RPI to 6% !

[dealtn]

The increasing spread between the RPI and CPI is beginning to smell a bit fishy, but not at all surprising. There's a reason they changed the official measure, after all.

There's a perfectly simple explanation:

The CPI calculates the average price increase as a percentage for a basket of 700 different goods and services. Around the middle of each month it collects information on prices of these commodities from 120,000 different retailing outlets. Note that unlike the RPI, the CPI takes the geometric mean of prices to aggregate items at the lowest levels, instead of the arithmetic mean. This means that the CPI will generally be lower than the RPI. The rationale is that this accounts for the fact that consumers will buy less of something if its price goes up, and more if its price goes down; it also ensures that if prices go up and then revert to the previous level, the CPI also reverts to its previous level (which is not the case with the calculation method used for the RPI). According to the ONS, this difference in averaging method is the largest contributing factor to the differences between the RPI and the CPI.[4]

Source: https://en.wikipedia.org/wiki/Consumer_ ... ed_Kingdom)

As prices are going up faster, CPI's use of the geometric mean serves to widen the gap between CPI and RPI.

MDW1954

While this may explain the increasing spread, you then have to question altogether if CPI is fit for purpose during times of higher inflation.. after all, the stuff that is going up the most like food, petrol and utilities don't have easy substitutes.

It's RPI that's not fit for purpose, not CPI - fashionable though it is every time inflation widens for many to claim it is the better methodology.

Food petrol and utilities do have substitutes. One of which is buying less of it. That's not a moral or judgemental comment, simply a typical and understandable human behaviour as a consequence of that price rise. It has to be taken into account when measuring inflation. In not doing so you are measuring something else (albeit closely linked to inflation, which is an economic term).

[BT63]

It's RPI that's not fit for purpose, not CPI - fashionable though it is every time inflation widens for many to claim it is the better methodology.

Food petrol and utilities do have substitutes. One of which is buying less of it. That's not a moral or judgemental comment, simply a typical and understandable human behaviour as a consequence of that price rise. It has to be taken into account when measuring inflation. In not doing so you are measuring something else (albeit closely linked to inflation, which is an economic term).

I want to know how much my cost of living changes assuming I continue buying exactly the same items and exactly the same quantities.

Adjusting to underweight items which went up in price and overweight items which went down in price is not measuring my cost of living.

[MDW1954]

It's RPI that's not fit for purpose, not CPI - fashionable though it is every time inflation widens for many to claim it is the better methodology.

Food petrol and utilities do have substitutes. One of which is buying less of it. That's not a moral or judgemental comment, simply a typical and understandable human behaviour as a consequence of that price rise. It has to be taken into account when measuring inflation. In not doing so you are measuring something else (albeit closely linked to inflation, which is an economic term).

As our resident inflation expert, I'm glad you weighed in, dealtn.

MDW1954

[dealtn]

It's RPI that's not fit for purpose, not CPI - fashionable though it is every time inflation widens for many to claim it is the better methodology.

Food petrol and utilities do have substitutes. One of which is buying less of it. That's not a moral or judgemental comment, simply a typical and understandable human behaviour as a consequence of that price rise. It has to be taken into account when measuring inflation. In not doing so you are measuring something else (albeit closely linked to inflation, which is an economic term).

I want to know how much my cost of living changes assuming I continue buying exactly the same items and exactly the same quantities.

Adjusting to underweight items which went up in price and overweight items which went down in price is not measuring my cost of living.

That maybe your requirement, but that isn't inflation.

If you want to measure that you can, but the ONS won't. How important is it to you, and is the error between your own measure of your cost of living requirement and the macroeconomic inflation equivalent significant enough?

Since you brought it up, if you're intention is to buy exactly the same items and same quantities, what do you do if your income doesn't grow by exactly the same? Do you adjust your savings, do you sell investments, do you borrow? How do you adjust for less interest earned, less dividends received, more interest paid, in your calculation? Others have similar intentions but maybe buy the same number but a slightly lower quality, some buy slightly less and the same (sometimes a higher) quality. Indeed in some extreme examples prices rising lead to people buying more!

[vand]

The increasing spread between the RPI and CPI is beginning to smell a bit fishy, but not at all surprising. There's a reason they changed the official measure, after all.

There's a perfectly simple explanation:

The CPI calculates the average price increase as a percentage for a basket of 700 different goods and services. Around the middle of each month it collects information on prices of these commodities from 120,000 different retailing outlets. Note that unlike the RPI, the CPI takes the geometric mean of prices to aggregate items at the lowest levels, instead of the arithmetic mean. This means that the CPI will generally be lower than the RPI. The rationale is that this accounts for the fact that consumers will buy less of something if its price goes up, and more if its price goes down; it also ensures that if prices go up and then revert to the previous level, the CPI also reverts to its previous level (which is not the case with the calculation method used for the RPI). According to the ONS, this difference in averaging method is the largest contributing factor to the differences between the RPI and the CPI.[4]

Source: https://en.wikipedia.org/wiki/Consumer_ ... ed_Kingdom)

As prices are going up faster, CPI's use of the geometric mean serves to widen the gap between CPI and RPI.

MDW1954

While this may explain the increasing spread, you then have to question altogether if CPI is fit for purpose during times of higher inflation.. after all, the stuff that is going up the most like food, petrol and utilities don't have easy substitutes.

It's RPI that's not fit for purpose, not CPI - fashionable though it is every time inflation widens for many to claim it is the better methodology.

Food petrol and utilities do have substitutes. One of which is buying less of it. That's not a moral or judgemental comment, simply a typical and understandable human behaviour as a consequence of that price rise. It has to be taken into account when measuring inflation. In not doing so you are measuring something else (albeit closely linked to inflation, which is an economic term).

If I google the term "substitute" it tell me that its "a person or thing acting or serving in place of another."

"Nothing" in place of "something" doesn't work as a substitute for me, sorry. That's literally the opposite of a substitute.

[dealtn]

The increasing spread between the RPI and CPI is beginning to smell a bit fishy, but not at all surprising. There's a reason they changed the official measure, after all.

There's a perfectly simple explanation:

The CPI calculates the average price increase as a percentage for a basket of 700 different goods and services. Around the middle of each month it collects information on prices of these commodities from 120,000 different retailing outlets. Note that unlike the RPI, the CPI takes the geometric mean of prices to aggregate items at the lowest levels, instead of the arithmetic mean. This means that the CPI will generally be lower than the RPI. The rationale is that this accounts for the fact that consumers will buy less of something if its price goes up, and more if its price goes down; it also ensures that if prices go up and then revert to the previous level, the CPI also reverts to its previous level (which is not the case with the calculation method used for the RPI). According to the ONS, this difference in averaging method is the largest contributing factor to the differences between the RPI and the CPI.[4]

Source: https://en.wikipedia.org/wiki/Consumer_ ... ed_Kingdom)

As prices are going up faster, CPI's use of the geometric mean serves to widen the gap between CPI and RPI.

MDW1954

While this may explain the increasing spread, you then have to question altogether if CPI is fit for purpose during times of higher inflation.. after all, the stuff that is going up the most like food, petrol and utilities don't have easy substitutes.

It's RPI that's not fit for purpose, not CPI - fashionable though it is every time inflation widens for many to claim it is the better methodology.

Food petrol and utilities do have substitutes. One of which is buying less of it. That's not a moral or judgemental comment, simply a typical and understandable human behaviour as a consequence of that price rise. It has to be taken into account when measuring inflation. In not doing so you are measuring something else (albeit closely linked to inflation, which is an economic term).

If I google the term "substitute" it tell me that its "a person or thing acting or serving in place of another."

"Nothing" in place of "something" doesn't work as a substitute for me, sorry. That's literally the opposite of a substitute.

So buying 99 litres instead of 100 litres is the equivalent of buying something and substituting it with nothing in your world is it?

I think you have "literally" not understood what I have said.

[tjh290633]

Looking at EARN01.xls downloadable from https://www.ons.gov.uk/employmentandlab ... ingsearn01 the current rate of wage inflation as at December was 4.9%, see sheet AWE_Total Pay.

TJH

[gryffron]

There's a perfectly simple explanation:...
As prices are going up faster, CPI's use of the geometric mean serves to widen the gap between CPI and RPI.

I accept that makes a small difference. But I would have thought the biggest difference by far was housing costs, which are included in RPI but not CPI.

It makes a great deal of sense that housing is ignored for setting interest rates. Otherwise you get positive feedback. Increase mortgage rates, increase RPI, increase mortgage rates, increase RPI,...
But it's a bit of a jump to assume housing costs will always follow general inflation over the long term.
And equally to use CPI for setting wages/benefits - given that housing costs are such a huge part of most people's outgoings.

Gryff

[MDW1954]

And equally to use CPI for setting wages/benefits - given that housing costs are such a huge part of most people's outgoings.
Gryff

Yes: but apparantly, when CPI was designed, that wasn't the intention -- benefits weren't the remit. That was a political decision, taken afterwards. There may be something about that in that wiki link -- I know that I've read it somewhere, but can't remember where.

And with all due respect, don't be too dismissive of the impact of the geometric mean. It does produce very different results.

My own view, FWIW, is that I have some sympathy for both the viewpoints expressed on this thread. As an individual, I'm interested in inflation as measured against the things that I habitually buy, in the quantities that I habitually buy them in. (Vand's view) If I were a policymaker or professional economist, I'd be interested in inflation as measured after people had made necessary adjustments to their spending. (Dealtn's view, ditto ONS)

MDW1954

[GoSeigen]

There's a perfectly simple explanation:...
As prices are going up faster, CPI's use of the geometric mean serves to widen the gap between CPI and RPI.

I accept that makes a small difference.

For those of us not immediately familiar with geometric mean vs arithmetic mean please could you do a quick worked example showing how small the difference is. How does the difference vary with the magnitude of inflation?

Thanks.

GS

[servodude]

There's a perfectly simple explanation:...
As prices are going up faster, CPI's use of the geometric mean serves to widen the gap between CPI and RPI.

I accept that makes a small difference.

For those of us not immediately familiar with geometric mean vs arithmetic mean please could you do a quick worked example showing how small the difference is. How does the difference vary with the magnitude of inflation?

Thanks.

GS

It's useful for things you multiply together: like cascaded gains (as the ubiquitous example I think of everything in) or musical frequencies or inflation

Try a set containing just 2 and18
Arithmetic mean = (2+18)/2 = 20/2 = 10
Geometric mean = (2 x 18) ^(1/2) = (36)^(1/2) = 6

Numbers chosen so that the sums are easy.

-sd

[GoSeigen]

There's a perfectly simple explanation:...
As prices are going up faster, CPI's use of the geometric mean serves to widen the gap between CPI and RPI.

I accept that makes a small difference.

For those of us not immediately familiar with geometric mean vs arithmetic mean please could you do a quick worked example showing how small the difference is. How does the difference vary with the magnitude of inflation?

Thanks.

GS

It's useful for things you multiply together: like cascaded gains (as the ubiquitous example I think of everything in) or musical frequencies or inflation

Try a set containing just 2 and18
Arithmetic mean = (2+18)/2 = 20/2 = 10
Geometric mean = (2 x 18) ^(1/2) = (36)^(1/2) = 6

Numbers chosen so that the sums are easy.

-sd

I was actually hoping Gryffron would do this and in a way that illustrates what he calls the small difference between RPI and CPI and quantifies how it varies with magnitude of inflation. I do know the basics of geometric vs arithmetic mean as per your example...

GS

[dealtn]

There's a perfectly simple explanation:...
As prices are going up faster, CPI's use of the geometric mean serves to widen the gap between CPI and RPI.

I accept that makes a small difference. But I would have thought the biggest difference by far was housing costs, which are included in RPI but not CPI.

Well CPI doesn't include housing but there is a (preferred) measure also released called CPIH which does. So if you wish to attempt to strip out housing and see if it is, as you think, housing costs that are the biggest difference, you can do so.

To save you looking it up compare CPI and CPIH.

In January 2022 the former is at 5.5%, up slightly from an annual rate of 5.4% the previous month. (The monthly change is -0.1% compared to -0.2% for the same month last year).

In contrast in January 2022 the CPIH measure is at 4.9% which is up by 0.1% on the 4.8% last month. (The monthly change was 0.0% compared to -0.1% for the same month last year).

On that basis it's difficult to claim the largest difference between CPI and RPI is the housing element, at least currently.

[Nimrod103]

There's a perfectly simple explanation:...
As prices are going up faster, CPI's use of the geometric mean serves to widen the gap between CPI and RPI.

I accept that makes a small difference. But I would have thought the biggest difference by far was housing costs, which are included in RPI but not CPI.

Well CPI doesn't include housing but there is a (preferred) measure also released called CPIH which does. So if you wish to attempt to strip out housing and see if it is, as you think, housing costs that are the biggest difference, you can do so.

To save you looking it up compare CPI and CPIH.

In January 2022 the former is at 5.5%, up slightly from an annual rate of 5.4% the previous month. (The monthly change is -0.1% compared to -0.2% for the same month last year).

In contrast in January 2022 the CPIH measure is at 4.9% which is up by 0.1% on the 4.8% last month. (The monthly change was 0.0% compared to -0.1% for the same month last year).

On that basis it's difficult to claim the largest difference between CPI and RPI is the housing element, at least currently.

I thought that everyone had agreed that house prices across the whole UK were currently rising rapidy, so why is CPIH lower than CPI?

[GoSeigen]

I thought that everyone had agreed that house prices across the whole UK were currently rising rapidy, so why is CPIH lower than CPI?

Most people already own their houses, the only effect rising prices will have is on the few thousand people buying on the margins. Not only that, the same rise in prices benefits the sellers (most of whom will also be buying a new property). Is this not obvious?

Housing costs are not determined by the rise in house prices but in the actual costs of housing to the people occupying. This is primarily rents and imputed rent; rents have been subdued for several years.


GS

[Nimrod103]

I thought that everyone had agreed that house prices across the whole UK were currently rising rapidy, so why is CPIH lower than CPI?

Most people already own their houses, the only effect rising prices will have is on the few thousand people buying on the margins. Not only that, the same rise in prices benefits the sellers (most of whom will also be buying a new property). Is this not obvious?

Housing costs are not determined by the rise in house prices but in the actual costs of housing to the people occupying. This is primarily rents and imputed rent; rents have been subdued for several years.


GS

OK, CPIH and RPI do not include house prices , but have a housing element which presumably represents residential rents? Rents were surpressed during the pandemic, but just like petrol prices, have since rebounded strongly.

Average rents hit nearly £1,000 per month
By Nicky Burridge 08 February 2022
The rate at which rents are rising is at a 13-year high, while demand soared by 76% in the New Year.
Based on new lets agreed, average UK rents are now £969 per month
The rate at which rents are rising reached a 13-year high of 8.3% in the final three months of 2021
Demand for rental homes soared by 76% in the New Year, compared with the same period between 2018 and 2021
(https://www.zoopla.co.uk/discover/prope ... per-month/)

Anecdotally, many landlords were giving rent holidays or reductions to tenants who were under financial stress during the pandemic. These reductions have now stopped, evictions are now allowed, and local rents are rising.

But there appears to be no element of this in the CPIH.

[1nvest]

We are now seeing a squeeze in living standards as wages fail to keep up with inflation. This is why inflation has always been historically destructive to the middle classes.

Range Rover a year ago £95,000, recent £100,000 (+5.3% more expensive)
Or 72.2 ounces of gold a year ago, 71.7 ounces recently (-3.3% cheaper)
Or in FT250TR shares 5.71 shares a year ago, 5.84 shares recently (+2.1% more expensive)

Being middle class has a greater tendency towards the option to decide which 'currency' to spend. Working class are more inclined to live hand-to-mouth and be limited to only cash/wage payment currency options.

[Wuffle]

Interesting that you pick a Range Rover for a couple of reasons ( and I am from a long line of Brummie car assemblers so I take a casual interest).
It used to be top of the tree in its market but over the last decade or two the global manufacturers have marched upmarket to provide ever more expensive products to sell to the increasing numbers of wealthy.
In the shorter term, it is fairly common knowledge that the finite amount of chips available have been going into the highest margin products across all global manufacturers - obviously the most expensive - whereas actually buying a base model is, to all intents and purposes impossible.

W.

[stevensfo]

We are now seeing a squeeze in living standards as wages fail to keep up with inflation. This is why inflation has always been historically destructive to the middle classes.

Range Rover a year ago £95,000, recent £100,000 (+5.3% more expensive)
Or 72.2 ounces of gold a year ago, 71.7 ounces recently (-3.3% cheaper)
Or in FT250TR shares 5.71 shares a year ago, 5.84 shares recently (+2.1% more expensive)

Being middle class has a greater tendency towards the option to decide which 'currency' to spend. Working class are more inclined to live hand-to-mouth and be limited to only cash/wage payment currency options.

Not that I disagree with you, but I think it would be more appropriate these days to use the terms 'Higher/Lower wage-earners' or similar, rather than Working/Middle class. I find the whole idea of 'class' in 2022 very anachronistic and well past its sell-by date. In any case, is there a consensus on what it all means?

Besides which, during my life I've noticed plenty of those you might describe as 'Working class' to be absolutely rolling in it, and some of those calling themselves 'Middle class' who barely have two pennies to rub together.

Steve (who considers himself slightly lower middle ruling class with pseudo lower class parents who had middle class aspirations)

[1nvest]

With inflation at 6%, you lose half your money in 11 years. The only way to outperform that is crypto. In less than 3 months alone, from firework night 2021 to 21st January 2021, it already lost half of your money