory retirement age would be an easy way for firms to get rid of less productive staff.
However, Jan van Ours has gathered evidence from Dutch manufacturers to suggest that productivity declines only  slightly with age, on average. He finds that academic economists’ productivity - measured by publications in top journals - rises to the age of 50 and then only levels off, which suggests that experience (or the ability to exploit PhD students) can offset a decline in mathematical skills.
What’s more, even in pure physical activity, the drop-off in productivity is low. He cites Ray Fair’s estimate (pdf) that the time required to run half-marathons rises by only 1.1% a year on average between 55 and 65.
 And van Ours estimates that whilst the productivity of Dutch manufacturing workers does fall after age 50, the drop is small and offset by lower wages.
This evidence suggests that it is not the case, on average, that 65-year-olds are low-productivity, high-cost workers that firms are better off without. So a mandatory retirement age, generally speaking, is not needed to weed out bad workers.
So much for averages. But what of selection effects? It might be that the sort of workers who’d like to work past 65 differ from the average. I suspect, however, that such effects cut both ways.
On the one hand, the people who want to carry on working will be more motivated than average, which points to them being more productive. But on the other hand, someone who hasn’t made money by 65 and needs to carry on working is likely to have poor productivity.
I suspect then, that there’s no harm in allowing people to work past 65. I just hope they don’t make it compulsory.

Kevin Pietersen & rational overconfidence

Some new economic thinking suggests we should not criticize Kevin Pietersen for the way he got out in his first innings. Bruce Weinberg writes:

This paper develops a simple, single-period, rational model in which it is optimal to be overconfident and to down weight new information about ability…

To see why, imagine someone faced with a choice of a simple task with a low safe return - say a defensive shot - and a trickier task with a higher but more variable return, an attacking shot.
The person who is overconfident about his ability will choose the trickier task more often, as will the person who sticks to his belief in his ability in the face of conflicting evidence. Weinberg demonstrates that, under reasonable circumstances, such double irrationality will yield better outcomes:

Moderately overestimating ability leads the person to undertake the challenging task more often, yielding higher expected output…
an individual who underestimates the variance of his priors is more willing to choose the challenging task, which raises expected output and utility.

This theory maps well onto Pietersen. He is both overconfident about his ability - hence the name figjam - and unresponsive to evidence, as we saw in the excuses he made for playing that shot*.
But on average, this overconfidence has served him well; the availability effect has caused his critics to overlook all the runs he has made by selecting extravagant shots.
And, compare his test record to that of men of comparable talent but lesser self-belief, such as Graeme Hick or Mark Ramprakash. Their under-confidence  led them to scratch around defensively until they got out. Which corroborates Weinberg’s theory, that overconfidence can be a good thing.
This is not just a theory about cricket. It applies to less important matters. For example, moderately over-confident traders or chief executives, on average, might do better than ones with more accurate beliefs about their ability.
And it also raises a puzzle about the nature of rationality. Belief rationality - proportioning one’s perception of one’s ability to the evidence - can conflict  with act-rationality, the choice of best (on average) actions.
* You might object that he is not moderately but extremely overconfident, but I‘m not sure; his ability is high, as well as his self-regard.

Marxism & the mainstream

Is there a difference any more between Marxism and orthodox social science? I’m prompted to ask by this paper by Jeremy Greenwood. He writes:

A large part of social change is a reaction to technological progress in the economy. Technological progress affects society‘s consumption and production possibilities. It therefore changes individuals‘ incentives to abide by social customs and mores…
This notion is applied here to the rocket-like rise in premarital sex that occurred over the last century…This is traced here to the dramatic decline in the expected cost of premarital sex, due technological improvement in contraceptives and their increased availability.

This justifies Marx’s famous claim that  “The mode of production of material life conditions the general process of social, political and intellectual life” - such as how much people shag around.
Of course, Greenwood is by no means unusual in doing this. Ever since Gary Becker’s pioneering work (pdf), mainstream economists have shown how the mode of production - technology and people’s perceived costs and benefits - affects social life, such as crime, family structure or obesity. In Marxian terms, “base” does influence “superstructure.”
This, though, is not the only point of convergence between orthodox and Marxist economics. The basic premise of  neoclassical economics, that people respond to incentives, echoes the Marxian notion that individuals are bearers of social relations. Both stress that an individual’s behaviour arises from the position he finds himself in - which influences the costs and benefits he perceives - more than from his character. Of course, both views can be pushed too far. But both remind us not to see human action as rising from mere idiosyncratic disposition.
There’s a further point of convergence - economic history. Marx said that technology shapes property rights and hence class relations:

In acquiring new productive forces men change their mode of production; and in changing their mode of production, in changing the way of earning their living, they change all their social relations. The hand-mill gives you society with the feudal lord; the steam-mill society with the industrial capitalist. (The Poverty of Philosophy ch 2, second observation.)

Non-Marxist economists such as Terry Anderson and Peter Hill have shown how this happens. For example, the invention of trains increased the value of land in the US-mid-west, and so encouraged its transfer into private hands, whilst the invention of barbed wire allowed huge tracts to be enclosed.
So far, I’ve considered the convergence between neoclassical economics and Marxists. But there’s an important convergence between Marx and behavioural economics. Marxists believe that false consciousness can bamboozle workers into accepting capitalism. If we want to know how this happens, the cognitive biases and heuristics programme helps us. For example, the fundamental attribution error leads us over-estimate the extent to which the poor are to blame for their poverty, and to under-rate the importance of environmental or societal forces. The availability heuristic leads workers to blame immigrants for unemployment rather than less obvious forces. The just world phenomenon and system justification cause us to believe that capitalism must be fair. The status quo bias causes us to accept existing evils rather than risk new ones. And adaptive preferences (pdf) cause the poor to resign themselves to their fates and want less, with the result that capitalist democracy sustains inequality.
Of course, both Marxists and mainstream economists will reject all this, and come up with endless guff about “methodology” to explain why they really really hate each other, and would prefer to languish in their little intellectual ghettos.
But the fact is that they do converge upon many important points. And you know why? Because both see the truth.

Regulation vs nationalization

The Treasury today published its recommendations for regulating banks. For me, though, the question remains: why is regulation is superior to nationalization?
I ask for four reasons:
1. The case for regulating banks is entirely different from the case for regulating other industries. We do the latter for fear that, left to themselves, unregulated businesses will pursue profit maximization too aggressively, to the detriment of others - so drugs companies will skimp on testing, or railways will rip off customers*. But we regulate banks not because they'll make too much money, but for fear that, left to themselves, they'll lose too much.
Whereas ordinary businesses are regulated to protect others, banks must be regulated to protect themselves.  Which raises the question: if privately-owned banks can’t act in their own long-term interests, what is the case for private ownership?
2. The Treasury recognizes that existing ownership structures are flawed:

There must be major changes to the way that bank boards function. Improved risk management at board level, changes to the balance of skills, experience and independence, and a better approach to audit, risk and remuneration are required. Institutional shareholders need to be more actively engaged in monitoring the board of the bank in which they have invested.

But if ownership is the problem, why use regulation and exhortation? Why not take direct control and structure banks better?
3. Nationalization helps ameliorate the principal-agent problem within banks. Banks failed in part because key employees were paid huge bonuses for short-term performance, and thus took too much risk. This has led to calls for lower, or longer-term bonuses.
But each individual bank finds it hard to follow this advice. Each one fears that less lavish bonuses will cause “talented” staff to leave for banks that do pay extrvagantly.
However, if banks were nationalized, this co-ordination problem would be reduced, as nationalization would allow all banks to pay lower bonuses. Then, “talent‘s” options would  be fewer, to go abroad, or start up a new bank.
4. The notion that nationalized banks are bad for the economy might be mistaken. This new paper claims that:

The view that government ownership of banks is harmful to economic growth is unjustified. If anything, our findings suggest that government ownership of banks has been associated with better long-run growth performance.

One reason for this is that, even if privately owned banks do engender small gains in allocative efficiency, these can be wiped out by the huge costs of banking failures; it takes only one Okun gap to offset many savings of Harberger triangles.
Now, I am not saying here that nationalization is a good idea. It’s certainly not sufficient for good bank governance, although it might be necessary; our current passive control of Lloyds and RBS is inadequate. I’m just wondering why the government regards nationalization as a desperate last resort rather than a reasonable policy option.
* For present purposes, let’s ignore the merits or not of this argument.