Sunday, September 03, 2006
Bank's credibility gap over 'true' measure of inflation
Posted by David Smith at 11:00 AM
Category: David Smith's other articles

We are, astonishingly, well into the 10th year of Bank of England independence. It has not exactly flown by but it does not seem that long since we were summoned to the Treasury a few days after the May 1997 general election, to hear Gordon Brown break the news.

Since then the Bank, and its monetary policy committee (MPC) has had its ups and downs. Mostly, though, it has been up, this being the most successful period for UK monetary policy in living memory.

True, independence came at a good time. Ken Rogoff, former IMF chief economist, told central bankers at their annual symposium at Jackson Hole, Wyoming last month that they have been lucky to have benefited so much from the anti-inflationary impact of globalisation; the China effect. But the MPC has also made its own luck.

Now, however, the Bank faces a problem, arguably its most serious yet. There are doubts, not just about its ability to control inflation in future, but whether it is under control now. Those doubts stem from scepticism about whether the consumer prices index (CPI), the Bank’s target inflation measure, accurately reflects people’s genuine experiences.

This is a serious matter, as Bank insiders are aware. If people stop believing the inflation numbers, the credibility of monetary policy suffers badly. That could begin the slippery slope to economic anarchy.

What is the CPI, and is inflation now 2.4% as it says? The CPI became the Bank’s target in 2003, Brown’s sop to Tony Blair when the Treasury rejected euro entry. The new index, constructed on the same basis throughout Europe, was his way of showing willing. Crucially, it does not include house prices or council tax.

At the time I wrote that the new measure would be seen as misleading and changing was a mistake. Mervyn King, the Bank governor, was also unhappy. “When defending a free kick from David Beckham, you don’t expect somebody to move the goalposts,” he said. But the change happened.

Had it not, I doubt we would have had this outbreak of scepticism. The retail prices index (RPI), one of our oldest official statistics, has inflation at 3.3%. The RPI excluding mortgage interest payments, the Bank’s old target, puts inflation at 3.1%. Nobody would have had serious cause to challenge them.

Instead, the switch to the CPI, and the highly visible rise in energy prices, has led to silly suggestions that “true” inflation is nearer 10%. It has also produced accusations that the CPI is really a “chav” price index - measuring inflation for those who buy (imported) branded sportswear and electronic products - leaving the middle-classes to bear the brunt of inflation.

In fact, the evidence is that people who suffer most from inflation driven by higher energy and food prices are those on lower incomes, particularly pensioners, because these items make up a higher proportion of their spending. Pensioner inflation, on reasonable assumptions, is nearly 4%.

Some middle-class families face big increases in independent school fees but that is nothing new - they have increased by 8% annually for the past five years. University tuition fees, since their introduction eight years ago, have risen roughly 2.5% a year, so have not contributed significantly to inflation. The new £3,000 a year fees will add to CPI inflation when introduced this autumn, though their real impact comes later; top-up fees are not payable upfront.

What is the real rate of inflation? “Core” inflation, excluding energy and seasonal food from the CPI, is just 1.1%. For many people, however, the things excluded from core inflation - life’s necessities - give the true measure. Charlie Bean, the Bank’s chief economist, recently rejected core inflation as of any real use.

Using figures from the RPI, I constructed a “necessities” index consisting of housing, domestic fuel bills, motoring expenditure, fares and food - nearly 60% of all spending. The result was a “necessities” inflation rate of 4.8%. The biggest contributor is energy, up by 29% in the past 12 months.

But inflation measures are constructed as they are for good reason, reflecting our spending patterns. Is it reasonable to exclude clothing and footwear, falling 4% a year? No family with children could do so. We have to buy some furniture, furnishings and durables, if only for replacement purposes.

Contrary to popular myth, consumer electronic products are not prominent in the CPI, and have a weight of just 27 parts in 1,000. I still have problems with the index but accept that either of the RPI measures gives us something like “true” inflation - in other words between 3% and 3.5%.

If you really want to know why true inflation isn’t 5%, 6% or 10%, look at spending. The CBI reported last week that August was the best month for retailers since December 2004. There is a caveat - the survey straddled the base-rate rise - but there was undoubted strength. That simply would not be happening if prices were rising faster, or as fast, as earnings, currently increasing by just over 4%. We would be in the deepest of consumer recessions.

The Bank’s real problem comes from sharply rising energy prices. It has had to deal with food price inflation before; during 2001 it hit 7%. It has not had to deal with a surge in energy prices or the folk memory that such surges have in the past been associated with very high inflation.

What should the MPC do to convince people it has inflation under control? The “shadow” MPC, which meets under the auspices of the Institute of Economic Affairs, anticipated last month’s rate hike from 4.5% to 4.75%. Three members, Tim Congdon, Andrew Lilico and Gordon Pepper, think the Bank should hammer home its advantage with another hike this week. That would make people sit up and take notice. The other six shadow MPC members vote to leave rates on hold, though three have a bias to raise rates further soon. Inflicting pain is the usual route central banks adopt to convince sceptics they mean business.

There is another way. The Bank’s inflation problem is largely an energy problem. Even if people are sceptical about the CPI, they appear to accept that this is not, yet, a general inflation problem. Public inflation expectations have risen, though only a little. Wage settlements remain very subdued. All that needs to happen for the energy effect on inflation to subside is that prices stop rising; they do not have to fall.

The Bank could do more to explain this. It has an interest in inflation staying low. If it overreacts, it runs the risk of convincing everybody that old-style inflation is really back and that would be disastrous.

PS Before I went on holiday (airport chaos, lost luggage, virulent foreign bug), I suggested scaffolding could be a useful new economic indicator. These days it appears to be popping up everywhere, even for simple household maintenance tasks.

Little did I realise what a Pandora’s box this was. The scaffolding boom, it seems, owes everything to bureaucracy, particularly the Work at Height Regulations 2005, which came in April last year. These, which had their origins in Brussels, define “at height” as any place where a person could be injured falling. In practice, work above two metres appears to require scaffolding. If your window-cleaner is hanging precariously from his ladder, he could be breaking the law.

Some other titbits - all provided by readers. Why does scaffolding stay up so long? Because many smaller scaffolding firms have nowhere to store it, so leave it up until it can be moved to the next job. There’s also what may be an urban myth; that many scaffolding firms started life by dismantling somebody else’s poles and planks and claiming them as their own. I’ll stick to my tried and trusted skip index.

From The Sunday Times, September 3 2006



You want a great measure of inflation? Take a look at college tuition. It represents food, energy, healthcare, wages, infrastructure repair,etc. It is the cost of running a small city. By that measure, inflation is running 5-7%.

John "Jack" Williams, one of the darlings of the internet, uses methodology to calculate inflation that, by many, seems old fashioned. What it provides is an apples-to-apples comparison. I've heard some criticisms of his work, but he puts inflation at 7-8%.

One thing I am confident of is that it is NOT anything near the ridiculous numbers coming of the government bean counters. Their credibility is already lost to those who are not being willfully obtuse.

Best regards
David B. Collum
Professor of Chemistry and Chemical Biology
Cornell University

Posted by: David B. Collum at September 3, 2006 03:48 PM


The CPI is derived from the GEOMETRIC mean of a set of prices for commodities purchased in different parts of the country.

The RPI is derived from the ARITHMETIC mean of a a set of prices for commodities purchased in different parts of the country.

The geometic mean of a set of data will allways be less than (or if the data all have a value of 1, equal to) the arithmetic mean.

It is therefore inveviatable that given the same set of price data that the CPI will be less than the RPI.

Now isn't that convenient.


Posted by: Rhys G Williams at September 4, 2006 11:15 AM

However, the index presented is not the geometric or arithmetic average of th basket - it is the change since the start date in the relevant average. In this case, the geometric average deals better with the kind of problem we have been having in a number of sectors such as furniture. After the sales, retailers try to rebuild margins. In current competitive conditions, this is accompanied by falling volumes. As the Index is, within the year, fixed weighted, the RPI arithmetic index incorporates the price change to rebuild margins as if there was no such thing as elasticity. A geometric mean-based index handles the elasticity effects of price changes better than an arithmetic mean based index.

Therefore, on technical grounds, the type of calculation is better for the CPI, but the big problem is the exclusion of much housing and council tax from the index. We had the RPI and RPIX running parallel for some time, which people could use as the differences were readily explainable. RPIX only useful for monetary policy geeks, RPI for everyone else. Fine. However, CPI presented as the only measure of inflation just excludes too much of spending to be felt reasonable, however sensible it may be as a policy tool to exclude items that are the direct results of policy changes.

Posted by: Paul Bivand at September 5, 2006 12:59 PM

You state "The CBI reported last week that August was the best month for retailers since December 2004. There is a caveat - the survey straddled the base-rate rise - but there was undoubted strength. That simply would not be happening if prices were rising faster"

Is this necessarily so? In Weimar Germany anyone with a wheelbarrow of notes was keen to spend it as fast as possible on an apple lest they could only afford half an apple a day later. Admittedly this demonstrates inflation in extremis but the notion that things will be more expensive in the future can bring forward a purchase. In fact, this very notion may be contributing to the growth in house prices, which has been above wage growth for some years. Potential purchasers feel compelled to stretch themselves because the window of opportunity to become a homeowner may be snatched away from them if they wait.

Posted by: Benjamin Walker at September 6, 2006 06:39 PM

Nice point! I was making the simple point that if real incomes are falling people would be unable to spend. But if we've got a touch of Weimar the Bank really is in trouble.

Posted by: David Smith at September 6, 2006 07:40 PM

Why was the TV license removed from the CPI basket of goods recently? Oh yes it went up in price quite considerably.

Surely this is a good thing to have in the basket as most households have to buy one! ?

Posted by: Rev at September 18, 2006 09:58 PM

It was removed because it has been reclassified as a tax and the CPI does not include any taxes, except when they are included in a prices, such as VAT or excise duty. The licence remains in the RPI. The CPI includes cable and satellite television subscriptions.

Posted by: David Smith at September 18, 2006 10:19 PM

>> It was removed because it has been reclassified as a tax and the CPI does not include any taxes, except when they are included in a prices, such as VAT or excise duty. The licence remains in the RPI. The CPI includes cable and satellite television subscriptions.

What a fortunate time to reclassify it.

Posted by: Rev at September 18, 2006 11:38 PM

I think there is huge merit in a True Inflation Index in addition to the CPI/RPI. Sure, we need these to get a picture of how much the TV licence costs (earlier comment) and there is merit in knowing what a large balanced basket costs. But it doesn't get to the heart of the problem. How much are our savings and pensions worth? Do we really think that the Government can run up such huge debts, spend so much and tax so much without diluting the value of the pound?

My view is that we need to measure the way the Government handles the economy (43% of it!) by constant reference to a True Inflation Index of the things we need to spend every week for austere ration-book existence. Is this basic cost of living moving up or down? Of course, personal rates of inflation are as varied as we are so we need a bottom line. It is all explained at my petition:
Do look and sign if you agree.

Mrs M A Westrop

Posted by: Mrs M A Westrop at February 6, 2008 12:09 PM