Stumbling and Mumbling

Bayes theorem: an application to Afghanistan

I said yesterday that MPs should know their Bayes’ theorem. To see its usefulness, let’s apply it to the question: do the deaths of British troops show that the mission in Afghanistan is failing?
Bayes’ theorem lets us answer this if we know just two numbers (well, actually four, but two are just one minus the others).
First, what did you think was the probability of success in Afghanistan before the mission began? This is the prior probability, which we’ll call Ps. The probability of failure, Pf, is one minus this.
Second, what is the probability that we’d see the number of deaths we have, if the mission were succeeding? Call this Pd|s. One minus this gives us Pd|f.
Let’s pin some numbers on this. Let’s call Ps 0.8. I guess this is the sort of number the government had in mind when the mission began. It can’t be much lower, because no-one would send troops to war without a high probability of success. But it can’t be much higher, because that would be over-confident, given that the British and Russian empires both failed to control Afghanistan.
Our second question is trickier. It could be that the mission was always going to be bloody, and in fact we’ve had fewer casualties than expected. On this view, Pd|s is high, and the deaths are no evidence of mission failure - in fact, the opposite. On the other hand, there was a vague hope that troops would “leave without firing a shot.” From this perspective, Pd|s is low, and the deaths are evidence of mission failure.
Let’s split the difference, and call Pd|s 0.5 so Pd|f is 0.5.
Now let’s take Bayes’ theorem to derive a number for the probability of the mission succeeding, in light of the casualities we‘ve suffered. This is Ps|d, the posterior probability:
            Ps|d = (Pd|s x Ps)/[(Pd|s x Ps) x (Pd|f x Pf)]
Plug the numbers into this, and we get:
            Ps|d = (0.5x0.8)/[(0.5x0.8) + (0.5 x 0.2)] = 0.8
In other words, the posterior probability is no different from the prior. The reason for this is simple. I’ve constructed the example so that the deaths are no evidence either way. So there’s no reason for the prior probability to change.
But what if the deaths were more consistent with failure? If Pd|s were, say, 0.2 and Pd|f 0.8, then our posterior probability would be 0.5.
Now, I write this not to take a view on Afghanistan. I know as much about Afghanistan as Abu Hamza does about playing the banjo. I do so rather to show how Bayes’ Theorem can help us think about the issue. I’d stress two uses of it.
First, it forces us to ask of any piece of information: what is this evidence of? Are deaths evidence of mission failure or not? It’s in this context that we need specific military expertise.
Second, we can reverse engineer the theorem. Rather than plug numbers into it and accept the answer, we can use it to ask: what numbers give us a high probability of success? How plausible are they? Again, military expertise can illuminate here.
Note that I am not saying that Bayes’ theorem is sufficient. It can’t define “success”. Nor can it tell us whether the success is worth the costs. It is, though, surely an important part of how to analyse the issue. And yet - in public at least - politicians do not think this way. They did not tell us Ps or Pd|s (or even ranges thereof) before the mission began. Nor do they tell us now. We can all think of reasons for this omission. But are they really satisfactory?

What should MPs know?

The FT asks: what should MPs know? Its list of questions contains a glaring omission. I would ask: describe Bayes theorem, and discuss some common deviations from it.
The reason for asking this is simple. It is a cliché that MPs, and especially ministers, must exercise good  ”judgment”. We should therefore expect them to know what judgment is, and to be on guard against the many ways in which it might fail.
However, I doubt if one in 20 MPs could satisfactorily answer my question.
The failure here, though, is not just MPs’. It’s also the media’s. For example, the FT’s panel of “experts” - they define the term sufficiently loosely to encompass Norman Lamont - does not contain a single decision theorist. I suspect even fewer journalists than MPs could recognise Bayes’ theorem.
This raises the question.  Why is there so little interest among MPs and the media in the nature of judgement?
One possibility - which I fear looms large in the FT - is that there is an unspoken presumption that the political class - MPs, candidates and senior civil servants - possess judgment anyway. After all, an Oxford education teaches one to think, doesn’t it?* One effect of this presumption is the belief that the irrationalities uncovered by behavioural economists apply only to little people and participants in financial markets, not to the political class. So, for example Kevin Barron, MP for Rother Valley, can say “we should not be frightened of taking decisions on behalf of our constituents, because that is to the general good” (via) without asking: why should our decisions be better than those of our constituents?
I fear, though, that there is a second possibility. Ministers’ and MPs’ claim to possess “judgment” is not an empirical claim at all. It is instead an attempt to legitimate their power. “Judgment” is intended to be a mystical entity which lies beyond scrutiny.
 * It doesn’t - it teaches one to bluff.

Tories & smackheads

The Tory Party is less popular than heroin. That - if anything - is the  lesson from the Glasgow North East by-election result.
The Tories got 1075 votes. The best estimate (table 3 of this pdf) is that there are over 13,000 “problem drug users” in the city of Glasgow - defined as users of heroin, methadone or benzodiazepines. If we assume Glasgow North East contains one-seventh of these - Glasgow NE being one of seven of the city’s parliamentary seats - then it has just under 1900 problem drug users.
It’s likely, then, that there are more smackheads than Tory voters in Glasgow North East.

Wage inequality falls

Wage inequality is falling. This is one message from yesterday’s Annual Survey of Hours and Earnings.
My table shows some measures of this. It shows the ratio of the 90th percentile of gross hourly wages for full-time workers to the 10th percentile. (That is, the wages of someone earning more than 90% of all full-time workers to the wages of someone earning less than 90% of all workers. The numbers include bonuses).
The ratio for all workers rose markedly between 1997 (when ASHE data starts) and 2005, but has fallen since then.
What’s more, the 90th percentile of senior managers have seen their pay fall relative to the 10th percentile of all workers ever since 2002.
And the 90th percentile of workers in the finance industry are doing no better, relative to the low-paid, than they were in 2001-02.
On these measures, inequality has fallen.
There are, though, some big caveats here.
First, these figures tell us nothing about the earnings of the super-rich - the top 1% or 0.1%. The 90th percentile shows only the lowest wage necessary to get you into the top 10% of workers. And this is quite low. It was £25.90 an hour in April - equivalent to just over £50,000 a year for someone working 40 hours a week.
Yes, that’s right. Just over £50,000 a year gets you into the top 10% of earners. Media chunter about “middle class” incomes  bears little relationship to reality.
Secondly, the fall in the ratio of “top” finance salaries to the bottom 10% of workers happened only last year, and might be a only a temporary artefact of the crisis.
Another caveat is: does this matter? If declining inequality reflects reduced rent-seeking, then great. But if it is due to lower productivity growth among higher-skilled employees, then there might be cause for concern.
The figures on wage differentials tell us nothing about how inequalities arise. For me personally, though, this is more important than the mere size of them.

Capital & full employment

Duncan has endorsed Keynes’ old proposal to maintain cheap and easy money, and so achieve “an increase in the volume of capital until it ceases to be scarce”.
I’m not sure this will work. We’ve tried something like it twice, and on both occasions it proved unsustainable.
In the 1960s, low real interest rates and tax breaks encouraged firms to invest. But this failed to secure lasting full employment because the high economic activity it engendered led to rising wages (and later oil prices) and so a profit squeeze so severe that firms could not make up in volumes what they didn't get in margins; Marglin and Bhaduri’s paper (pdf) provides the analysis.
Then in the 00s, we had the savings glut and easy monetary policies. This should have encouraged non-financial firms to invest. But they didn’t do so to anything like the level required to achieve full employment; even at the peak of the boom, there were four million out of work. It was housing that boomed in response to easy money, not corporate investment.
Quite why this should be is unclear. It just seems that, despite high observed profit rates on existing capital, firms couldn’t see profitable uses for further capital.
Experiments, then, seem to suggest that Keynes was too optimistic. Flooding the economy with cheap finance capital - at least on the scale that’s been tried - is not enough to produce lasting full employment.
I fear Keynes might have been misled by an ambiguity in the word “capital” when he wrote that “there are no intrinsic reasons for the scarcity of capital.”* Yes, there’s no good reason for the scarcity of finance capital - the cash available to firms to invest. But there might be good reasons for a scarcity of profitable (future) physical capital. It might be that a lack of innovative ideas or working class militancy does produce such a scarcity.
In this sense, Kalecki was far more correct than Keynes. He pointed out that there was no inherent tendency for capitalism to generate lasting full employment, and that it was possible that declining innovations could lead to sustained under-investment in physical capital and hence lasting unemployment**.
Its’ not obvious that Keynes’ proposal to socialize investment would overcome this. If capitalists can’t maintain sufficient innovations to maintain full employment, there’s no hope of the state doing so. And we know from looking at the Post Office today - or from many industries in the 70s and 80s - that mere state ownership does zilch to overcome the class conflict that Kalecki thought would prevent full employment. The only way state-controlled investment could secure full employment would be at the expense colossal allocative inefficiency.
I mention all this to raise a possibility. Conventional discussion of macroeconomic policy falls into two camps. On the one hand, there are those who argue that such policies can be effective in creating jobs. On the other are those who see them as distorting a market mechanism that would, if left to itself, generate full employment.
But what if there’s a third possibility - that both these views are half-right or half-wrong? It could be both that the economy isn’t a neatly self-righting mechanism tending towards full employment, and that macroeconomic policy interventions are insufficient to generate lasting full employment.
* He might also have been misled by thinking of physical capital as an aggregate. But there’s no such thing. There’s only this particular project, that one, and so on.
** Kalecki’s works are scandalously hard to get. I’m thinking of his papers “The process of economic development”, “The development factors” and “Political aspects of full employment.” This paper (pdf) by Malcolm Sawyer is a nice discussion of Kalecki’s thinking.

A feminine economy?

Our economy is becoming increasingly feminized. That’s one message of today’s labour market figures.
These show that, over the last 12 months, male employment has fallen by a net 447,000 whilst female employment has dropped just 43,000 net. Thanks to this, there are now almost as many female employees as male - 12.3m vs. 12.6m, though men still dominate in self-employment. This is the smallest gap on record.
One reason for this is that there’s been a massive shift from full-time jobs, which are mostly male jobs, to part-time ones, which are mostly female ones; in the last 12 months a net 641,000 full-time jobs have been destroyed, and a net 101,000 part-time ones created, 96,000 of them in the last three months.
There is, though, another reason. The industries that have suffered the most in this recession have been male-dominated ones. For example, manufacturing lost 253,000 jobs in the 12 months to June, 190,000 of them men. And business and financial services lost 259,000 jobs, 145,000 of them men.
By contrast, the sector that’s expanded is a female-dominated one. The education, health and public administration sector has put on 184,000 jobs, 131,000 of which have been women.
The question is: will this feminization of the workplace continue?
Two things suggest not. One is that, as the economy grows, full-time jobs should be created, to the benefit of men. Also, public spending cuts could well hit female employment more than male: 69.9% of the jobs in the  education, health and public administration sector are done by women.
On the other hand, though, deindustrialization is likely to continue - it has been going on for over 40 years. This would be detrimental to men, as these comprise three-quarters of the 2.16m jobs left in manufacturing.
Also, if we look at younger workers, there’s almost no gender gap in employment. Indeed, among 16-17 year-olds, there are more women than men in work. And among 18-24 year-olds, the gap is tiny: 1.65m women are in work against 1.75m men. By contrast, among 25-49 year-olds there are 1.5m more men than women in work. If this reflects an increase in women’s employability relative to men, then the gender gap in employment might continue to close, and even be reversed.
Personally, I’m not bothered either way by this prospect. But should I be?

What media influence?

How much influence does the media really have on our behaviour? This question is bedevilled by the problem of interpreting causality: do people vote Tory because the read the Daily Mail, or do they read the Daily Mail because they vote Tory?
This new paper (pdf) by Baris Yoruk sheds light upon this. He studied the “Give Five” campaign which ran in the US in the early 90s, which was intended to encourage people to give 5% of their income to charity and do five hours a week of voluntary work.
He found that people who had heard of this campaign through the media were more likely to give to charity and do voluntary work. But this is not necessarily because of the power of the media. It could be because people who wanted to donate to charity were more likely to seek out adverts and information about charities.
Mr Yoruk controlled for this effect. And when he did so, he found that there was no link between awareness of the campaign and charitable giving. In this sense, the media had no impact upon what people did.
However, there was an effect upon doing volunteer work.
Why the difference? It could be, says Mr Yoruk, that the campaign changed the behaviour  (a little) of people for whom there is a low opportunity cost of time. People who watch lots of TV, such as the retired or unemployed, can afford to give a little time for voluntary work.
Where the opportunity cost is higher - for example where money is at stake - the media have no impact.
How can we reconcile this with other evidence, which suggests that local newspapers can  (with exceptions) enhance local democracy, and with evidence that TV news can affect voting behaviour?
Maybe easily. Voting is “cheap talk” - its opportunity cost is low. And a preference that’s cheap is one that’s easily swayed.  When our own money is at stake, we are less susceptible to media pressure*. Perhaps what's going on here is a variant of Sayre's law - the media matter only when the stakes are low.
For me, this strengthens my prior support for demand-revealing referenda, because if people had to risk money when they voted, they’d be less swayed by the trash press.
* Yes, of course, our money is at stake in general elections, but only collectively so. What I mean is that there’s no individual link between how I vote and how much I lose individually. And this is what matters in shaping incentives.

Children & happiness

Having children makes you miserable. That’s the message of this paper by Luca Stanca, which draws upon data from 94 countries:

Having children is negatively related to subjective well-being. Conditioning on individual characteristics shows that the effect of parenthood on well-being is positive and significant only for widowers, older and highly educated individuals…On the basis of a purely economic approach, the optimal number of children for a rational agent is zero.

This partly corroborates evidence from the UK (pdf), which shows that children increase the well-being only of married couples and widowers, but reduce the well-being of single or separated parents.
However, the reason for this is rather mundane. Children make (many) people less happy only because they are expensive. Mr Stanca finds that children improve people’s satisfaction with non-financial aspects of their life, but worsen their financial happiness. This corroborates some other evidence (hat-tip).
Even so, this raises questions. Why do people not eventually adapt to having children, with the result that their happiness is unaffected by them? If having children reduces well-being for so many people, is this a failure of rationality? If so, why? Is there just a problem of self-control here, or the power of the desire to conform, or some other bias? If it's rational for (many) people not to have children, why does such a choice often bring with it huge stigma? 
And if it is irrational to have children, why should we respect the rights of parents over those children - for example, their right to give them a religious education?

Against a transactions tax

Gordon Brown’s proposal for some sort of transactions tax seems to have been rejected by Tim Geithner, which means the idea is, in effect, dead: such a tax is only remotely feasible if applied globally. This is good. There are (at least) three arguments against such a tax.
1. It would have done nothing to have stopped this financial crisis, and might even have made things worse. The two markets upon which a tax would impinge most - FX and stock markets - played little role in this crisis; they were, as near as dammit, innocent bystanders. The tax would not have stopped RBS overpaying for ABN-Amro, would not have stopped HBOS making bad loans, and probably wouldn’t have stopped Northern Rock funding its mortgage lending by borrowing in wholesale markets.
What the tax might have done, though, is reduce the liquidity of mortgage derivatives. But this was, for many banks, precisely the problem. As Alistair Milne describes in The Fall of the House of Credit, the problem with many “toxic assets” is not so much that they were devalued by defaults, but rather that they became illiquid, untradeable. A transactions tax might have exacerbated this problem.
The only way such a tax might have helped is that it might have deterred mortgage securitization in the first place.
2. A transactions tax does not necessarily stabilize markets. It might do the opposite. As Andrea Terzi points out (pdf), such a tax doesn’t so much reduce short-term trades as trades with low expected gains. However, these trades are often stabilizing trades - those done by arbitrageurs hovering up pennies.
If the tax bears more heavily upon these than it does upon noise traders, then it might make bubbles more likely, not less.
Note that short-term measures of volatility don’t help us decide how great his danger is: if prices rise 1% a day, then daily volatility is zero - but we’ll soon have the mother of all bubbles. This paper (pdf) discusses the issues here
3. Tim is right. The burden of a transactions tax doesn’t necessarily fall upon evil bankers. If people anticipate lower liquidity when they come to sell, they’ll not pay such high prices now. The effect of a tax will then to be depress prices now, as Bank of England research suggests.
In this sense, a tax bears upon people sensitive to current share prices - such as those coming up to retirement - rather than traders.
If these objections are valid, you can draw two possible inferences. Standard libertarians might say that interventions in markets are (often) ineffective or counter-productive. Leftists, though, might show that it just shows that the financial system needs much more radical change than merely new taxes.

Immigration & localities

The house next to mine is up for rent. But I have no say over who the tenant should be. Is this right?
I’m prompted to ask by Martin Wolf’s argument for immigration controls.
He points out that immigrants add to congestion. But if next door is rented out to a three-car family, I’ll suffer from extra congestion. Why do supporters of immigration controls think I should have no say over this, and yet should be able to control the numbers of people moving into areas I never visit?
Wolf goes on:

Diversity brings social benefits. But it also brings costs. These costs arise from declining trust and erosion of a sense of shared values. Such costs are likely to be particularly high when immigrants congregate in communities that reject some values of the wider community, not least over the role of women in society.

Now, leave aside the dog whistle he’s blowing here. This is also an argument for me to have a veto over next door’s tenants. If these are untrustworthy, or don’t share my street’s values - say because they play loud music or pollute the neighbourhood - I lose. But, again, I don’t lose if untrustworthy types move into an area miles away.
Surely, if social cohesion has any meaning, it is because of the benefits it gives real people in actual streets. I can see the possibility of cohesion between me and my neighbours, but what does it mean to speak of cohesion between me and people living hundreds of miles away?
The question I’m driving towards is: why should immigration controls apply only at the national level? Insofar as immigration brings costs, it does so because of particular individuals - those who “reject some values of the wider community” - moving into particular neighbourhoods. Doesn’t this mean that it would be better for people to control who moved into their locality than the state to try to manage immigration at the national level?
Of course, such controls would in many cases be blind to race or nationality. I - and I suspect the overwhelming majority of people - would much rather my neighbour be a decent foreigner than an obnoxious Englishman.
So, if we must have immigration controls, why don’t we have them at the micro-level? One possibility is that these would violate property rights - the right of landlords to choose their own tenants. But nationwide immigration controls also violate such rights. Another possibility is that we just make a fetish of the state.
Is there a more coherent reason?

Jedward & social choice

Is it just me, or do several X Factor contestants look like Arsenal players: Danyl and Eduardo; Joe and Cesc, Jedward and Nicklas Bendtner’s bollocks?
Which is a preface to saying that Jedward raise an important point about social choice theory - namely, that the public‘s “choice” can depend upon the voting rule. Take for example, three plausible rules:
1. First-past-the-post. On this rule - if the Sun is to be believed - Jedward would win, thus heralding the End of Days. This can happen because, with eight contenders, just 15% of the vote might get you victory.
2. Vote to reject a candidate. On this rule, Jedward would get the heave-ho, as most X Factor viewers do have some sort of hearing ability.
3. Sequential voting for one’s favourite candidate, with the loser eliminated at each stage. This is the rule the X Factor uses. And on this basis, Jedward are likely to be eliminated sometime, as Danny points out.
Depending on which rule we use, then, Jedward are either the most popular act, the least popular, or middleingly popular. "Public opinion" is indeterminate.
Sunder sees this as an argument electoral reform - for adopting an X Factor-type rule.
I’m not sure. This gives weight to second, third or even lower order preferences, which might be weak. This might explain why some X Factor winners - Steve Brookstein, Leon Jackson - have not had much success; the public just didn’t care enough about them to buy their records. It's moot whether weak preferences should count in politics or not.
Instead, my inference here is an Arrovian one. If we have so little information about public preferences - just votes, rather than any data about the reasons for or intensity of those preferences - then it can be impossible to identify for sure what the public want.
Which in turn raises a question about the nature of democracy. Could it be that what is valuable about democracy is not the voting rule as such? Instead, what matters is the lead-up to the vote, as discussion reveals the strength and (ir)rationality of preferences. If so, what is needed is not so much electoral reform as greater deliberative democracy.

Johnson, Burke & MPs' pay

It’s insufficiently appreciated that Alan Johnson’s sacking of David Nutt has direct bearing on the question of MPs’ pay.
What I mean is that there are two views of what an MP should do. On the one hand, there‘s the Burkean view, that MPs should exercise independent judgement:

Your representative owes you, not his industry only, but his judgment; and he betrays, instead of serving you, if he sacrifices it to your opinion.

On the other hand, there’s the notion that MPs must follow public opinion.
Now, the Burkean view suggests we should pay MPs the sort of salary good professionals command, as Philip Stephens proposes, because we want them to have similarly good, independent judgment.
However, if MPs merely follow public opinion, there’s no need for such high wages. Any pub bore can echo the prejudices of the mob. 
Which brings us to Johnson. In rejecting scientific evidence, and pursuing a drugs policy that merely panders to the most base and ignorant public opinion, he is rejecting the Burkean view  in favour of the populist one.
But if our representatives are to do this, why should we pay them as if they are taking complicated decisions? I can see a case for paying people good money for sifting scientific evidence, weighing arguments and making tricky judgments under uncertainty. But if they are just reading Daily Mail editorials, we should pay them as much as this skill demands - which is peanuts.
So far, so clear-cut. But here’s the problem. Could it be that the choice between the Burkean and populist functions hasn’t been a deliberate one at all, but rather is a product of our age?
Put it this way. I know nothing about environmental, foreign or military affairs. I would be quite happy for decisions on these to be taken by people of integrity and intelligence, if only such could be found.  However, this attitude - though, I suspect, common 50 years ago - is increasingly rare. In our ego-driven culture, people not only hold ignorant views, but expect these views to be respected, listened to and acted upon. The Burkean MP has become an impossibility.

Tories & the pound

Are the Tories planning on a huge fall in the pound? This is one question raise by Giles Wilkes’ paper, Slash and Grow, in which he argues that it’s unlikely that capital spending and/or net exports will grow sufficiently to keep GDP growing well in the face of government spending cuts.
Giles is, in effect, arguing against this recent paper (pdf) by Alberto Alesina and Silvia Ardgna, which shows that there have been several precedents for cuts in public spending proving expansionary.
My hunch is that he’s right to do so.
In the 1980s and 90s, it was possible for spending cuts to be accompanied by economic growth because: monetary policy could loosen in response; government bond yields could fall, thus crowding in capital spending; and because household debt was on a rising trend because, in many cases, consumer debt markets had only just been liberalized.
However, Giles rightly says these offsets are not so readily available to a future government. Index-linked yields are already zero for shorter-term maturities, so there’s no chance for orthodox crowding in.  Monetary policy is already loose, and even optimists can’t see much chance of significant increases in consumer borrowing.
There is, though, one other mechanism through which government spending cuts might be expansionary - if they cause a collapse in sterling.
Imagine it's 2010-11, and Osborne announces big spending cuts. The Bank of England responds by keeping interest rates low. However, the Fed and ECB start to raise rates. The UK could soon end up with almost the lowest rates in  the world. Carry traders around the world will then short sterling. The pound will fall, possibly very sharply. This effect would almost certainly swamp any uplift the pound gets from improved confidence about the public finances.
Would this boost growth by raising net exports?
Only to a limited degree. For one thing, there are long lags between exchange rate moves and export performance. During these lags, FX markets will question whether sterling is helping, and might therefore sell pound even more.
Also, many exporters respond to a lower exchange rate by raising prices more than expanding sales.
And even if sales do rise, many UK exports have a high import content - think of all the imported raw materials and components - which mitigates the effect on growth.
All of this means that the pound would have to fall a very long way indeed to offset any fiscal tightening.
Which poses my question. Is this what the Tories are hoping for? And if it isn’t, what is the mechanism through which a fiscal tightening will boost growth, given that the obvious ones can’t be as powerful now as they were years ago?

Competition vs safety

Alistair Darling says he wants a “safer, more competitive banking system.” But isn’t there a trade-off between these two goals?
I mean, competition generates instability, because firms that lose crash out of business. Schumpeter’s “perennial gale of creative destruction” will blow some companies away.
We saw this in the UK mortgage market in the earlier 00s. This was competitive; margins were bid down as some incumbents, notably Northern Rock, expanded and alternative lenders entered the market. But - we now know - it was also unstable. Not only did Northern Rock collapse but non-bank lenders have now all but disappeared.
In this context, we should ask whether Darling’s proposals to split up the banks really will increase competition.
The reason for doubt is simple. Competition requires not just there be many sellers, but that the sellers want to expand; it’s companies’ desire to sell more that bids prices down. However, if banks are required to have high capital-asset ratios, or large liquidity buffers, then their ability to expand will be limited. And this limits competition.
It might be this that explains a curious finding by the World Bank - that, across countries, there’s no significant link between concentration in the banking sector and net interest margins. Numbers alone aren’t always sufficient to produce real competition.
I say all this not to dismiss Darling’s proposals, but rather to raise a more fundamental question about our attitudes to economic systems. Fierce competition gives us innovation and lower prices, but also instability and uncertainty. Is the gain worth the pain?
It’s easy today to believe that the costs of instability are prohibitive. But is the availability heuristic biasing us towards this? Or is it instead the case that, as James Tobin said, it takes a lot of Harberger triangles to fill an Okun gap?

Drug laws, cognitive biases & democracy

Everyone seems to agree that the Home Secretary’s sacking of David Nutt as chairman of the Advisory Council on the Misuse of Drugs shows that policy is based upon prejudice rather than evidence, and that the function of experts is not so much to inform policy as to legitimate it by cloaking it with scienciness.
This is true. But it misses something. In rejecting Nutt’s advice, Johnson is pandering to a large chunk of public opinion. Which raises the question: why is it that public opinion is more supportive of drug laws than the evidence would warrant? Here, I suspect, are the relevant cognitive biases:
1. The fear of the unknown. Most of the supporters of criminalizing drugs have, I suspect, not actually taken them; I’m excluding newspaper editors and journalists here. And people are naturally scared of the unfamiliar - hence the Frankenstein complex and (a few years ago) the fear of GM foods.
2. The status quo bias. People prefer existing evils to unfamiliar ones.  For this reason, Nutt’s point that cannabis or ecstasy is less harmful than nicotine falls on deaf ears simply because, as Laban says, we’ve been smoking for 400 years but doing Es for barely 20.
3. Representative heuristic and ignorance of unintended consequences. Having decided that drugs are evvviiilll, people think that a policy of cracking down on drugs is somehow associated with less use of drugs -  they think outcomes must somehow resemble inputs.  What they don’t appreciate so much are the adverse effects of such laws: the gang wars as dealers fight for the high profits caused by restrictions on supply; the diversion of police time; the higher crime as addicts must steal more to pay higher prices.
Here, though, comes a problem. At least one recent opinion poll shows that these biases are not universal and that majority opinion - just - favours some legalization of drugs.
Which raises a depressing possibility. It seems that when public opinion is wrong - for example on immigration - politicians pander to it, but when it is right they ignore it. The function of representatives in representative democracy, it seems, is take all the idiocies of public opinion, and when these are insufficient, to then add some of their own.