How inflation is calculated - does one very bad month affect inflation for the next 11 months?

Pat38493

Ron246235 said:

These answers are really helpful. It looks like I was right - the figure that is reported is misleading as it depends on what happened in the whole of the last year. But scaling up the inflation that happened within the last month would also be problematic as there is a lot of volatility. 

I have never seen a journalist talk about this issue either. I suppose they want to avoid confusing people, so they just say "this is how inflation is measured". To be fair to the journalists, some of the general public think that if inflation goes down then that means prices are decreasing - so perhaps they are just trying to avoid that even worse misunderstanding. 

The answer from @hugheskevi is interesting. Prices rose by 1.8% from January to May, so there will have to be almost zero inflation for the rest of the year - almost impossible - if annualised inflation is to stay at 2%. It would be good if everyone could understand that. 

I have seen some headlines from US news sites in the past where it appeared to me that they do announce monthly inflation figures as well as rolling annual ones, at least in the financial part of the news.

Hoenir

Pat38493 said:

hugheskevi said:

The principle is correct.

At the moment, CPI inflation is 2% over the last 12 months. But that is largely because there was almost no inflation between May 2023 and January 2024. Since January 2024 prices have increased by 1.8%, so well above an annualised 2% increase.

Yet all we are going to hear about for the next few weeks is speculation low inflation will lead to interest rate cuts, and how the govt's economic plan to bring down inflation is working. This is despite prices now having to remain almost unchanged until January 2024 if annual inflation is going to remain at 2% (ie rising annualised inflation is extremely probable in the coming months).

But looking on a month-to-month basis would exaggerate any volatility in the figures, so it is necessary to take account of both short-term changes and longer term trends.

CPI index numbers

2023 MAY	131.3
2023 JUN	131.5
2023 JUL	130.9
2023 AUG	131.3
2023 SEP	132.0
2023 OCT	132.0
2023 NOV	131.7
2023 DEC	132.2
2024 JAN	131.5
2024 FEB	132.3
2024 MAR	133.0
2024 APR	133.5
2024 MAY	133.9

I have never once heard any UK journalist discuss these points or mention for example that most of the inflation in the current annual number actually happened in the last 5 months.  Surely this is what the BOE considers when they make their decisions.

Today's headline rate is low. However within that "Services" inflation is running at over 5%. Likewise the increase in the minimum wage takes effect in April every year. Takes a few months for the full impact on wages to filter through to the real economy. The quarterly adjustment to the energy cap creates one off movements. Data and the future projections there of are what the BOE are usng as the basis for any decision. Akin to steering a yacht into the wind there's no straight course to follow. 

MattMattMattUK

Ron246235 said:

These answers are really helpful. It looks like I was right - the figure that is reported is misleading as it depends on what happened in the whole of the last year. But scaling up the inflation that happened within the last month would also be problematic as there is a lot of volatility. 

Something is not misleading just because you do not understand it.

Ron246235 said:

I have never seen a journalist talk about this issue either. I suppose they want to avoid confusing people, so they just say "this is how inflation is measured". To be fair to the journalists, some of the general public think that if inflation goes down then that means prices are decreasing - so perhaps they are just trying to avoid that even worse misunderstanding. 

The general public have little to no proper understanding of economics, let alone the complexities of inflation.

Ron246235 said:

The answer from @hugheskevi is interesting. Prices rose by 1.8% from January to May, so there will have to be almost zero inflation for the rest of the year - almost impossible - if annualised inflation is to stay at 2%. It would be good if everyone could understand that. 

That is not how inflation works.

Ron246235

MattMattMattUK said:

Ron246235 said:

These answers are really helpful. It looks like I was right - the figure that is reported is misleading as it depends on what happened in the whole of the last year. But scaling up the inflation that happened within the last month would also be problematic as there is a lot of volatility. 

Something is not misleading just because you do not understand it.

Thank you for that. I do understand it, but I think it is misleading - it has the potential to mislead, or it doesn't show what you might intuitively think it would show.

MattMattMattUK said:

Ron246235 said:

The answer from @hugheskevi is interesting. Prices rose by 1.8% from January to May, so there will have to be almost zero inflation for the rest of the year - almost impossible - if annualised inflation is to stay at 2%. It would be good if everyone could understand that. 

That is not how inflation works.

What do you mean? 

zagfles

The main point of inflation figures is a benchmark for stuff like pay rises, benefit rises, rises in stuff like excise duties, some tax thresholds (that aren't frozen) etc. For things like that, ie which tend to change annually, a 12-month rolling average is exactly what you need. You want to know how much prices have rises compared to a year ago, and if you use the same month's figures every year then stuff will keep pace with inflation. So that's why inflation figures are quoted as they are. 

They are not a very useful guide to the current rate prices are rising. Any more than how many miles you drove in the last hour is a guide to your current speed. You can get a more accurate figure for that using the indices eg for CPI CPIH INDEX 00: ALL ITEMS 2015=100 - Office for National Statistics (ons.gov.uk) so Apr index was 132.2 and May 132.7 so a monthly change of 0.378% which is an annual rate of 4.63%

But the monthly rate could be skewed by seasonal variations, eg going for a meal at Christmas is expensive, and shops have sales in January etc, so it might be misleading to rely on monthly changes.