Friday, 27 February 2009

Government insurance and a paradox of externalities

HBOS today announces losses of about 6% on its total corporate loan book, according to Robert Peston (ignore the 47% figure that he also gives - that is only generated by cherrypicking the scope of the statistic to get the most serious-looking number).

Can a bank charge a sufficient additional premium on its interest rates to cover this size of loss? It's unlikely. Small businesses regularly pay a spread of 6-10% over base (I've heard from some bank managers that they are charging up to 15% to some) but most loans are to much larger companies which have been able to get much better terms in the past.

So, if a bank can't charge enough money to cover its costs, should it stop providing the service? Strict microeconomic logic says yes: the service is not viable in the marketplace at the current price, so its price should go up and the volume of loans provided should be much lower.

But this is a classic example of market failure due to uncompensated externalities. Most discussions of externalities focus on negative externalities - for example the cost of pollution which is not paid by the polluter. This means that too much of a service is provided, because the cost to seller (and buyer) is artificially low.

As Tim Harford points out in one of his books (I forget whether it's Logic of Life or The Undercover Economist) positive externalities sound nice at first. For example, beekeepers make honey, which they sell for a market price - and the positive externality is that crops and flowers are pollinated for free by the bees. However, positive unpriced externalities cause an economic distortion just like negative ones; it may mean that too little of a service is provided, because the provider is not being compensated enough to encourage the amount that society wants. If people stop eating honey, beekeepers will go out of business, few bees will be kept and crops will not be pollinated. You can imagine the knock-on effects of that.

In this case, the positive externality is the economic activity generated by the loan, and the multiplier effect that grows the economy as a result. I don't know if there is an accepted model for this, but I would expect that £1 billion of corporate lending (which has a market price of say £30-50 million depending on the type of borrowers) would generate perhaps £300 million of extra annual activity including around £75 million of profits for the borrower and about £100 million of tax revenue. So the externality is substantial, because the overall benefit is far in excess of the price being paid to the lender, and also well in excess of the benefit to the borrower.

This is a strong argument for the government to subsidise corporate loans so that the economy can keep going. Even on a simple cost-benefit count, the government may make a profit because of the additional tax revenues generated. And society will benefit by much more.

Subsidies to corporate loans in non-recessionary times cause other distortions - as does fiscal stimulus of all kinds - and this argument becomes much less strong. Indeed, the subsidy is unlikely to be needed in those times, because losses on corporate loans are tiny compared to the figures for this year. But in times when there is a significant output gap - times when Keynesian stimulus is justified - a subsidy through partial guarantees of corporate loans is completely justified.

Indeed, the existence of externalities means that these guarantees are the only way the free market can operate properly, without distortions. That sounds paradoxical but the logic is sound.

Thursday, 26 February 2009

Fred Goodwin's pension

Robert Peston has another example of an economic decision which looks odd from a simplistic rational point of view, but makes perfect sense when perceptions and expectations are taken into account.

Firing someone often makes an important statement about the direction of a company. But (as most of us would certainly wish) it's difficult for a company to fire people at will. Thus, companies usually have to pay off the individual - whether with a pension or a lump sum.

For a company the size of RBS, a £16 million payoff (or whichever part of this was actually discretionary) is easily worth the improvement in public image that comes with such a major break. Of course the payoff itself - since it inevitably has to be made public - has a perception impact, but that's probably much smaller than the damage of keeping the same person in charge.

Would you argue that Goodwin's actions were egregious enough to justify sacking without compensation? Most people have made judgment calls at work which caused substantial losses, though they are rarely as visible and hardly ever as expensive as Goodwin's. We tend not to run the counterfactuals over most of our work to see what opportunities we have missed or mistakes we've made.

Perhaps in future, CEOs will be more likely to have "at-will" style contracts where they waive their rights to compensation in advance. But it's a little pointless to argue that we should have had them in place ten years ago.

Transparency and mendacity

The Richard Thinks blog has an item about transparency today. It implies that The Economist has fallen for the self-serving nonsense of some anonymous traders:
'not having full-transparency allowed them to fully exploit the potential of secret trading strategies and that with full disclosure they would have little incentive to correct market inefficiencies through arbitrage.'
And:
'when investors start thinking that other people are privileged to lots more relevant information and that they have an unfair disadvantage they are likely to resist activity in the market.

So "[s]ymmetry, not the amount of information, matters"'
How convenient for the financial institutions!

The first argument is clearly fallacious. Arbitrageable market inefficiencies only arise from lack of transparency. If the markets were transparent, the market inefficiencies would be visible and there would be an immediate and high profile incentive to correct them. And a much stronger one - since it is less risky and more urgent - than whatever incentives the proprietary traders may have.

The symmetry argument sounds appealing at first, but given the intrinsic asymmetry of the financial markets (where a small number of 'insiders' or experts know far more than the masses), surely the release of information by insiders is much more likely to increase than decrease symmetry?

Richard himself then goes on to say that:
"full-disclosure can result in less transparency than you had to begin with, if the users of information are overwhelmed by the sheer volume of it, and its complexity"
While this might be a short-term problem, it's likely to disappear quickly. People are nowadays pretty good at providing tools to sift through and interpret data on behalf of others. Dennis Howlett's item about XBRL is a good counterargument to this - companies will spring up to provide simplified interpretations of this information if it is released by companies. This is what has always happened in the financial markets and I see no reason why it would stop now.

The disadvantages of transparency are so small and short-term, and the benefits so strong, that there can be little debate about the public good. The only question is the protection of existing 'owners' of information, and that is a question for politics, lobbyists and interest groups, not for economists. The economic recommendation is clear: release the data!

Confiscated savings and bank runs

One year ago, you invested £100,000 in a money market investment account.

Unfortunately, you chose to keep it with a small UK bank which has lent too much money to subprime borrowers. You signed up for a fixed-term deal where you receive 6% interest for four years in return for keeping your money in there.

Reading Robert Peston's blog one day, you find out that the bank is about to get into trouble. It has had to write off 15% of its asset base. Your deposit is not covered by government insurance and the bank is going to be partially liquidated; you are going to lose 11% of your money! Overnight your balance will drop to 89% of its previous value. The bank is still going to keep trading so you won't get your money back - the term deposit will remain in place, but at least they will keep paying interest on the reduced balance. What remains of the bank will be taken over by the government so there seems to be no risk of a further writedown.

You figure there is about 30 seconds to act before the news spreads and the bank bolts the stable door. You can get onto their website, pull your money out and transfer it to your current account. Because you are on a four-year fixed deposit, you would only get your original deposit back and not the interest paid so far. You might be helping to cause a run on the bank - irresponsible perhaps - but at least you'll have your cash. Decide now! Do you do it?


If you said yes, your motives are part of the reason for the financial crisis. But not because you caused a bank run.

Loss aversion - which means that we think a loss of existing resources is bigger than an equivalent gain in resources we don't have yet - is the only reason to answer yes to this question. If you take the money out, you probably won't earn more than 1-2% interest on it for the next three years. Your £100,000 may become £105,000 if you're lucky. But if you keep it in, even with the 11% haircut, you'll end up with £112,000.

High interest rates are linked with higher risk, but even when the risk goes bad, they may still be a worthwhile investment. Don't worry so much about capital loss if you are getting high interest in return. Capital defaults are rarely all-or-nothing - you'll get to keep something, and even in default you may well end up with more than if you invested in low-risk, low-return government bonds.

Wednesday, 25 February 2009

Practicality versus justice

Some commenters on Stephanie Flanders' blog (and of course in other places) are suggesting the banks simply be liquidated - no nationalisation, no bailouts.

This has the appeal of being fair - creditors who made bad decisions lose money.

But it may not make the system work well. If it destroys liquidity and causes people to become irrationally risk-averse, then justice may be served at the expense of the common good. By forcing a subset of creditors to take a hit to their assets - particularly those which are liquid and immediately available - we might cause knock-on effects which would destroy far more economic value than the counterfactual, the cost of a bank bailout. This is what the phrase 'cut off your nose to spite your face' was invented for.

It's impossible to know for sure what would happen if we did let banks fail. Some people are working out ideas on this. But the feeling in government is clearly that the damage would be so great that it's worth putting hundreds of billions in public money at risk to prevent it.

Could it ever be otherwise? What criteria (and what definition of justice) would specify an economic system where justice and utilitarianism always coincide? Where the means and the ends are always consistent? Could such a system exist? Your comments on this are very welcome.

Word cloud, B2B pricing and value modelling

Thanks to all Marginal Revolution readers who have stopped by in the last couple of days. I am looking forward to generating next week's economics word cloud on Sunday and will summarise whatever trends emerge in comparison to this week.

Eric Mitchell at the Professional Pricing Society has an interesting article about pricing in a B2B environment. I broadly agree with the comments of John Burdass and Rick Robinson; pricing should, as far as the supplier can influence it, be driven by value and not by either cost or competitive comparisons.

However they imply a focus on objectively demonstrable financial value. For me, value is highly subjective; and perceived value is driven by framing and other cognitive factors distinct from the objective attributes of a product. This is especially true in B2B service provision where direct comparison of a service - either against another supplier or against a similar service purchased in the past - is difficult.

Much of our recent research at Inon has been in this area. We use a technique called value modelling to build a hierarchical representation of subjective value for a given customer. For example, some customers may place a high value on winning new business; others on getting home on time each day. These values in turn can be expressed in terms of simpler, lower-level drivers such as material gain or pain reduction. The structure of the hierarchy, and the weighting of different factors, has a strong influence on decision-making.

Pricing interacts with this decision process in a complex way; we cannot yet model the whole process in detail but we do have some insights into it, and we have spend the last 18 months building software tools that provide a framework for setting prices so people are more likely to make the decision to buy.

I think that over the coming years we will see more real-world applications like this of behavioural economics and decision theory; as the models become better understood and closer to reality, the field has the potential to enable powerful microeconomic applications.

It's fascinating to engage with the theory in these two distinct ways: from the pure theoretical side of designing and exploring the models; and from the application side, making the models tractable and creating tools usable by people not trained as economists. If you are working purely on theory, or purely on tools, I recommend engaging with those on the other side too.

Tuesday, 24 February 2009

Misunderstanding management decisions

Robert Peston posts an odd blog today, which implies that he thinks all business decisions are either absolutely right or wrong, and not contingent on the situation or on business judgment.

He cites £1.5-2bn of costs which Fred Goodwin did not take out of RBS - despite his reputation as a cost-cutter - and which Stephen Hester has now identified.

Well, surely in today's business environment some costs are no longer appropriate which were nevertheless justified in 2007. The private plane which he cites may be the perfect example. When there is competition among investment banks to win new M&A instructions or IPOs, being able to pick up executive in a private plane for meetings in your office could be a good marketing investment. When there is no competition - or indeed no mergers - or when the executives become much more price-conscious, it may no longer be cost-effective.

As we've discussed before, it's dangerous to assume that apparently-extravagant expenditure is really wasteful. Companies have bought their private jets for a reason, and they organise sales conferences in Las Vegas for a reason too.

Undoubtedly there is an agency problem here: it's easy for executives to decide that an expensive sales conference is a good use of money when it isn't their money. And it may be too simplistic to just consider this as just part of the compensation package for managers - it may be much more cost-effective to instead give them a bit more money.

As a matter of public policy (since RBS is majority state-owned), if we are looking to stimulate the economy as a whole we don't necessarily want companies to cut back on their spending. We ought to be encouraging spending which maximises the velocity of money, as well as achieving economic efficiency.

I don't know the right answer - but all these factors merely show that this is a complex decision and that the answer is far from obvious.

The real reason for making certain decisions is symbolism. It's important for taxpayers who feel worried about the recession to see that high-profile people, especially those paid by the public, are making decisions that do not look extravagant - regardless of the management rationale behind them.

Fortunately the bandwidth of public discourse is limited, meaning that there is only room for so many high-profile people in one economy. Thus RBS can sacrifice its plane and make us all feel prudent, without every company and every person in the country having to cut back on their spending too. And a good thing too!

Monday, 23 February 2009

Why are recessions bad?

Stephanie Flanders (and earlier, Tim Harford) raise an interesting question today. As they point out, changes in individual circumstances tend to dominate the macroeconomic aggregates - increases and decreases in individual income are usually much greater than the 3% increase or decrease in the output of the whole economy.

Of course, a 3% decline in GDP means that a few more incomes have fallen than risen - but on the face of it, this is a minor effect.

So what explanations could there be for the fact that we worry so much about recessions?

First, some of the effects are very visible. Lots of people lose their jobs - and a million more people out of work will always create some high-profile news stories.

Second and related, the media (and the public) select stories which confirm their overall narrative. Any newspaper whose main focus of the last two weeks was on the 9,000 new jobs at KFC rather than the 850 job losses at Mini would not look credible - its narrative would be out of step with what its readers expect.

Third, there are two possible statistical breakdowns of the overall effect - each of which has its own negatives.

The first possibility is a general slow decline in incomes - everyone loses about 2-3% of their income across the country. Due to loss aversion, people feel very strongly about a reduction in income. Thus a 2% decline is much more noticeable than the 2% increase we all had the previous year. The secondary psychological effect (the "hope" versus "fear" effect) strengthens this.

The other possibility (which Stephanie and Tim have indicated is the case this time round) is a big divergence in incomes. 48% of the country gets rich and 52% gets poor. In this case, due to declining marginal utility of income, the effect on the happiness of the people who get poorer is much greater than on those who get richer.

For all these reasons, a relatively small recession can have a much bigger impact on our overall wellbeing and satisfaction than the numbers would suggest.

Sticky policy making

Three posts in 24 hours, Mr Peston? Back from your holiday and full of energy it seems.

One of the enduring phenomena in macroeconomics is stickiness. This most commonly shows up in the idea of 'sticky prices', which in some models are responsible for business cycles. In short, prices of goods and wages increase beyond the level of stable demand (perhaps driven by higher short-term demand or by inaccurate expectations of rising demand). The demand curve turns out to be a bit lower than was thought, but prices do not immediately adjust downwards in response. Therefore the amount of goods supplied is less than the optimal (equilibrium) amount and we get a recession.

There are many other examples of stickiness, or as I referred to it in an earlier post, friction. But one that I have not seen discussed is policy stickiness. Robert Peston highlights this in an item today about Northern Rock.

The British government is changing its original policy of running down Northern Rock's loan book to minimise losses to the taxpayer on risky loans. While this policy was probably the correct one in an economy which was still growing and credit-driven, it is unlikely to be right in a recession with a shortage of credit available.

Now the government intends to change Northern Rock's policy and start extending new loans. One key feature of these loans will be a higher loan-to-value (LTV) ratio than is typically available in the private sector - 80-90% instead of the usual 75%. This seems like a good idea, though it does remind me of the effect last year where Northern Rock - because it had the only full government guarantee of all deposits - soaked up an artificially large proportion of the savings market for a time. I don't think this would be a major problem in the current environment, but it is something to watch in the future.

But it has been reasonably clear for several months at least that this was the case. The government has been trying to encourage lending and it has a captive lender representing a large share of the mortgage market. I believe we have a rational government and that the Treasury provides good advice to its ministers.

So why hasn't policy changed earlier? There is clearly a stickiness. Part of this arises from criticism of "U-turns" - if the government changes its mind quickly, it is often pilloried for lack of resolve. Part arises from a desire to let a policy play out to see if it is going to be effective. Part comes from a game-theoretic calculation (whether conscious or not) in which government does not wish to be seen as easy to influence and thus create an incentive for private parties to try to change its mind instead of following its dictates.

So there is a stickiness in the private sector and a different stickiness in the public sector. We hope that the non-sticky parts of each sector will help to cancel each other out - emergency fiscal stimulus is a good example. A non-sticky, quickly adjusting economy is the holy grail of macroeconomic stability and is a goal worth spending lots of time on.

A completely rational actor in either sector will behave in a less sticky way than do real people. It will be interesting to analyse specific departures from rationality and see whether they imply that stickiness is correlated. If so, it will show us where we need to work to improve the flexibility of the economic system.

One of the successes of interest rate management in recent years has been the ability to flex without creating the appearance of a policy U-turn.

Economies need 'escape valves' so that the total amount of demand and supply - as well as the amount in individual sectors and regions - can adjust as appropriate. When the private sector is not flexible enough, and the central bank and credit system are at the limits of their flexibility, the public sector has to provide flexibility instead. Thus an ability to run less sticky policy could substantially reduce the impact of recessions and make a difference of tens of billions - or hundreds of billions in the case of the US economy.

In my list of cognitive biases, anchoring is the most obvious candidate to explore this issue; loss aversion and limited cognitive power have an impact too. Another item for the research list.

Update: Jeffrey Sachs has a different opinion - he wants to avoid short-term swings in policy (via Greg Mankiw).

Sunday, 22 February 2009

The economics zeitgeist, 22 February 2009



This is a word cloud from all economics blog postings in the last week. I'll be generating this every week so please subscribe using the links on the right if you'd like to be notified each time it is published.

It has been constructed from a list of economics RSS feeds mainly generated from the Palgrave Econolog, and uses Wordle to generate the image, the ROME RSS reader to read the RSS feeds, and Java software from Inon to process the data.

You can also see the Java version in the Wordle gallery.


Thanks to an anonymous poster on an economics blog somewhere for the idea. Unfortunately I haven't been able to find the original reference but if anyone can remind me who was asking for this within the last month or so, please let me know.

Friday, 20 February 2009

Surprised at the WSJ

The Wall Street Journal today says that the UK government
"...has already spent £600 billion on its financial bailout"
This is simply misleading. The government has made a lot of liquidity available in exchange for other assets through several Bank of England schemes. It has barely spent anything at all; even if you include the bank recapitalisations, which are also in exchange for equity, the figure is around 5% of the number quoted.

While this kind of stuff is to be expected from the Daily Mail and other tabloids, I am pretty surprised at it in the Wall Street Journal. I'm glad that the FT still maintains a standard of reporting that allows us to rely on its interpretations as well as its raw facts.

Thursday, 19 February 2009

Your advice wanted - starting a business in a recession

My friend Margie, who's at business school, is less optimistic than me about the recession. She thinks things will be dismal for a long time yet.

However, she's still considering starting a business when her course is finished. So what kind of company would you recommend starting up in a recession?

And if I'm right after all and the recovery starts soon, what sort of companies would be especially successful in a recovering economy?

Please suggest your answers in the comments, or by email to leigh@inon.com and I'll post a selection. The best recommendation wins a free skinny latte (other coffees are available).

BearingPoint (ex-KPMG) is bankrupt

Ed Kless at VeraSage points out that BearingPoint has filed for Chapter 11 protection, wiping out shareholders equity.

VeraSage campaigns for a switch from hourly-based to value pricing (mainly among accountants) and so Ed asks how BearingPoint could possibly have lost money on hourly contracts, unless most of its consultants were sitting around doing nothing for months on end.

However, even though BearingPoint came out of KPMG, I doubt they were charging on an hourly rate for most projects. In my experience consulting and IT firms are much more likely to use fixed-price agreements than accountants.

A number of such firms have made serious losses on these agreements in recent years - IT projects of course being notorious for going over budget. I would expect that this is the reason for BearingPoint's problems. While government IT projects used to be notorious for going over budget at the taxpayer's expense, they are now getting just as bad a reputation for causing losses on the supplier's side.

I certainly don't believe this is a strong argument against fixed-price agreements - in my own software firm we almost exclusively use fixed or value pricing - but it does highlight that pricing is an area where these companies seem to make some expensive mistakes.

I'm sure most firms make a profit on the majority of their engagements; but IT projects in particular are highly asymmetric. If you fix a price of $1 million for a project, the highest profit you can make on it is $1 million. But the loss you can make is, in principle, unlimited. Inevitably a large loss will occasionally happen, and can wipe out the profits from many other projects.

This makes it incumbent on consultants to become good estimators. There is a certain amount of research on this in the software field, but I think not enough. At any rate there is still no universally accepted method.

Many consultants will accept hourly rates when they can get away with it, specifically in order to avoid such problems. Others will write their fixed-price specifications skilfully and therefore be able to charge more when circumstances change. Others will get themselves stuck with too-loose specs and lose money on changes.

I don't know whether the greater acceptance of fixed price agreements among the IT sector is because they see the opportunity for greater profit, because of a better understanding of how to profit from intellectual property, because of a greater attachment to 'a logical way to do business' or because clients have got wise to IT overruns and put greater pressure on suppliers to fix their prices. Probably a combination of the above. I'm sure both the consultants and the accountants have things to learn from each other.

A story about evidence

There were two traders who used to cross the desert in Kenya, carrying incense from one city to another and metalwork back the other way. The trip was a long and hot one, and they travelled with two camels each and always with just a bit more food and water than they strictly needed.

For many years their trade carried on successfully - occasionally with a sandstorm, once losing some goods and camels to a thief, but never with a real crisis. Their wealth gradually accumulated, and they were able to expand their caravan to six and eventually ten camels; but no matter how rich they became, they never forgot to carry an extra waterskin and packet of cured beef: they were wise enough to know that riches cannot stop you dying of thirst.

But still it came to pass that on one difficult trip, the wind was strong and they had to camp for an extra day; when they reached the next oasis it was dry; and the spare water was soon gone. With three days to walk to the next water, and the sun stronger than ever, their mood was grim.

"What shall we do?" said the younger trader to the older. "I fear that we will never see the city again."

"My grandfather told me once of a time when he nearly perished," replied the older man. "Eighty years ago, in the greatest sandstorms of his time, he was caught in the desert with six camels and a cargo of wine. There was a tribe who camped nearby but they were hostile and he had only escaped recently from their swords. After a day of dehydration he was near to succumbing. In desperation he slaughtered a camel and drank its blood; drank a pint of wine; and finally he worked himself into a frenzy, drew his sword and raided the supplies of the tribe for milk and meat."

"And...what happened?"

"He lived to tell the tale, so it must have worked. Perhaps we can follow his example."

"But there is no tribe nearby, and we have no wine," said the young trader.

The older man nodded. "We still have the camels."

"Your grandfather's story is inspiring. But even if true, there's no evidence that drinking blood on its own will save us. I have worked hard for my camels and I will not kill them for nothing."

The old man sighed and drew his knife. "If you have no evidence, sometimes you only need common sense. We are going to die in this desert if we do nothing. I am slaughtering one of mine."

So he killed the camel, drained its blood, filled his waterskins and drank from one. And then they set off once again towards the distant oasis with the nine remaining camels.

One day later, in the height of the afternoon heat, the young trader gave in and begged for a share of the blood; his colleague offered it and they both survived to reach the oasis two days beyond.

Once back in the city they were soon able to replace the slaughtered camel and continue to build their wealth. The older man never mentioned the fact that he'd sacrificed his camel to save both their lives; but the young trader never forgot it.


In recent months I have frequently seen the complaint that there is "no evidence" that fiscal stimulus works, and therefore we should not try it. Sometimes, as the old trader said, evidence is not needed if you have enough common sense.

The moral of this story is: If you have insufficient liquidity, you should try anything that might work. Sacrificing some wealth in order to make it to the next oasis is better than dying rich.

Incidentally, the two traders retired and wrote their joint memoirs a few years later: The General Theory of Employment, Interest and Camel Blood, by Kenyans.

Wednesday, 18 February 2009

Bizarre blog links

Every so often someone links to an article on this site - recently, it's usually the "stimulus for bloggers" article, but occasionally something else. I usually discover it through Technorati or by noticing a new referrer in the server stats.

Generally this is simply because somebody found the article interesting and linked to it. Often they copy a few lines of the article to give context to their readers.

Today, however, I discovered a very odd phenomenon at this blog. My article on behavioural economics and the knowledge firm has been linked, but it seems to have been put through a machine translator - twice! As far as I can tell it has been translated into another language (the blog is in South Africa, so I would guess Dutch or Afrikaans) and back to English. Google Translator doesn't produce these specific results though, and neither does Babelfish. Anyone seen this phenomenon before?

The original, from my article here, was:
This blog has two primary themes:
  • behavioural economics
  • the economics of knowledge
I believe they are closely linked, because behaviour is derived from the knowledge that people have about the world (or more strictly their mental model of the world, which may not actually be accurate knowledge).

Knowledge, in the economy, is influenced by many things. But at least one type of entity specialises in influencing it: the knowledge firm.

Knowledge firms include professional services firms,
The quote on the page is:
This blog has two core themes:behavioural economicsthe economics of knowledgeI believe they are strongly coupled, because behaviour is derived from the skill that people have about the world (or more sternly their mental typical of the world, which may not actually be accurate wisdom). Knowledge, in the wealth, is influenced by many gear. But at slightest one kind of entity specialises in influencing it: the expertise fixed. Knowledge firms embrace professional army firms,
How curious! I can't quite imagine why someone would do this. But every link's a good link...right?

Note in particular that "services" was translated into "army" and "gear" to "things". Is this some kind of Googlespamming linkfarm specifically focused on army products?

Tuesday, 17 February 2009

Reshaping economics

I'm linking to Gregory Clark's article not for the same reason that some people have - enjoyable though it is to poke fun at professors - but to examine his suggestion that economics as a discipline is being reshaped.

There are a number of details which qualify only as interesting gossip - the suggestions on academic economists' salaries and the dismissal of the "not real" "Nobel" prize.

But the substance of the article is that:
  • the "mathematical contortions of academic economics" are useless
  • nobody predicted the downturn
  • economics has not advanced in 80 years; the terms of the bailout debate date back to the 1920s and 1930s
I'm conflicted about the last point. Selfishly, I want to believe it because, if true, it opens up more space in the discipline for new research such as my own. But I don't want to be working in a intellectually moribund field; and I want to believe that the rich and interesting range of work done since 1940 is actually worth building on.

As an aside, I have now worked (to varying extents) in four academic disciplines: mathematics, physics, computer science and economics. If Clark's assertion is true, the only one whose mainstream has advanced substantially since the 1930s is computer science, and that's mainly because it wasn't invented until then.

One could argue that the last serious, interesting and definitive advance in the mainstream of mathematics was Godel's theorem; in physics, the development of general relativity and quantum theory; and, according to Clark, in economics it's Keynesian fiscal theory (and perhaps game theory, around the same time). Naturally there has been lots of work since 1940 in all three fields, but I would not accuse any of it of advancing the core of the subject. Essentially everything taught at undergraduate level - in maths and physics at least - is pre-1940.

Now I don't really think this is the case for economics - Clark has forgotten monetary theory, to name but one - but it's surprising that the discipline has sufficient lack of credibility to invite such attacks.

Touching on the first two points - which are related - the mathematics of economists is perhaps a bit like the mathematics of meteorologists. It is good for predicting short-term behaviour - tomorrow's weather - and it can give lots of general rules which apply most of the time. But it can't predict a storm two weeks out, and there's still intensive debate about the extent and speed of global warming.

In fact, I know a meteorological physicist who has become an economist and now bemoans economists' lack of mathematical abilities. I kind of agree; it's not advanced mathematics that causes problems for economists, it's a lack of testable models and sufficient input data to calibrate them. Some people say that macroeconomic behaviour, because of complexity theory, can never be predictable. I am not convinced - I think there is room for huge improvement in the models and their power.

Despite that, economics still does a much better job than no economics. So let's do what most academic economists are busy doing: work at improving it without dismissing what it has already achieved.

Incidentally, the Atlantic also has a much better article along similar lines from Jim Manzi today, along with a nice followup note from Gregory Clark. Along related lines, this item from Anatole Kaletsky pushes for new directions in economic thought.

Schizophrenic inflation forecasts

Stephanie Flanders expresses an opinion which seems to reflect the conversation throughout most economics blogs, as well as - she suggests - in the City. That is, we are either due for severe deflation or high inflation - nothing in between.

Interestingly, the deflation story mainly comes from the authors of the blogs, while the runaway inflation opinion is mostly in the comments. You might say that shows which opinion reflects economic orthodoxy, but not necessarily which one is more likely to be accurate.

The most likely scenario may in fact be "both" rather than "either". Deflation in the short term, which we hope will be arrested by the immense monetary and slightly-less-immense fiscal stimuli coming out of most rich-country governments. Followed by inflation in a couple of years, which central banks will try to contain by mopping up excess money supply - with no consensus on their likelihood of success.

Most economists would say that the inflation is a price worth paying for avoiding a depression. Inflation does impose a deadweight cost because it makes prices harder to predict and compare, and distorts purchasing and investment decisions accordingly. But assuming we don't get hyperinflation (which is still extremely unlikely) this cost is probably less than the damage that would be done by a severe reduction in output.

On this blog I am interested in perceptions, knowledge and behaviour. What is the impact of inflation on these factors - or of these factors on inflation?

Hyperbolic discounting is one filter through which we can consider these effects. This means that people apply a higher discount to near-future deferral of income than far-future. The classic example is that if I offer you £100 now or £120 in a year, you'll probably take £100 now. But if I offer you £100 in a year or £120 in two years, you'll probably take £120 in two years. So you apply a discount rate of more than 20% now, and less than 20% in a year's time.

In principle this might reflect an expectation that inflation will be lower in a year than it is now, but the result seems to appear in experiments over time regardless of inflation levels.

In a world where we expect deflation today and high inflation in a couple of years, our money should be worth more in a year and less in three. This has problematic effects on both consumption and investment. Hyperbolic discounting could perhaps counter this.

I am a firm believer that government should use behavioural research just as much as classical economic theory when setting policy. Thus, by offering consumers a one-year advance on monetary transfers such as tax credits or benefits - or equivalently a deferral of tax payments - government can take advantage of its ability to borrow short-term at low cost (reinforced by expected deflation), and boost current spending without substantial impact on deficits. The repayments may need to be spread over a longer period to avoid a demand crunch in a year's time, but that is easily affordable in the context of current government borrowing.

So an additional fiscal boost now, financed by pre-announced tax increases in 3-5 years, should both have an oversized impact on current spending and a damping effect on expected medium-term inflation. Sounds like it's worth a try.

"In the real world...": aggregate demand versus price

"(In the real world things are more complicated, but never mind.)"
The rallying cry of economists through all history. This time it comes from Paul Krugman (the paper is a few months old but he points to it in today's column), but I don't mean to pick him out unfairly. It's just a less formal way of saying "we need to simplify our models otherwise we don't know how to work with them". Sometimes, though, it would be useful to be explicit about how far from the real world we believe our models are.

I just want to take up one point in the paper. Paul assumes that, under normal circumstances the AD (aggregate demand) curve slopes down and the AS (aggregate supply) curve slopes up. His point is that in a liquidity trap (like now and in the 1930s) the AD curve may not slope downwards after all, so standard results about the equilibrium point are not applicable.

Fair enough. But I would question whether the AD and AS curves behave as described even in normal times.

It may be another example of the one-versus-all fallacy. The demand curve for a single product will of course slope downwards (with rare exceptions such as Giffen goods). But this is because, for a given amount of income, we can choose between buying one good and another. Thus, the price of a product in the standard demand graph is the price relative to other products.

In the case of the aggregate demand curve, total demand is meant to be based on an average price level, but a price relative to what? If this describes total demand in the whole economy, what is the substitute good that we are supposed to buy instead?

One might say that demand for goods is a substitute for wealth - i.e. that the average price level is taken relative to the balance in our bank accounts. But I question whether this is really a substantial effect. While a hypothetical rational agent would annuitize their wealth at a certain rate and spend from it, the real-life phenomenon of mental accounting indicates that we spend mainly out of income, with little effect from total wealth. And income, according to our favourite accounting identity, is the same as aggregate demand.

If this is true, then the AD curve would be vertical all the time, not just in a liquidity trap. The effect of wealth, while small, is non-zero and so the curve may veer slightly off the vertical. But as Professor Krugman says, "never mind".

Intuitively, this does make sense. Price is, after all, just a number. It is there to arbitrate the relative value of things and not their absolute value. There can be no meaningful price for "all the output of the world economy" because if I wanted to buy everything, what could I trade you for it?

By analogous reasoning the aggregate supply curve is near-vertical too. And thus, on this theory, the total levels of production in the economy are not determined by price at all - not only in a recession but at all times.

So what does determine total production? In the short run, I believe real-time dynamic factors dominate - including the momentum of existing production and agreements, changes in prices (not absolute levels) or expectations of such changes, and people's perceptions of their need for goods relative to the immediate past or future. In the long run, it should be determined by trade-offs between absolute quantities: chiefly, I suspect, the relative preference for leisure versus goods.

Obama's Elf




I believe this piece is referring to the position of Secretary of 'Elf and 'Uman Services.

Monday, 16 February 2009

Behavioural economics and the knowledge firm

This blog has two primary themes:
  • behavioural economics
  • the economics of knowledge
I believe they are closely linked, because behaviour is derived from the knowledge that people have about the world (or more strictly their mental model of the world, which may not actually be accurate knowledge).

Knowledge, in the economy, is influenced by many things. But at least one type of entity specialises in influencing it: the knowledge firm.

Knowledge firms include professional services firms, consultancies, marketing and media companies. Their distinguishing characteristic is that their work is not about manipulating physical objects but influencing the minds of people. This is done by creating messages and communicating them.

Therefore, knowledge firms are in the business of second-order influence of behaviour. Traditional firms offer products to which people respond based on their existing perceptions and preferences. Knowledge firms actively change the perceptions and preferences that people use to make their decisions.

Fortunately for these companies, people are not rational. In a world of perfectly informed rational actors, there is really no role for this kind of business. It's the various departures from rational behaviour that create the opportunity - and requirement - for knowledge firms to exist. And so, by understanding those departures, we can help knowledge firms to operate more effectively.

The list of departures from rationality is long. Here is the list I've been working from in some recent research:
  • Hyperbolic discounting
  • Loss aversion
  • Habit-forming
  • Framing
  • Satisficing
  • Peer defaults
  • Heuristics
  • Limited cognitive power
  • Overconfidence
  • Risk miscalculation
  • Animal spirits
  • Reciprocity
  • Mental accounting
There are plenty of others, for example at Wikipedia. Depending on your model of the mind, many different items in these lists may originate from the same cause.

Just to take a couple of simple examples:
  • Marketing firms take advantage of framing to create adverts. By creating subconscious associations between emotional goods (such as comfort or success) and products (such as fabric softener or cars) marketers influence people to buy products that they might not otherwise buy, or to pay a higher price for the same thing.
  • Accounting firms allow people to use satisficing to make financial decisions. Instead of having to evaluate all possible risks and returns in a decision to invest or lend money, people can rely on an audit statement to decide that a decision is "good enough".
  • Law firms can take advantage of the existence of heuristics. A law firm, when it creates a contract that governs the behaviour of two parties, restricts the likely behaviour of each one to a limited set of dimensions. Each party can operate as if the other will behave according to the contract, and avoid the time required to predict and evaluate every possible behaviour.
Knowledge firms have little reason to exist in a world of rational-maximisers. Every firm should understand its place in the ecosystem of cognitive bias in order to know how to position itself, what skills to develop and how to sell its services.

BMW cuts 850 jobs and one shift

It's been accepted by most analysts for years that the global car industry has substantial overcapacity and that factory closures, or even closures of whole companies, are needed. Now BMW is cutting 850 jobs at its Mini plant - even though Mini sales are up this year. So is this a good thing?

Probably not - at least not right now - because it causes a reduction in aggregate demand, and that means reduced GDP growth, or a deeper recession.

But that's true of capacity reductions at any time. So why would it have been good before and not good now? What is the difference?

The first thing to understand is why a cut would have been desirable. Let's imagine that, in more normal times, a car company (Chrysler for example) shuts down. The immediate beneficiaries of a company closing would be the other car companies. With less competition they could sustain higher prices; some people who would have bought Chryslers will now buy other cars, boosting both revenue and profits at the remaining companies.

But this isn't just a matter of companies better off at the expense of consumers and ex-employees. The people who were making (and marketing, and financing) Chryslers would, after a while, find work making something else which is more valued by society - I assume this on the basis that if customer won't pay for Chryslers at the breakeven price, but they will pay for some other product, the other product is more highly valued. Ultimately resources would be employed more usefully.

There are certainly adjustment costs - both in the obsoletion of useful capital (machinery and those big factories in Detroit) and in a period of unemployment for employees. But in the long run - and even the medium run - these costs are outweighed by the greater productivity and usefulness of the alternative use of resources. The reduction in aggregate demand is another one of these short-term costs.

So far so good - orthodox economic theory at work.

But what about now, in the middle of a recession? The period of unemployment is likely to be much longer, and the leftover capital is likely to find no other profitable use for a while. Thus the adjustment costs of closing down a factory now are higher.

And will the reduction in aggregate demand have a greater effect now? It will certainly be more noticeable - but will it really make more of a difference?

Probably, yes. A reduction in private demand has a multiplier effect, just the same as an increase in government spending. One of the main determinants of the multiplier effect is the proportion of idle resources in the economy, and that proportion is certainly higher now. Thus, an immediate reduction in demand of around £40 million per year (a conservative annual cost for these layoffs) will have a total impact of say £80 million now, while in an economy running close to full capacity it might have an impact of only £20 million.

This does not necessarily mean that BMW should be prevented from closing factories (and note that they are not closing a factory, just eliminating a shift - so the capital reallocation effects are much smaller). If we try to freeze the economy in its current state by retaining the capacity of all firms, economic growth will probably be damaged even more, by missed opportunities for innovation and productivity growth. But we also don't want whole swathes of the economy to close down in a vicious cycle of reduced demand.

These cycles are largely a function of the friction that I described in my article of a few days ago. Because of frictional phenomena, adjustment takes time to happen. In a recession, demand is more sensitive to levels of unemployment, and equally unemployment is more sensitive to falling demand.

Naturally this presents a strong argument for fiscal stimulus, where the government helps to stabilise total demand so that the knock-on effects are less likely. It also suggests that, where possible, closures should be staggered so that they don't all happen at once. While there is obviously reduced demand for cars, there is certainly someone somewhere who will pay for a Chrysler (or a Mini) at some price, just not the prices that are being asked for them now. It's likely that in poorer or emerging markets, the marginal utility of one more car is higher than it is in the UK. Perhaps Indian or Zimbabwean families could be buying up Minis if we offered them at a reasonable price? Perhaps even unemployed people in the UK.

Setting the right price would be an interesting task - and allocating the cost of the subsidy between the government and BMW would be another one. But rather than pay all the adjustment costs and lose all the tax revenues from that £40 million, it would be a lot cheaper for the government to subsidise low-income people to buy the excess Minis for 18 months until the recession is over and it's safe to let the factory fail on its own merits.

In the meantime, shift workers who know they are likely to lose their jobs over the next 18 months will have a strong incentive to retrain for new industries, and this is definitely a role for government-subsidised adjustment assistance. If they finish training and find jobs before that, so much the better.

Danger: metaphor alert

In physics, there are two kinds of friction. Kinetic friction acts as a continuous force between objects that are moving relative to each other. It continually uses energy and generates heat, but in practice rarely affects the ability of the objects to keep moving, and therefore does not change the qualitative behaviour of the system.

The other kind, static friction, acts on objects which are not yet moving, and makes it harder for them to start moving. This kind does change the behaviour of the system. Static friction is usually stronger than kinetic friction, and in the economic analogy this is true too. A car engine is harder to start than to keep going once it's running.

An economic system which is moving at high speed can absorb friction much better than one which is stuttering or stagnant. Thus we need counter-cyclical adjustment assistance - encouraging excess capacity to close when the economy is doing well, so that it won't all happen at once when the brakes are thrown on. That, incidentally, is an argument for targeting higher levels of productive investment (relative to consumption) in a growing economy. But that's an argument for another time.

Sunday, 15 February 2009

A generally applicable skill of management

Just a pointer to an excellent article on management by Daniel Davies, which fits nicely into the vision of my think tank, Intellectual Business. Thanks to Brad Delong for the pointer.

Another article, this time on homo economicus and by Gavin Kennedy. I have started reading Kluge and though it is not broadly speaking a book about economics, it still has some useful things to say for behavioural economists.

Thursday, 12 February 2009

If we can't eliminate friction, we need a stimulus

Synthesising the whole stimulus debate into a few lines, there seem to be a small number of messages:
  1. Stimulus is good because it gets idle resources producing something
  2. Stimulus is bad because it moves productive resources into less productive use
  3. Stimulus is bad because its cost has to be paid back, reducing future efficiency
  4. Stimulus won't work because rational consumers will save as much as the government spends
  5. Stimulus is good/bad because government spending is efficient/inefficient
These messages are not necessarily contradictory - some of them are orthogonal. The truth of each assertion all depends on your model of how the economy works - and especially on one big factor: how much friction is there?

In an economy without friction, much of the argument would disappear. Idle resources would be immediately reallocated to some other use, since there is always somebody with savings that they could switch to consumption. Prices would adjust. Resources would always go to the most productive application.

Even the efficacy of the rescue of the banking system is substantially dependent on this friction question - if the government nationalises and shuts down a bank, how quickly will its assets (if any) and the capital of its depositors return to productive use? How quickly are savings redistributed as investment? In short, what is the velocity of money?

If all friction disappeared, our only debate about stimulus would be the argument about how efficient government spending is, which really comes down to (as Lynne Kiesling says) the knowledge problem versus the collective action problem. Even that question would get a lot easier with the help of the immediate feedback that a frictionless economy would provide.

So why is the economy so frictional? There are several reasons and none of them are going away soon.

One cause is the existence of contracts. Contracts are there to give people predictability - perhaps they don't trust in the efficiency of the market, so they want to provide some security of income. Normally a seller reduces its price (or an employee reduces potential wages) in return for this security. The buyer is generally happy to accept this and absorb some risk.

We can see a transition very clearly in some commodity markets which have moved over time from contract to spot trading. The gas market in the UK is one - once a liquid market was created, suppliers and buyers were willing to move from the contract structure and trust the spot market. This increases economic efficiency and thus both parties should benefit (at least in the long run).

However, many markets are not like this - including most employment markets and many markets for the provision of intermediate goods and services in business-to-business sectors. So, contracts create friction.

Another cause is anchoring. This is a psychological effect which, again, is likely to have evolved to increase predictability and also to reduce cognitive load. We learn that a particular object in the world has a particular value - for example bread or petrol may have a particular price - and we make certain decisions about it. In order to save time in future, we look at the price and remember our last decision. If the price changes, we need to make the decisions all over again - therefore prices are likely to stay the same.

As an aside, this is one reason for price discrimination between new and existing customers. Existing customers have already made the decision to buy at the old price. Therefore sellers want to get the benefit of the customer's memory of the decision they already made. Non-customers, on the other hand, if they have seen the price before, have made a decision not to buy at that price. Therefore you need to get them to make their choice anew - and thus, give them a new price so they have to reconsider.

Related to anchoring is loss aversion or a closely related phenomenon, the endowment effect. This means we overvalue what we own in comparison to what we don't; making us less likely to trade. We may be unwilling to sell a house which has fallen in value, even though the new house we want to buy is also cheaper. It means we may not leave a job that we're unhappy with, because we have an attachment to it which is not, strictly speaking, rational.

A fourth cause is information lags. It takes time for information to disseminate around a population. If the price of a product is reduced or a new service becomes available, it won't immediately have an effect because the people who might buy it won't know about it right away. Technologies like the Internet naturally act to combat this, but they are far from perfect.

A special case of the above is negotiation time. When products (and particularly services) do not have a published price, it often takes time for two parties to reach an agreement. This also occurs when a service is not of a standardised kind, in which case the nature of the service has to be negotiated as well as the price.

The final cause, and the one most handled in the economics literature, is the generic category of transaction costs. Some of these are the physical cost of consuming something - for example the effort of moving house, or the delivery cost of a book from Amazon. Others include taxation (stamp duty on shares or houses), costs of telecommunications or postage, or time spent queueing to buy something.

So next time someone says a stimulus isn't necessary, or that the market will sort itself out, or that we should just close down all the insolvent banks and let new ones start up, ask them how they plan to eliminate friction from the economy.

If they don't know, then they better be ready for a three-year depression while people and companies form a whole network of new relationships, figure out the right prices for services and products, the appropriate savings rate for them, and how much to trust each other's and the government's promises. This is no small undertaking and I really don't want to waste three years of my life on it. Do you?

Do Click: twitter vulnerability

If you are at this page, you probably clicked on a "Do Click" link on twitter.

Well done - you know how to follow instructions. If you were one of those rebellious types, you might have clicked on the "Don't Click" link which is also propagating on there.

Fortunately you aren't like that. If you were, you would probably have been the victim of a cross-site scripting vulnerability (explanation at Coding Horror). This runs some Javascript which starts posting messages in your name on twitter.

In this case the message seems relatively harmless - it simply posts another copy of itself to try to spread itself to other twitter users. But, not having clicked on it, I can't say for sure that it is innocuous. In general it's dangerous to click on this kind of thing.

One way around this is to insist on using the 'preview' feature on tinyurl. Any tinyurl link (such as the one that's being passed around with the "Don't Click" message) can be modified by putting "preview." in front of it. For example, http://preview.tinyurl.com/secretstim/ instead of http://tinyurl.com/secretstim/

If you use this, then you will see the URL of the site you're being forwarded to, giving you a chance to decide whether to click through. You then get to use your judgement in deciding whether to click or not.

In the case of the "Don't Click" message, you would have seen a URL ending in .php within the directory http://www.umoor.eu/blog/

I went to the above site manually, omitting the PHP link - and decided it didn't look very trustworthy. Therefore I didn't click on the link.

Plus, common sense should warn you off when someone sends you the message "Don't Click". Most of us have learned not to click links like that in our emails.

Didn't anyone tell you at school that "gullible" isn't in the dictionary? xkcd warned us about this just two days ago.

Here is a page discussing an older version of this problem.

Bounded cognition and the stimulus bill

According to the Sunlight Foundation, nobody (outside of the ten-person committee that negotiated it) has yet been allowed to see the stimulus bill which the House and Senate have agreed. This includes the members of the House and Senate who are supposed to vote on it today!

I know that I've argued that cognitive limits and imperfect information act as a constraint on economic efficiency; and no doubt also on political efficiency. But that doesn't seem like a good reason not to publish the bill. Maybe they have slipped in that clause on a stimulus for bloggers and they want to get it passed before the twitter crowd start asking for their own version.

Utility versus income

Greg Mankiw asks us to spare a thought for the top ten percent of earners who apparently bear the main burden of the economic downturn.

I wish he had figured out how to measure whose utility has been hurt the most, as opposed to just whose consumption has fallen. Unfortunately this is a blind spot of many economists. Poor people simply get more value out of a dollar than rich people. This also means that they are hurt more by a smaller fall in income.

Diminishing marginal returns have been part of economic theory for two hundred years, so I'm not sure why people keep forgetting it.

Illiquidity measures

Continuing the discussion of a few days ago, Nick Rowe joins in correcting Niall Ferguson's "average debt" logic. Maybe we should give the guy a break - he is only a historian, not an economist. No doubt whenever a historian writes a book about the history of [X], they get brickbats from all the practitioners of [X]. But then that's the price of intellectual engagement.

The comments on Nick's post set me thinking about better ways to measure the total impact of debt in the economy. One commenter (Patrick) reminded me of the issue of debt maturity mismatch, which seems to have been a major contributor to of the current financial crisis (aside from the effect on the real economy). Here's the response I posted there:


It sounds attractive that there would be some meaningful measure that at least partly captures the way in which debt influences the economy. We could certainly build models where a large amount of gross debt has almost no effect at all (e.g. A owes B, B owes C and C owes A an identical fixed amount with identical fixed repayments, and each debt secured on the next), so it can't simply be the quantity of gross debt.

To adjust Declan's example, the 10% interest rate should be there to compensate for the fact that one or two of Harry's friends will default on their loan; this reduces the accumulation of money in FI. But without quibbling on the details, there is definitely less "stability" in that system than in one with no debt at all.

Patrick pointed out the issue of liquidity, or more generally the length of maturity of debt; although by the same logic as the original argument, the net amount of debt at any given maturity is also zero.

One could generate a measure reflecting the amount of short-term liabilities held by entities who have lent long. This would be appropriately asymmetrical, because we don't care about those who have borrowed long and lent short (except that their short borrowers may be unable to repay! but that is captured in the original measure anyway).

More precisely, we could draw a maturity graph for each entity, showing the debt they are due to repay in each period net of the assets they are due to collect (including their money holdings). This illiquidity measure (IM) at any given time t would be the integral of that curve between 0 and t.

The appropriate aggregate measure would be the sum of all positive IMs, ignoring negative ones, for a given period. So you could generate an one-year IM for the whole economy, or a one-month IM or a ten-year IM.

What does this tell us? I don't know exactly, but a lot more than "average debt". I imagine it could offer guidance for the appropriate size of central bank liquidity measures such as the Bank of England's SLS and the Fed's money market purchases.

Update: Enrico Perotti and Javier Suarez have a proposal for mandatory liquidity insurance at VoxEU.

Wednesday, 11 February 2009

Price of a Senate seat, revisited

In December I calculated the price of a US Senate seat at conservatively $1 billion.

I'll disregard for now the $2.5 trillion secondary effect of the new Tim Geithner bank rescue plan, and whether we might therefore revise the value of the marginal seat upwards to $5-10 billion.

But Brad DeLong has worked it out in a different way. The price of three Republican Senators is 600,000 jobs. So let's crunch the numbers.

At 200,000 jobs per senator, and one senator per billion dollars, it's easy to work out the price per job: $5,000. This compares extremely well to the $280,000 per job estimated by Greg Mankiw.

Even on the enhanced valuation of $5-10 billion per senator, we are still only looking at $25-50,000 per job. Perhaps the stimulus plan can be restructured to focus on buying the seats of Republican senators and converting them into jobs. Obama's goal of 4 million jobs could be achieved for as little as $20 billion, by simply buying up twenty of the remaining Republican senators.

We leave open the question of whether Democratic senators are worth more or less than Republicans.

p.s. server logs show several visits to the above page from both the US Senate and the House of Representatives. My recent posting on a stimulus for bloggers also received visits from the House, while the Senate clicked on my analysis of the Buy American clause. Perhaps policymakers are taking these important blog postings seriously. At last, some sanity prevails in the halls of government.

Tuesday, 10 February 2009

Nationalisation and the knowledge problem

In pure asset-allocation terms, nationalisation of bust banks seems to be the most reasonable option. Shareholders have taken big risks (or allowed their appointed representatives to do so), the risks did not pay off, and government now has to step in and help. Therefore, government should now own the banks.

I have broadly supported this argument, but it does hide a big problem: the knowledge problem. (There's a whole blog about this subject)

In order to best allocate economic resources, those who control them need to have knowledge about where they can get the best return. This knowledge is distributed all over a population of billions of people. Therefore how can we ensure the information can most quickly and reliably find its way to the person who controls the resource?

First we need to note that 'knowledge' is a tricky concept. At the risk of stepping through a vast minefield of epistemological thought, I'll assert that nobody knows anything about the world with absolute certainty. There are varying levels of confidence, conflicting views, self-delusion and simple gaps and errors in the information and beliefs that people have.

The market is a good (though hardly perfect) way of testing beliefs, finding which ones work and incentivising those with better information to apply it. And this is why Tim Geithner is so keen to ensure that the private sector has a stake in the toxic assets that the government is now acquiring. Market involvement helps to find an accurate price for debt, stakes in banks, and other assets - and this accurate price is the market's way of revealing all available information.

Government has other ways of gathering knowledge, and they are useful for certain things, but most people would agree that they are generally less effective than the market. If government ends up choosing the loan policy or marketing strategy of banks, it's likely that the resources will not go to the most productive places. I am well aware of the counter-arguments: a) bank management hasn't done a very good job in the past, and b) there are public goods and externalities which government will achieve better than private, competing banks. Nevertheless, the problem still remains. A future article will have more to say about this.

The hybrid public-private structure of the proposal not only saves money for the government, but achieves a reasonably equitable asset allocation while still retaining the informational advantages of market involvement.

Undoubtedly the solution is not perfect - but in a situation of limited meta-knowledge - knowledge about knowledge, and knowledge about the best solutions for the problems we face - it sounds like a reasonable proposal.

Monday, 9 February 2009

A bad Barclays

According to Robert Peston, Barclays' share price appears not to reflect the underlying state of the bank. It raised £10bn in new (private, not public) capital last year and made £6bn in pre-tax profits (between £4bn and £5bn post-tax). Thus, in theory, its balance sheet should contain around £15bn of new capital since this time last year. Not to mention whatever was on there before.

And yet, its shares are valued at under £9bn. Presumably this is because of uncertainty about the value of some of its assets, and the fear that they will be written down - eliminating the capital base of the bank.

So why doesn't Barclays take - on its own initiative - the advice many economists have given to the US government, and create its own "bad bank"? It could simply split itself into two or more pieces, with the better half of the assets in Bank A and the riskier half in Bank B; and provide the minimal required capitalisation to bank B. B would carry off a big chunk of capital but, on the basis of the accounts, by no means all of it.

Bank A, "good Barclays", would be left as a strong and low-risk institution whose market value would surely exceed £9bn; Bank B, "bad Barclays" would be much riskier, but even if its shares became completely valueless the shareholders would be better off.

This might not be "playing fair" - it would risk shoving the liabilities of Bad Barclays onto the government - but it would surely be in the interests of the shareholders, who are so far still in control of the bank. And, arguably, in the interests of the economy by providing at least one large, secure, well-capitalised bank capable of providing credit to willing borrowers.

Sunday, 8 February 2009

Behavioural models of pay, and the agency problem

One of the primary criteria in designing any system of pay is to influence the recipients' behaviour.

Traditional pay structures work on the assumption that people behave rationally. Thus, they will pay people for generating profits for their company; on the basis that if you pay them more for higher profits, they will generate more profits. This has the attraction of simplifying the manager's job - assume that the employee can figure out the best way to make profits, and leave it to them.

However, this structure has at least three intrinsic problems. One is asymmetry: you can pay someone for making profits, but you can't penalise them for making losses. Virtually no employee will work for a company if there are circumstances in which they would have to pay for the privilege.

The other is the agency problem: the employee has control over the company's resources but does not get all of the benefits of using them in the most productive way. Thus, a temptation exists for them to use these resources for their own private benefit.

The final problem is that it is based on a false assumption. People do not behave in the traditional "rational" utility-maximising way.

If, instead, we apply some of the results of behavioural economics, we may be able to design a pay structure in which the same behaviour is in the interest of both firm and employee.

Some of the most important results of behavioural studies include:
  • People are more likely to act in their short-term interest than their long-term interest. Many different behavioural experiments have demonstrated this. We can reflect this in pay and management structures which provide immediate reward for doing things that are unambiguously in the interest of the company. This is challenging, because it isn't always obvious in the short term what the company's interests are. But if we have a way to, for example, measure the amount of risk that is being taken in an investment portfolio, and give bankers an immediate reward when that risk is reduced, this may help to override their longer-term tendency to manipulate the system in their own interests.
  • People are loss averse. This means that it has more impact on someone to lose £1000 of money that they already have, than to gain £1000 that they don't have. To take advantage of this, a clawback provision would probably be effective. As suggested in a number of articles recently, bankers could be obliged to return some past bonuses if a company goes on to make losses on positions they took in previous years.
  • People are motivated by social factors. This result is informed more by traditional psychology than by behavioural economics. But a number of social factors - status in a group and adoption of social norms - compete with individual self-interest to influence our behaviour. Companies and their ecology provide two of the strongest social groups for many employees - their colleagues and competitors. By creating measures of status within those groups, and linking those to social norms related to the long-term stability and growth of the company, managers could create influences that might dominate financial incentives.

    Money, in some ways, is just another social status marker anyway. Anecdotally it seems likely that people work hard to increase their bonus not mainly because of the associated material benefits but because it governs their place in the hierarchy of their social group.
  • People are, under some circumstances, risk-seeking. Traditional economic theory says that people will give up some rewards in order to avoid risky situations. Recent behavioural research says that, sometimes, the opposite holds. Thus, particularly in situations where the risk of something happening is very small (perhaps less than 0.1%) people will sometimes make decisions that increase this risk instead of reducing it. Clearly that can cause major problems when they are managing large amounts of money - just look at Bernie Madoff, Jerome Kerviel or Nick Leeson to see the results. So if a third party can monitor traders' behaviour and take a dispassionate view of the risks, they could help to provide a rational discipline which helps reduce them.

    While traders would at first no doubt resist having someone monitoring all the choices they make, it is possible to demonstrate that this is in their own interest. In software development we sometimes use a technique called pair programming which has two programmers working at the same keyboard writing the same piece of code. While most programmers find this distracting and wasteful at first, the resulting improvement in quality and reliability - and often productivity - becomes clear in many situations, and most programmers who regularly work this way are won over.
Lots of modern management theory explores how to motivate people in a non-financial way. However few of the books and articles I've seen base this discussion in real behavioural research; they tend to be based on heuristics and informal experience. While experience is clearly valuable in making management decisions, behavioural economics provides an opportunity to supplement and enhance it with knowledge derived from scientific research.

Computers more ironic than humans

Visiting Daniel Little's Understanding Society blog I noticed the following automatically generated news headline in the right-hand column:
Computers cleverer than humans will create even greater inequality
Financial Times, UK
I am more worried that artificial intelligence will greatly increase the inequality of wealth and power among humans. You have viewed your allowance of free ...
Not quite so clever yet...

Saturday, 7 February 2009

How much debt is too much?

Another article (this time by Niall Ferguson, of whom more later) on the too-much-debt theory. The subtitle is: "Governments cling to the delusion that a crisis of excess debt can be solved by creating more debt".

Well, this isn't a crisis of excess debt. To the extent that credit problems are responsible for the recession, it is a reduction in the availability of credit that has triggered it.

Ferguson may think that companies and individuals owe too much. But who do they owe it to? Er, other companies and individuals!

He doesn't present much evidence for the "too much" theory, except an assertion:
The Western world is suffering a crisis of excessive indebtedness. Governments, corporations and households are groaning under unprecedented debt burdens. Average household debt has reached 141% of disposable income in the United States and 177% in Britain. Worst of all are the banks. Some of the best-known names in American and European finance have liabilities 40, 60 or even 100 times the amount of their capital.
Ferguson obviously thinks these numbers sound high. But who knows? What is the correct number? What should our target amount of debt be? Government debt in the past has reached 200% and 300% of GDP (after major wars) and been paid off. So has private debt. Why shouldn't bank liabilities be 100 times their capital - after all, their assets are 100 times too.

While building up my company, I've certainly had much more debt as a percentage of disposable income than the numbers Ferguson cites. I don't any more. If I hadn't taken on those loans, I wouldn't have the business now. The loans were an essential part of coordinating current and future activities in order to create a productive business. And my ratio of liabilities (or assets) to capital? At its peak, maybe 2000 or so. It just doesn't take that much capital to run a software business.

This attitude arises from forgetting the difference between me and us. It's a useful aspiration for individual people, as they go through life, to build up financial assets. Not perhaps as useful as it is to build up income, but still, the two numbers are partially correlated. Few 20-year-olds have much in the way of investments to draw on, but the average 55-year-old (in the developed world) has a reasonable sum. This is usually what enables us to retire around 60 (and withdraw our productive skills from uses that can make people better off - but that's a different argument).

But this does not mean that the whole world can build up financial assets. Every asset I have (that is, each pound I have invested or lent) is a liability of someone else. It's only someone else's debt that enables my credit.

Does it feel tragic to realise that, for all the saving, investing, profits and dividends that people and companies have built up over the years, the whole world's financial asset base has a value of: zero! It shouldn't. These assets are just a promise for one person to pay money (or resources) to another person in the future. Every promise has two sides, the giver and receiver of the promise; and thus every debt has an asset.

So these figures - 141%, 177%, 500%, whatever - are simply measures of how many promises people in the world have made to each other. Nobody has a useful theory (that I've seen) of the number or strength of promises that people should make to each other; Ferguson does not propound one in that article, except the implied idea that "if I had that much debt I wouldn't feel comfortable". Well, that's a statement about someone's feelings, not economic reality.

Incidentally, Ferguson does have a few reasonable policy proposals in the article. But they come from a flawed premise, and thus the chain of reasoning seems likely to be incorrect too. Funny how our instincts can keep us on track in building an argument, pulling us back from the obviously-wrong conclusions that our premises may logically imply. But this is just a newspaper article; no doubt in an academic paper one would argue more clearly from premise to result.

Back to the recession. Each reduction in a debt is also a reduction in someone else's asset. And if they both shrink faster than the holders desire, both will be unhappy; and their ability to carry on the productive work that they were previously doing, will be reduced.

This withdrawal of credit and the dislocations that result have led to people and resources sitting idle, and output falling. The reason governments are trying to boost the debt in the economy is to counter this fall. Excessive debt did not cause this crisis - the problem with debt is that we can't get enough of it any more.

Update: Brad DeLong comes to similar conclusions about Niall Ferguson's reasoning, though for slightly different reasons.