Friday, 9 May 2008

Dump the CPI

Does the consumer price index adequately capture inflation? If you try to tell anyone in a supermarket today that inflation is growing at just 2.5 percent a year, the likely reaction is cynicism verging on outright disbelief.

The CPI is just an index, constructed from a large number of sub indices, each capturing individual price changes. The weights used to construct the series, coupled with what sub indices are included, have enormous influence on the final rate of inflation.

This point is amply illustrated when we compare the CPI with a much older measure of inflation - the retail price index. Since 1997, the RPI has consistently grown faster than the CPI.

Today, the divergence is around 13 percentage points. In practical terms, this means if we were to compare two individuals - one whose shopping basket was accurately reflected by the RPI and another by the CPI - then the RPI consumer would be paying 13 percent more for their basket compared to their CPI neighbour.

Housing cost is the major difference between the two indices. The RPI includes housing, while the CPI excludes it. What happens if we take housing costs out of the RPI? The divergence goes down, but the RPI is still considerably higher than the CPI. Over the last 10 years, the RPI has cumulated a four percent difference over the CPI.

Of course, removing housing costs from the basket is total nonsense. People need to live somewhere, and it costs something. If housing costs are rising, then other things being equal, people are getting poorer. Any meaningful inflation measure would properly reflect that development.

The Bank of England is currently using the CPI as their inflation target. However, the exclusion of housing costs has allowed it to portray the level of inflation as being lower than most people actually experience.

The Bank of England has also placed heavy emphasis on its core inflation measure, which excludes food and fuel. This measure is utterly meaningless. No one lives without food, fuel, or housing. So the Bank of England's perceptions of inflation are disconnected from the day to day realities of inflation. This has lulled the Bank into keeping rates low, and fueling inflation and the housing bubble.

It's about time that this foolishness was stopped. The Bank of England need to do two things. First, it must revise its inflation target. In future, it should target the retail price index and not the CPI. Second, it should be banned from ever mentioning core inflation again.

These two simple changes would go a long way to preventing a repetition of the housing bubble and maintaining price stability. Higher mortgage costs will be reflected in the RPI, and when credit begins to rise, the Bank of England would be obliged to raise rates. In the short run, as the bank raised rates, this would generate a slight peak in the RPI, as higher rates raised mortgage payments. However, over the medium term it would ensure greater price stability and help avoid dangerous asset price bubbles.

It will be a happy day when the Bank of England dumps the CPI. People might again believe in official inflation numbers, and the MPC might again begin to get serious about bearing down on inflation.


Josh said...

Yes, the CPI just doesn't capture what is going on in supermarkets; bread, milk, petrol, are all rising faster than wages.

Anonymous said...

Good post though as you get to the conclusion I start to disagree. You can probably guess why.

Yes the BoE should dump the CPI. They should replace it was an accurate measure of money supply. Increases in prices only have two inputs: (i) money supply increases and (ii) demand growth relative to supply.

The BoE has no influence whatsoever over real supply/demand factors. They can't tell the Saudis to increase oil production any more than they can tell the Indians to drive less cars.

Therefore the only thing they can control is how many pounds they can print. Target that.

IMHO management 101 says only assess people against factors they can control.


aSteve said...

CPI is a valid metric for inflation as it relates to goods traded between countries in Europe. It is pretty awful in all other contexts.

Unfortunately, you can't (realistically) demand that the BoE targets RPI when setting interest rates - because the policy would be self-defeating.

If RPI went up, say (and assume inflation going up is our primary worry) then rising rates would push RPI higher (on account of immediately higher mortgage costs) - and if interest rates were dropped in the context of rising RPI inflation, the immediate effect would be to reduce RPI. This, of course, is dangerous because lowering rates does not reduce inflation in the medium term. Similar arguments arise if your primary concern is deflation... which, economically, is often seen as even more undesirable than inflation.

If RPI were to be adopted as the inflation target for rate-setting... then an inherently unstable system would emerge... and disaster is all but guaranteed.

What we need to do is to take house prices into account - not the debt that some choose to accept in order to buy them. This would allow the cost of debt to be continue to be used to regulate the economy.

I don't think that CPI should be updated to include house prices, I think house prices should be considered a separate index. I think the BoE should target not only CPI but HPI too. An additional HPI target of 2% (+/- 2% say) would be reasonable, in my opinion... after the crash, of course.

aSteve said...

Nick, I think there is a massive flaw in your proposal.

How can we accurately measure the money supply? What is included and what isn't? If, for example, you recommend targeting M4, I've a simple way to subvert the system and get very rich. I securitise debt and establish a liquid trade in the resulting bonds... If the bonds are immediately convertible to cash - so they'd be accepted in payment... but they wouldn't affect M4. This would undermine the national currency, of course, and would make taxation all but impossible... but it would present an opportunity to make vast profits.

I see the question: "What exactly is money" as being fundamental. I can't think of any definition that would result in a system that would not be highly susceptible to exploitation. For example, is your pension "money"? What about trust funds for children? Are your stock market investments "money"? What about futures contracts for corn or oil? What about life insurance? What about bottles of single malt scotch, fine-wines or cannabis? Where do you draw the line (no pun intended)?

Anonymous said...

money, money, money,
its so funny,
in a rich man's world....

powerman said...

The money supply of a currency is the supply of all credit created by the central bank and fractional reserve banks. Issuing your own bonds doesn't create credit because the bonds have to be bought by somebody using the Central bank credit.

aSteve said...

Powerman, that's a valid definition of money supply in a general sense... but I do not think it sufficient in the context of "Austrian Economics" - where the money supply expansion is taken to be the definition of inflation. In that context a wider view must be taken, IMHO, since it is possible - by increasing the velocity of money - to use alternate assets as proxies for money. I'd argue that in order to use money supply as an effective way to control interest rates - you need to consider not only M4 (as you describe) but every liquid asset... which is no simple task.

powerman said...

I see your point Steve, but I still think central bank credit is a useful measure for the following reasons:-

All other liquid assets only have a currency denominated value in relation to the central bank credit which counts as legal tender.

Stocks, bonds and commodities, whilst liquid are only marketable as money at some price which depends on many factors (most of which are beyond a central banks direct control, as the original poster noted). But in part the price depends on the total supply of the currency in which the stock price is tendered.

ii) Central banks and fractional reserve banks effectively have a licence to counterfeit. Central banks are only constrained in this activity by political considerations and the tacit trust that their monetary committees are sufficiently wise to fix the price of borrowing money.

As we can see in the most recent credit-based bubble, these political constraints can work counter to good sense. Even the wisest and most altruistic of mandarins with degrees from the finest schools are still subject to Hayek's calculation problem.

I personally think Mervyn King has done his best to resist an awful lot of pressure throughout this. I don't hold him much to blame considering the targets and inflation models Gordon Brown gave him. However, I know more trust him to be able to calculate the 'correct' price of borrowing money, than I would trust any central planning comittee to determine the correct price of bread, vodka or tractor tyres.

Total Sterling-denominated credit in circulation is under the direct control of the BoE and UK top-tier fractional reserve banks, so that's the figure I'd like to see measured.

Accidentally factoring things totally beyond their control such as the effects of oil pipelines being blown up, or bumper wheat harvests by trying to measure inflation indirectly via the price of commodities makes the index in question a pretty murky instrument for trying to determine monetary police IMHO.

aSteve said...

I agree that the monetary base you describe is extremely relevant... but I remain convinced that it is far from the be-all and end all... and that using it alone as a target for the purpose of setting interest rates would be disastrous.

I share your views about Mervyn King - and, I strongly suspect, he would also argue that the targets he has been set are asinine - if professionalism didn't constrain his tongue.

One thing I've never properly fathomed is why, where the markets have worked so well in other fields, we don't let the market set the rate for borrowing. At the level of bonds this seems to be efficient enough - and, while fractional reserve banking throws a spanner into the works, it doesn't seem entirely unreasonable to set a capital adequacy level; enforce it and give up trying to manipulate prices.

Anonymous said...


The BoE has no business whatsoever to "set" interest rates. That is effectively soviet central planning by fixing the most important price in the economy - the price of money. I find it astonishing people would allow interest rate fixing in an allegedly free amrket.

I'm with Powerman on money supply being measurable. M4 is too inclusive in my opinion. Keep it to cash and fractional reserve-created credit.

Best of all worlds is just go on to a gold standard, replace the BoE with a currency board who's sole job is to expand and contract the money supply depending on it being presented with gold or pounds, in order to keep the exchange rate even to gold.

The end result should always be a stable currency. The most stable currency is a gold standard. Implement that and all the other problems disappear.

BUT it prevents governments printing their way out of trouble, so they'll not go along with it.


aSteve said...

Well, I have to disagree - I am convinced that M4 is too narrow a measure of money for an Austrian approach to interest rates. I am, in principle, however, in favour of a interest rates set by the market. I am utterly opposed to the concept of a gold standard which, thankfully, I anticipate will be interpreted as being the crackpot idea it is. A monetary system based on gold allowed a monopoly on wealth and impoverished civilisation for centuries. Fiat has its problems, but represents an infinitely superior proposition.

aSteve said...

Well, I have to disagree - I am convinced that M4 is too narrow a measure of money for an Austrian approach to interest rates. I am, in principle, however, in favour of a interest rates set by the market. I am utterly opposed to the concept of a gold standard which, thankfully, I anticipate will be interpreted as being the crackpot idea it is. A monetary system based on gold allowed a monopoly on wealth and impoverished civilisation for centuries. Fiat has its problems, but represents an infinitely superior proposition.

powerman said...

I'm not sure about a gold standard either, but I do in principle agree with the Austrian idea that fiat money which can be created by certain private organisations arbitrarily (except within political constraints) is somewhat dangerous.

There's nothing magic about gold, it's just a convenient non-perishable commodity one can use as a money base to prevent arbitrary creation of credit.

I disagree that the gold standard impoverished our civilisation though.
Our economy saw substantial, prolonged economic growth under the gold standard. What it did do was act as a firm break on how much money our government could arbitrarily create, and we only came off it to break this credit/inflation limit so we could fund WWI.

steves said...

The real problem with the gold standard is the value of the gold itself -

As Friedman pointed out when the gold rush started in the states the value of gold plummeted - as with any market (look at Gordon Browns ridiculous example of announcing he was going to dump the Uk's gold reserve onto the market)

If something else replaces gold in the eyes of the world this will reduce it's value.

While I will concede that the gold standard is preferable to the current system, I would prefer a truley private solution, let people use whatever currency they wanted, let the good money drive out the bad, if the BOE still want us to use their £ sterling this will make them look after it better.

Anonymous said...

Gold has all the characteristics needed for a stable currency:

- It can't be printed;
- It's supply increases by only about 2% a year;
- It's fungible;
- It doesn't disappear;
- It's as useful in one country as any other;
- It can be moved from a country with impending political risk to a safer one.

It's also stood the test of time for 5,000 years and provided long term interest rates of about 3%. There's not a single fiat currency in history that comes close to matching gold. Look at how badly the dollar has done since Nixon tored up Bretton Woods.

Calling a gold standard crackpot exposures a failure to evaluate history and a naivete towards accepting the latest hits as the greatest hits. The problems we are having now are the same problems fiat currencies have always had in the periods when countries came off the gold standard.

In fact the US's finanicial problems right now are now any different in kind from Rome under Nero.

I like the discussions here, and the technical detail is interesting, but let's not lose historical perspective. Economic principles haven't changed in thousands of years.


Anonymous said...

I would even go further and argue that price index is logically invalid concept.

There is a long argument by against price index at How much can we trust the price indices (e.g. CPI)?- basically, all price indices are pseudo scientific. They should all be dumped.