Stumbling and Mumbling

Oil: a case for tax cuts?

Raedwald ponders the impact of $200pb oil prices. My question is: if this were to happen, wouldn’t it be a case for loosening fiscal policy?
The case for doing so is simple. Higher oil prices are, in effect, a tax upon oil users - which, directly or indirectly, is everyone. So, shouldn’t this be offset by a loosening in orthodox fiscal policy?
I am NOT talking here about a fuel price stabilizer. The incidence of higher oil prices falls not just upon motorists, but upon workers who lose their jobs because of the lower demand caused by higher oil prices. The case, then, is for a general fiscal loosening - lower taxes or higher spending - so that aggregate economic activity is unaffected by the oil shock, whilst we encourage people to consume less of the devils' excrement.
Let’s do some numbers. Last year, the UK consumed just over 1.6 million barrels of oil a day. This means that every $10pb rise in the oil price, if sustained for 12 months, is equivalent to a £3.7bn tax rise. That’s 0.25% of GDP. The $200pb oil discussed by Raedwald would, then, be equivalent to a tax rise of 2.5% of GDP - enough to cause a recession even in normal times.
Of course, only an arrogant charlatan can claim to be certain where oil’s going. But there’s a risk here. Oil prices are so volatile that a 37% rise in the oil price - equivalent to a 0.9% of GDP tightening - is a one-in-six chance.
In principle, the recessionary effect of this could be offset by looser monetary policy.  But there are two arguments against using this. One is that there is a (slim?) chance that the temporary inflationary effect of higher oil prices will become a permanent one if it raises inflation expectations. Looser monetary policy would add to this danger. The other is that looser policy would have to take the form of quantitative easing, and the effectiveness of this in boosting real economic activity is doubtful.
Fiscal policy, then, is the more obvious way of offsetting an oil shock. A cut in VAT of one percentage point would put £4.1bn (pdf) into the economy, just offsetting the effect of a sustained $11pb rise in oil prices.
So, why shouldn’t fiscal policy loosen in the event of oil prices rising? Two obvious counter-arguments seem weak to me.
1. “It would add to the deficit.” True, but not obviously relevant. For one thing, the revenues of oil producers are likely to be recycled into western government bond markets; in the last five years, oil exporters' holdings of US Treasuries have almost trebled, to $218bn. This would hold borrowing costs down. Also, if oil does take off, there’ll probably be clamour for safe assets - equity prices at $200pb would not look pretty - and this high demand for index-linked gilts will make the deficit easily financeable.
2. “Loosening policy now would reduce credibility.” This matters only to the extent that it would raise gilt yields - and I suspect (though we can’t be sure) any adverse effect here would be offset by point 1. It could also be mitigated by ensuring that the fiscal loosening were temporary - for example, by a one-off cut in VAT.
A more valid counter-argument would be the Ricardian equivalence one; a temporary tax cut would have no stimulatory effect as people would merely anticipate higher future taxes. This argument, however, implies that a loosening would do no damage either.
There is, therefore, surely a case for considering using fiscal policy to offset an oil shock.

Financial illiteracy: what's the problem?

Tim Harford worries about financial illiteracy:

You buy a new £1,000 computer and borrow money to pay for it. You have a choice: either (a) pay 12 monthly instalments of £100; or (b) borrow money at an APR of 20 per cent, meaning you pay back £1,200 at the end of the year. Which offer is better – or are they (c) identical?

Only 7% of Americans got this right. This is a remarkably low proportion, given that a very trivial principle would have given you the right answer - hang onto your money for as long as possible.
So low, in fact, that I suspect something else is going on. If we have to pay £1200 in a year’s time, we need to save and we might not trust ourselves to do so. We might therefore prefer answer (a) as a means  of committing ourselves to saving. In effect, we pay to enter a Ulysses pact.
This raises a general problem in establishing how rational people are: what looks like simply wrong behaviour can easily be rationalized. Granted, the paper (pdf) Tim got this from gives other examples of financial illiteracy. But these seem to be cases of simple innumeracy as much as practical ways in which people get into trouble.
This, though, is not my main reason for being sceptical about proposals to improve financial literacy.
For one thing, I fear that the cure here - even if there is one - might be worse than the disease. In some cases, a little learning is indeed a dang’rous thing because our confidence sometimes rises faster than our knowledge. Lessons in financial literacy might therefore increase people’s willingness to take on debt, in the belief that they can handle it.
And for another thing, I’m not sure how far financial illiteracy is a problem. Most people who get into debt trouble do so because they have fallen on hard times, or been overly extravagant or are just desperate - not because their stupidity has caused them to overpay for otherwise manageable debt.
So, what is the answer here? Partly, as Tim says, we need to simplify. Financial products - debt and savings instruments - are vastly more complicated than they need be. How far this is because of providers attempt to bamboozle folk into paying over the odds, and how far to bone-headed regulation is moot. 
Also, though, we need to recognize that financial education is subject to the Pareto principle. A few simple ideas get  you a long way - and these have little to do with maths. Not least of these is that, unless you buy an appreciating asset - which is less likely than you might think - borrowing will make you poorer. Acknowledging this, however, requires both the self-control to lead lifestyles within our means and a lack of illusions about one’s future; as  I’ve said, what matters is not so much financial planning as character planning.
The problem is, though, that it's not obvious that capitalism can thrive if people stick to these principles.

Overpaid MPs

Here’s some good evidence that politicians are over-paid. A new study of German MPs says:

The average politician earns more than the average voter, even after controlling for observed characteristics which are commonly identified to affect earnings. Depending on the estimation method and the respective specification applied, the politicians’ wage gap is 70–90% compared to the full sample and declines to 40–60% compared to citizens with an executive position. Hence, we show that the widely used claim that politicians would earn significantly more in the private sector is not confirmed by our data.

Does this apply to the UK? Granted, German MPs earn slightly more than their UK counterparts: the median MP in this data gets £73,000 whilst British MPs get £65738. It's unlikely that British MPs are greatly more skilled than German ones. And their pay is more than twice the median wage of 40-49-year-olds. Even if we apply a graduate premium of 25% to the median wage of 40-somethings, we get a salary of just over £36,000. Which suggests MPs are 80% overpaid.
You might object here that this ignores unobservable facts which affect earnings. It could be that MPs’ ambition, self-control and capacity  for enduring dull work and duller people would cause them to earn big money outside politics. But that requires evidence that there are huge returns to non-cognitive skills.
This raises the question. If MPs were paid no more than their experience and qualifications would suggest (say, £40,000), how would it change our politics? Would it really cause many talented would-be MPs to pursue other careers? And would this be a serious loss anyway?

Oil & unemployment

One thing that peeves me about reporting on oil prices is the tendency to see them only in terms of the “misery” they cause motorists. But history suggests that oil matters more than this. My chart - stolen and adapted from Andrew Oswald (pdf) - shows the point for the UK. It shows that oil prices predict unemployment, with a lag of around 18 months. Rises in oil prices in 1973-74, 1979-80 and in the mid-00s led to rising unemployment. And falling oil prices in 1986, 1998 and 2003 led to falling unemployment.
The reason for this is simple. Higher oil prices hurt companies as well as motorists. One reason for this is that if people are spending more on petrol they’ve less cash to spend on other things, so retailers suffer. But also - obviously - higher oil prices raise firms’ energy and transport bills. At the margin, this causes them to shrink production and lay off workers or even go out of business altogether.
So far, so simple. But four things complicate the relationship:
1. What causes higher oil prices? If prices rise because of supply restrictions - which was obviously the case in 1973-94 and 1979 - it is unambiguously bad for jobs. If, however, price rises reflect a global economic boom or loose monetary policy - as was the case in 2005-07 - their effect is ambiguous. This is because the things that cause higher oil prices also help create jobs. It’s not entirely clear which is the case now: there might be an element of betting on a worst-case scenario - rather than genuine supply problems - in the Brent price. 
2. What are firms’ expectations for oil prices and demand?  Firms do not lay off staff immediately as oil prices rise. If they think there‘s a good chance of prices falling back - which there is, given their volatility - or if they think economic conditions generally will improve, they might soldier on in the belief that the squeeze on their cashflow is temporary. It is only when they realize that higher prices and weaker demand are here to stay that they sack workers.
In this sense, responses to higher oil prices are like responses to exchange rates: the volatility of both dampens firms’ immediate reaction to them. This means its hard to find a precise mechanical relationship between oil and employment decisions, as the link is mediated by animal spirits.
3. How do wages respond to higher oil prices? If workers were to quickly and easily accept lower incomes to offset high oil costs, unemployment needn’t rise. However, in efficiency wage models, this doesn’t necessarily happen.
4. Are the firms most exposed to higher oil prices the same ones who are sitting on large cash piles, or not?
Higher oil prices interact with credit constraints. If a firm can’t raise the cash to see itself through the squeeze imposed by higher oil costs, it will be quicker to go bust or lay off workers.
Because firms, in aggregate, have been accumulating cash for years, you might think they are not credit-constrained. Such a belief, though, is an example of the representative firm fallacy. It is possible that the firms sitting on cash are not the same firms which suffer most from rising oil prices. To the extent that they differ, the oil price hit will be quicker and harder.
These four uncertainties mean we can’t say for sure whether, when, or by how much unemployment will  be affected by higher oil prices. What we can say, though, is that there is a danger here.

Pensions & cuts

The Times and Telegraph report that public sector workers could lose huge amounts of pension rights if their jobs are transferred to the private sector. Three quick points:
1. Pensions are only deferred wages. A cut in pensions is therefore a wage cut. If the private sector can only provide a service cheaper than the public sector because it cuts wages, what does this say about the supposedly greater efficiency of the private sector?
2. Final salary pensions are not inherently (pdf) better than defined contribution schemes. There’s no magic money tree that ensures that cash paid into a final salary scheme grows faster than cash paid into a DC scheme. Insofar as final salary schemes are generally more generous that DC schemes, it is only because employers pay more into them.
The issue here is not - or should not really be - final salary vs. DC pensions. It should be level of payments into them.
3. A big reason why pensions are so expensive is simply that annuity rates are nugatory - 3.5% for a 60-year-old man wanting an RPI-linked pension, which means you need pension pot of £571,000 to buy an index-linked pension of £20,000 a year.
However, this could change. A big reason for low annuity rates is that the global savings glut/investment dearth has depressed gilt yields, which have dragged down annuity rates. But this glut mightn’t last. It’s possible that some combination of reduced Chinese savings and increased (US?) government debt might, over the long-run, at least cause real gilt yields to return to their historically normal levels (2-4%). A one percentage point rise in annuity rates would reduce the cost of that £20,000 pension by 22%, to £444,000.
The claim that final salary schemes are unaffordable rests in part on the assumption that annuity rates will remain extraordinarily low - which is a strong assumption.

Jeremy Kyle & revealed preference

Many of you, I gather, watch Question Time or the Jeremy Kyle show - the two are indistinguishable  - only to express dissatisfaction with them. This is a common trait, with important implications for the TV industry and economic theory, as a new paper shows.
Researchers at the University of Milano-Bicocca ran an experiment. They got subjects to spend 10 minutes watching TV, where they had a choice of three programmes which they could switch between at any time. One was an 18th century costume drama, one a documentary about immigrants and the third a talk show. For some subjects, the talk show was a calm discussion and for others it was a violent argument.
Where the talk show was a violent argument, subjects watched significantly (27%) more than they did when it was a calm discussion. However, when asked their opinion of the show afterwards, they expressed less satisfaction than they did with the calmer show. This is consistent with Bruno Frey’s finding that watching more TV generally makes folk unhappy. The researchers conclude:

In the presence of arousing content, subjects watch more of a given program, against their own will and interest.

This is true whether the violent talk show is a low-brow gossip show or a high brow current affairs programme: as I say, Question Time and Jeremy Kyle are the same thing.
There are two implications here. One is for the TV industry. Viewing figures alone are no evidence of programme quality. It’s entirely possible for a show to get a big but unsatisfied audience - and for TV executives to chase such audiences by staging controversy.  This is not  just a snobbish opinion, but an experimental finding.
The second implication is wider. This result flatly contradicts orthodox consumer choice theory, which says that people choose the option which gives them the highest utility. In this case, they do the opposite. Free choices make them worse off. It is an open question how far this finding generalizes.
But even if it does generalize, it would be a large leap of logic to infer from the fact that choice diminishes utility that such choice should be curtailed. Utility isn’t everything.

Universities as bouncers

It looks as if many universities will charge the maximum £9000 tuition fee, to the annoyance of David Willetts.
Let’s unbundle this.  For a humanities student, teaching comprises lectures, tutorials and reading. What should these cost?
The LSE library charges business users £180 a year (five cards cost £750 + VAT), albeit for limited access rights. It’s unlikely  to be subsidizing such users, but let’s err on the side of generosity and price a year’s library use at £300.
Let’s also assume that teachers charge £50 an hour for tutorials and lectures, so someone teaching 25 hours a week earns £37,500 a year. Then a student buying the equivalent of four hours of one-on-one time a week - say, two one-to-one tutorials plus a contribution to lecture fees - would spend £6000 a year.
Combine this with library charges and we have a £6300 fee.
So, where does £9000 come from? Clearly, there are overheads such as the maintenance of wine cellars buildings and the cost of support staff. And it’s possible that humanities students will subsidize more expensive science tuition. But it’s not obvious why a student would want to pay for such things - or, indeed, why he should want to deal with the universities, rather than contract with teachers directly.
I suspect instead that what’s going on is that universities aren’t charging for merely for education at all. They are also acting like nightclub bouncers, limiting access to a prestigious good.
Universities offer more than tuition. They offer credentials, signals. And it’s these that they are charging for.
But they might be mistaken. A lot of the credentials you get from Oxford come simply when you get the acceptance letter, so you don’t need to pay £9000 a year to signal to future employers that you’re good enough to get to Oxford. What’s more, the benefit of this signal is short-lived anyway. Evidence from the US (pdf) (pdf) and Israel (pdf) shows that able people end up earning much the same, whether they go to a prestigious university or not. As Alan Krueger says:

Your own motivation, ambition and talents will determine your success more than the college name on your diploma.

This raises an interesting possibility. If potential students believe this, then - having got their acceptance letter from a “top” university, they would (in a free market*) shop around for a cheaper one. “Top” universities will then no longer automatically get the best students, unless they drop their fees. Which raises the question, of whether they will remain “top” at all.
My point here is one that should have been learned from the experience of top-flight football over the last 20 years. When you start charging a high price for something, you change both the product and the customers. And not necessarily for the better.
* Which we don't have, as students are limited in the options of university they can express.  Students are only customers in the sense of handing over money - not in the sense of having free choices.  

Armies & transactions costs

Alex points out that in Libya, as in Egypt, the government has been hiring expensive mercenaries to attack protestors. This raises an interesting issue in transactions cost economics, as inspired by Coase (pdf) and Williamson.
The obvious question is: why should a government resort to markets to buy protection when it has huge organizations - the police and army - which are supposed to do this job?
The answer, I think, lies in the fact that, occasionally, the alleged benefits of organization - that they reduce the costs associated with market transactions - actually break down. I’m thinking of three things here:
1. Incomplete contracts. It’s often impossible or expensive to write a contract specifying what a supplier should do in all contingencies. Where this is a big problem, internal organization is one solution. It allows the buyer to simply order the supplier around.
However, if a government looks like collapsing, this mechanism doesn’t work. The supplier of protection - soldiers - will not accept orders to kill protestors for fear that they will be prosecuted for human rights abuses if the revolution succeeds. In such cases, the implicit contract within organization is incomplete and unenforceable. 
By contrast, because mercenaries can disappear more easily - there’s no paper trail proving their employment - they are more able to get away with murder.
2. The hold-up problem. A buyer might want a supplying firm to invest in particular plant or area. The supplier, though, might be loath to do so, for fear that it will be left with unprofitable assets if it loses the contract with that buyer. In such cases, the supplier holds up the buyer’s production. The solution is for the buyer and supplier to merge, so organization replaces a market transaction.
However, this does not automatically solve the problem. Suppliers might still be reluctant to obey orders, if - given point 1 - they cannot be assured of a reward.
3. Asset specificity. It’s widely thought that if a buyer needs a highly specialized asset, it should own it directly, as this gives one greater control over it. By contrast, general purpose assets can be bought easily in the market.
The thing is, the very specificity of the Egyptian army made it difficult for Mubarak to control. Many Egyptian soldiers took a specific view of their job, figuring “I didn’t join up to fire on my compatriots”. Mubarak therefore needed a more general asset - thugs and arseholes - which he bought in the market.
As for why such arseholes are so expensive, I suspect there’s an element of Smith’s compensating advantages - governments have to compensate for the risk of being caught by protestors - but also of efficiency wages; governments have to buy a modicum of loyalty.
There are, I think, a couple of general principles here. First, transactions costs are such that they can cause markets to supplant organizations even in very unexpected places, such as the army. Organization does not necessarily solve the problems of incomplete contracts, hold-ups and asset specificity. A smarter writer than I would write a piece along the lines of "Mubarak's lessons for the AOL-HuffPo merger."
Secondly, even where markets are superior to internal organization, they can fail badly. Just ask Mubarak.

Their politics and mine

Norm’s criticism of this post of mine illuminates a way in which I often misunderstand politics.
For me, there are, roughly speaking, two types of political discourse. The dominant form is an exchange of tribal grunts; this is what passes for political debate in most blogs and, I understand*, the media.  The other sort, which exists largely in my own head - I never, ever, have political conversations with anyone - and in a few books, blogs and the occasional newspaper article, is a form of idealized dialogue about values, mechanisms and evidence.
Sometimes, though, I mistake one for the other. I did this, for example, when I tried to find meaningful substance in George Osborne’s letter to Mervyn King. And I seem to have done the same in reading Aaronovitch. I thought his criticism of scepticism about democracy referred to my sort theoretical arguments. Norm, however, thinks he was referring to something else - apologists for tyranny.
I’m happy to be corrected.
I’m also more than happy to acknowledge that my concerns and interests are peripheral, and are to be ignored by columnists. Why should the likes of Aaronovitch try to engage with the likes of me?
But there’s something I’m not at all happy with - that people are paid good money to engage in what is merely tribal grunting, and that I should think that what they do is interesting. Or, for that matter, important: are the people Aaronovitch is opposing really likely to be convinced by his finger-wagging?
* I say “understand” because I never watch any current affairs TV, or listen to shows such as the Moral Maze or Any Questions. From what I gather, though, these exemplify what I have in mind.
Another thing: I stand by my “suddenly everyone gives a damn about Egpyt” sneer. I was driving at two things. First, I was opposing the tendency for people to become instant experts on subjects they hitherto knew nothing about. Second, the extent to which we care about people should depend upon their amount of hardship, not upon their media coverage. Once we start doing the latter, we might easily stop opposing injustice.

AV: a doubtful argument

There’s one argument for AV that puzzles me. It’s the claim that, as Andrew Rawnsley says, “It will be a worthwhile improvement if MPs have to gather some form of support from at least half of the voters.” Both the Electoral Reform Society and the Yes to Fair Votes Campaign cite this as the first argument for AV.
But imagine the following pattern of votes:
Tory = 44%
LibDem = 31%
Labour = 15%
UKIP = 10%
It’s possible that, under AV, UKIP voters’ second preferences would all go to the Tory, giving him a majority.
But this is barely different from the present system. All that changes under AV is that Ukippers’ second preference, which is tacit under FPTP, becomes expressed. This is true of any constituency where AV returns the same MP as FPTP - which might well be the case in most constituencies.
Many MPs now do have majority support - it’s just that the support is tacit.
So, why should we put so much weight upon the difference between an expressed preference and a tacit one?
There is, of course, another possibility. This is that AV does change the MP. In my example, this would happen if Ukippers don’t express a preference and if Labour voters’ second preference switches to the Lib Dems.
However, in such cases, a large chunk of the majority vote is a weak and perhaps grudging preference. Why should this outweigh the larger and  perhaps stronger preferences of over two-fifths of voters? It’s entirely possible that, in such a case, a majority vote  produces an outcome with which there is more aggregate dissatisfaction - because the antipathy of Tories to the LibDem is stronger than Labour supporters’ endorsement.
It seems to me, then, that this case for AV requires further argument. You have to explain either why an expressed preference is better than a tacit one. Or you have to show either that majoritaranism will not clash with welfarism, or that where it does, majoritarianism should prevail. 
I do not say this to oppose AV: I’ve no view yet, though I suspect that the most interest ingarguments about it involve the possible behavioural changes it will cause. Instead, I just want to suggest that one argument for it isn’t self-evidently strong.