The Piketty fallacy (2)
One aspect of the wealth debate seems to me less than fully explained by conventional economics. It is perhaps most obvious if you spend your Sunday afternoons visiting stately homes. The National Trust seems to have an unlimited supply, and then there’s English Heritage, others who belong to the Historic Houses Association – not to mention the ones you never get into because they remain private. And while they may not all be as large as Blenheim or Chatsworth, they must each have cost a good deal to build (and built they were, most in the eighteenth century when the supporting population base was far smaller, many with the ill-gotten gains from warfare, sugar or slavery). But more to the point, with the infrastructure of servants they require they would not be built today. Only the tiniest handful of Russian oligarchs could afford to attempt it now.
While I haven’t spent years amassing the statistics (even if they were available, and of course they’d then be challenged by those who disagree with my thesis, so it hardly seems worth the trouble), my guess is that the rate of inflation of stately home building costs is far higher than that of the regular basket of goods and services that get measured by RPI. That is probably true also for the accoutrements of noble living: servants’ labour costs, old master paintings, top of the range “toys” (useless luxury objects) etc. Anecdotally, and over a shorter period, you can probably confirm this: opera tickets, medical bills and short run scholarly books have gone up far more than “average” goods, certainly in my lifetime. But economics, in its quest to reduce everything to a single parameter, ignores these differences, and serves up only one RPI for every citizen. (Of course if Chris Giles has his way, RPI would be abolished in favour of CPI, but that’s a different story.)
Now think about this. Suppose A has an index-linked pension of £25,000 p.a. B on the other hand just has a cash sum of £500,000. Say this is so when actuarial rates make these equivalent. Now suppose interest rates fall, without inflation being changed (is this possible? well the Bank of England think they can do it, because this is exactly what QE was all about). Which of them is affected? Common sense tells you that nothing has happened to A: each month he gets his pension and spends it, so changes in interest rates are irrelevant. B however is able to obtain lower returns on his cash, and is clearly worse off.
But economists tell you something different. B’s wealth is unchanged: he is still “worth” £500,000. A however has a stream of income which now needs to be evaluated by discounting each payment by a lower interest rate than before, so its net present value (or “NPV”) has gone up – and he is richer!
You can try to wriggle out of it in various ways. One, particularly relevant when it comes to annuities, is that economics generally assumes that you can borrow or deposit at exactly the same rate, whereas in practice there is a “bid–offer spread” which prevents A and B swapping positions; when annuity providers offer annuities at multiples of 20 times, but cancel them at 37 times, you can’t really do this. But the paradox remains even when that is removed.
The real point is that, even if the NPV of A’s pension has gone up, so to has the present value of the consumption of goods and services on which he spends his money (the value of the items doesn’t change, but the discount rate does) – and overall he’s no better off. B on the other hand may still be “worth” £500,000, but his future consumption is now more than that, so he’s worse off.
Accountants usually ignore this because they think A and B are free to do what they like with the money. But these individuals are small people, not grands seigneurs. They spend exactly what they have. Rich people don’t: that much Piketty (and many other economists) grasp fully. But the differential effect of rich-peoples’ RPI on their surpluses remains I think underanalysed.
The simple r > g calculus is inadequate precisely because luxury goods do not obey standard economic behaviour. People pay nonsense sums for a Jeff Koons piece simply because the number is silly. Moreover wealth ceases to be additive in any normal sense: to be twice as rich as someone else in a functional sense, you probably need to have 10 times as much accounting net worth.
Clearly these are very inchoate ideas, but it seems to me that the real challenge to Piketty’s conclusion that the rich are getting wealthier is that we should not be looking at this on an arithmetic scale, but on a logarithmic one since that is how superwealth behaves. But that brings with it its own refutation: an unequal division of the cake results in shrinking it. You can see that already with bonuses in banks. Give 100 people £10,000 each, and most will be happy. Give £1m to one instead, and not only will you have 99 unhappy people, but I doubt if the favoured one will be even ten times happier than each of the 99 would have been.