Wednesday, 30 April 2008

UK Food inflation running at 6 percent

The Bank of England might claim that inflation is running at just over two percent, but prices are increasing far faster down at the supermarket. For the last four months, UK food inflation has come in at 6 percent.

Here is an interesting question - how important is food prices in the CPI measure? What would you think? Before you answer, think about what proportion of income you spend on food.

According to the Office of National Statistics, food is given a 9 percent weight. This implies that the average household spends just 9 pennies in every pound on food.


Anonymous said...

Oh, so the CPI isn't accurate? Well, there's a surprise. What's that you say, it's just a political tool to disguise real cost of living rises due to economic mismanagement and debasement of the currency? That it's deliberately understated so the government can underpay on inflation-linked debt and entitlements.

Hmmm, it's almost like we should look at increases in the money supply rather than fraudulent measures such as CPI.


aSteve said...

According to this document (Page 6, Table 1)

Food and non-alcoholic beverages has a CPI weight of 10.9%... with restaurants and hotels - by comparison - with 13.7%... and Alcohol & tobacco with 4.2%.

For a broad interpretation of "restaurants" - then this doesn't seem too unreasonable to me as an indication of the distributions in which people, on average, spend.

By comparison, from Table 2, RPI weights (which sum to 1000) put food at 11.1%; "catering" at 4.7% and Alcohol at 5.9% (excluding tobacco)... or 21.7%... and this is in spite of RPI including owner occupied mortgage costs.

Why the discrepancy, you might well ask... and why does it seem everyone with an interest in international finance is moving to CPI from RPI?

First... Why the difference. RPI has some interesting rules for what it tries to model. It excludes entirely what the most affluent 4% of the population spend their money on (as well as excluding benefit claimants.) The justification is that the spending patterns of the affluent vary wildly from year to year - as their purchases often represent investments... and it would be unhelpful, for example, if the month that Bugatti take payment for Veyrons for the relative expense of motoring to shoot up. This makes RPI a relatively good metric to model the cost of living for the majority of folk living ordinary lives. This is why RPI and not CPI is a helpful metric to estimate fair wage settlements for the “working class”.

That CPI is geometrically derived, while RPI is arithmetic (which is a major public misunderstanding of CPI) is not, in my opinion most relevant in this context. In the context of international trade, it matters not what the majority of people's costs are - but how the majority of money is spent. Our domestic policy is influenced heavily, for example, by the economies of Germany, Spain, Italy, France - etc... but for our domestic policy it doesn't matter if we export to their affluent, middle-class or poor nationals... and similarly vice-versa... for consumption or (supposed/real) investment. In the context of international trade and economic harmonisation (especially with a view to adopting the Euro) CPI is a far more useful metric. This is why CPI is considered the sole measure of inflation in the context of Maastricht and one of the key economic tests of UK suitability for Euro adoption.

It is, in my opinion, to completely miss the point to argue about what CPI should or should not measure. The correct question that everyone should be asking is this: Is CPI the correct sole inflation target for domestic interest rates? This is a deeply political question... and I can see merits in various answers. I feel that it was dishonest of Brown to adopt CPI as he did... because CPI adoption represents a policy only to regulate prices with respect to European trade and not to domestic demand... and I do not believe that is the remit that those who elected New Labour believed they were endorsing. Even Eurostat (who invented HCIP which we renamed CPI) state that it is dangerous to fix national policy solely on this inflation measure. I wouldn't personally care much - as long as the rules aren't changed as soon as it stops working for government's "chums"... if left to run its natural course, CPI - I'm sure - would prove a fair target for interest rates. The problem is principally that the public has been deceived.

In summary, I believe both CPI and RPI are extremely accurately calculated... though spectacularly misunderstood metrics. Neither reflects the real cost of living for any meaningful demographic - let alone everyone. I believe that adopting CPI was politically dishonest and an abuse of trust on Brown's watch. I think that the CPI target significantly exacerbated the UK's ballooning debt problem (and house price crisis) while simultaneously allowing government to proclaim "robust economic growth" that, in reality, was anything but. It perpetuated a feel-good factor among those who they'd hoodwinked to believe that secured debt is synonymous with wealth in order to secure re-election.

powerman said...

Actually, inflation when you exclude imported consumer electronics and other trivia is rather higher.

Alice Cook said...


You are way ahead of me on this one. I had intended to do a series of short posts pointing out the underlying anomolies of the CPI and RPI. I was keen to point out the absence of key items such as housing and how the focus on this limited measure distorts perceptions of inflation.

I was also going to point out the extraordinary difference between monetary growth and CPI inflation.

Excellent comment with some very perceptive observations; I enjoyed reading it.


Anonymous said...

"I was also going to point out the extraordinary difference between monetary growth and CPI inflation."

I look forward to this, but please take due care in what you use as the measure of money. It's just as difficult as CPI calcs. Personally I'd suggest M Prime.


Alice Cook said...


I think household holdings of M4 is probably the best measure of monetary growth.


Anonymous said...


I look forward to seeing the analysis. I'd appreciate if you specify exactly what makes up your M4 numbers and why it is appropriate (i.e. no double counting, silly estimations, political tinkering, such as with CPI).


aSteve said...

Alice, I'm glad you liked my comments about RPI/CPI... if you wanted to take my words as a basis and tweak and add to it to arrive at a blog post - then that's fine by me. I spent a couple of months' free time in 2007 trying to get to the bottom of publications by the ONS. I felt I understood RPI and CPI - but I am still utterly bemused by the details of GDP, for example... and the national accounts proved too steep a learning curve for me, at first pass, at least.

I'm definitely extremely interested in M4, but I'm not entirely convinced that it is as amenable to meaningful direct analysis as RPI and CPI are. I started out thinking M4 was hugely important - maybe exclusively important - but now I don't. I think that just as treasury bills are distinguished by maturity date, personal wealth also conforms to a variety of 'maturities' and that changes in distribution of funds among society is probably more important than absolute M4 totals. As a very simple metaphor/example, if a sane 18 year old man has £250K in savings and somewhere free to live... he'll buy an outrageous sports car; crash it; replace it with a faster one... fill wardrobes with designer clothes and spend nights at the wildest nightclubs - jetting off to frequent luxurious holidays – the cash will have gone by the time he's 21. If the same man were 60 with identical assets, he'd far more likely decide that the money represents a secure comfortable retirement spending £10K+inflation per year for the next 30 years. When such long-term ideas are considered, other forms of investment are equally relevant – e.g. - endowments; life-insurance; pensions; shares, funds & trackers... they all represent money - but not all are included in M4... many even have an an extremely uncertain value.

I understand the idea that we need a better measure of inflation - but I'm not sure that any pre-cooked metric suffices. I don't believe that UK "credible inflation" (whatever that might be) has been running at >10% for a decade – it simply doesn't feel right to me when analysing anecdotals... everything significant that I want (except a nice house) has been roughly static or falling in price for a decade – not going up. The big (and I do mean big) problem facing anyone who wants to go beyond an RPI/CPI approach in establishing where money ends and speculative assets begin. There is not a clear divide... if M4 expansion were to be used as the control for interest rates, securitisation would be employed by banks to bring interest rates as low as they wanted to suit their business model – whatever that meant for inflation and the economy. It is not easy – and this, I suspect, is why Alan Greenspan exclaimed in ~1998 that it is impossible for central banks to control asset bubbles. He was wrong, in my opinion – he should have said that it is impossible to control asset bubbles in an apolitical way... but that is rather different. I expect this idea to receive greater focus by clever people in the immediate future... and await with interest what gets suggested by the likes of the IMF and World Bank. I think everyone knows that the current strategies are flawed... the challenge is in finding an alternative that isn't (potentially) worse. I've a hunch that moving the CPI target lower in the context of sectoral asset price rises at significantly above the risk-free rate for money (BoE base rate) should help... by restricting funding when markets start to get giddy... but it would be a tough sell politically... new governments want to reward their supporters who lost out under the previous government.

I think another error of judgement made by Brown was to mandate that the BoE should set policy for economic growth and to accept full-employment as an ideal... this certainly upset the balance of things. By analogy, if we consider the Government to be a teacher, and the citizens its pupils, GDP is similar to the total class exam results. Where exam results are the prime goal, success is optimally achieved by training for rote answers rather than by education about subjects. Similarly, GDP grows optimally by expanding debt and neglecting risk. Brown's mandate to the BoE was, effectively, not to create situations that give rise to questions about the validity of government policies with respect to the economy – i.e. to support the obfuscation of risk with overly lax monetary policy. It is critical, in my opinion, in future, that metrics presented as a measure for government success are not also used as targets – this leads to extremely unhelpful feedback effects.