The Office for National Statistics released a new dataset last week (28 January 2022) – CPI Consensus Inflation Estimates for UK Household Groups (Wealthy Weights) – It goes on to analyze the different effects of price increases on different groups of people. Other national statistical agencies have been doing this for several years. For example, the Australian Bureau of Statistics regularly publishes its- Selected Cost of Living Index, Australia (Last edition September 2021 quarter, published November 3, 2021).This type of data fills the hole left by the standard CPI version
These datasets are designed to address the variable impact of cost pressures on different household types.
There are many ways to construct a deflator to capture the impact of rising economic costs.
The Consumer Price Index (CPI) is a method in which a fixed basket of goods and services is assembled and an index of price changes over time is calculated based on survey evidence.
Problems can arise if, over time, the basket becomes unrepresentative of the goods and services consumers purchase (for example, if consumers’ choices change over time).
This problem has led statisticians to regularly update the basket and the weights in the basket to reflect current consumption patterns.
There is also the question of whether the basket represents the spending patterns of any particular group. Different income groups have different consumption patterns.
Low-income households spend a higher percentage of their income on basic necessities and less on international travel to ski resorts.
To address this issue, the ABS calculates its— Selected Cost of Living Index – By asking:
How much does after-tax monetary income need to change to allow households to buy the same amount of consumer goods and services as they bought in the base period?
They defined several different households representing different spending patterns — pensioners (age-independent recipients of government transfers), employees, old-age pensioners, and self-paying retirement households — to better refine the index.
Such an index differs from the CPI because it is designed to measure “changes in the purchasing power of households’ after-tax income” relative to their specific spending patterns.
The CPI “is designed to measure price inflation across the household sector” and is therefore “conceptually not an ideal measure for assessing changes in the purchasing power of household disposable income”.
The new ONS series follows this distinction and aims to provide information on how changes in the prices of goods and services affect different income groups, especially low-income households.
In the ONS Blog (26 January 2022) – Measuring the changing prices and costs that households face – ONS Head of Inflation Statistics explained:
…of course the headline inflation rate is just an average number and doesn’t necessarily reflect the price changes experienced by different people…the average annual inflation rate can mask a lot of things. When you collect 180,000 prices on a regular basis, you observe a lot of variety…
The CPIH title measures the average, but everyone has their own personal inflation rate. If you drive to and from get off work every day, some people may spend most of their income on gas and electricity, or gas…
Before I go any further, you may be wondering what “regal weight” means.
A price index is simply a weighted average of individual prices.
There are two broad ways to calculate weights. The first is called the democratic approach, in which the share of each good in total household spending is averaged across all households to form the total weight of the index.
The second is the chaebol approach, in which the share of each good in total household spending is weighted by each household’s share of total consumption spending.
In the UK, headline (headline) inflation was 5.4% in December.
The question is whether this is a reasonable summary of the impact on the purchasing power of British citizens.
If one’s spending matches the goods and services in the CPI basket, and the proportions are based on the weights used by the ONS, the answer is that the headline figure will be an accurate aggregate.
But, in general, the answer is probably not.
This is why the ONS has resumed publishing its population-specific measures after a break due to Covid-19 difficulties with data collection.
In the latest version, we learned that:
1. “Since 2014, high- and low-income households have experienced similar annual inflation rates.”
2. “The inflationary experience of high-income households is driven by rising transportation costs, which spend a larger proportion of their expenditures on transportation than low-income households, and for low-income households, which have more housing-related costs a factor.”
3. “Retirement and non-retirement households have experienced similar inflation rates since April 2021.”
The chart below compares the January 2020 index of the lowest decile of 100 income recipients with the highest decile of income recipients.
The patterns were remarkably similar, with the cost of living rising by 6.6 per cent for the highest earners over two years, while the cost of living for the lowest rose 6.1 per cent.
There is hardly any difference between income deciles, which either means that the different spending patterns in the income distribution are themselves offsetting – as in point 2 above, or that the spending patterns are not very different.
The headline CPIH ratio actually closely reflects the experience of all deciles.
There is also not much difference when calculating the series for retirees and non-retirees, whose cost of living has increased by 6.1% since January 2020 (to December 2021), while the former has increased by 6.4%.
ONS data also breaks down data by 12 spending categories and equivalent disposable income deciles:
- 1. Food and non-alcoholic beverages
- 2. Alcoholic beverages and tobacco
- 3. Clothing and footwear
- 4. Housing, water, electricity, gas and other fuels
- 5. Furniture, household equipment and repairs
- 6. Health
- 7. Shipping
- 8. Communication
- 9. Entertainment and Culture
- 10. Education
- 11. Restaurants and Hotels
- 12. Miscellaneous goods and services
I recently saw a headline in The Guardian UK saying the UK is “cutting the poor out of food”.
The ONS will release more data in May 2022, which will allow us to say more about this issue.
What the existing dataset allows us to track is the change in the share of expenditure in COICOP (United Nations Classification Framework – Classification of Personal Consumption by Purpose) from 2005 to December 2021.
I created a table to summarize the percentage change in stocks.
It’s too complicated to provide a continuous review for each interesting finding.
But the bottom two decile now spend a fraction of their disposable income on food, while the top four decile spend more of their income on food.
All deciles spend a larger proportion of their spending on housing and (mostly) education.
Restaurants and hotels have a lower overall share, which is clearly the impact of the pandemic.
The ONS data also allows us to see the contribution of price increases in each spending category to the annual increase in the cost of living for the second and ninth quintiles between January 2006 and December 2021.
The graph below shows the contribution from the food and non-alcoholic beverage categories, while the second graph shows the contribution from the housing, water, electricity, gas, and other fuel categories.
During the global financial crisis, food was a major contributor to low-income households and to a lesser extent for high-income households.
They’ve all experienced the effects of higher food prices during the pandemic — more or less equally.
Rising house prices have hit low-income households more severely.
Finally, transportation contributions are currently higher for high-income households.
Australian experience
The chosen cost of living index is not strictly comparable to the ONS UK data because first, household types are different, although the ABS data does distinguish between ‘old age pensioner households’ and ’employee households’, which is not exactly the same as that in the UK data. Retirees and non-retirees are the same.
The ABS also has a “Self-financed retirement family” category to distinguish those on the government’s old age pension.
The chart below sets the index at 100 for the December quarter before the pandemic and tracks the evolution of the cost of living index for four household types.
It is clear that the cost of living for employee families and unemployment benefits recipients has fallen in the first year of the pandemic (in part due to the government’s policy decision – free childcare, for example in the March 2020 quarter).
In 2021, as supply constraints intensify, transportation (gasoline) and housing costs (declining affordability) push the cost of living index up at a similar rate for all groups.
During the pandemic (through the September 2021 quarter), our cost of living has increased as follows:
1. Employee households – 1.4% higher.
2. Older households receiving pensions – 2.9% higher.
3. Other households receiving government transfers – 2.6% higher.
4. Self-paying retirement households – 3.1% higher.
If we use the average CPI measure, then we conclude that the cost of living has risen by an average of 4.4%, which tells you why the measure of choice is a better guide to declining purchasing power.
in conclusion
This blog post is actually about my familiarity with this new dataset that ONS is releasing.
It gets richer as they publish more data (crosstabs, etc.).
It already tells an interesting story about what’s going on in the income distribution.
My Helsinki Lecture Series 2022
I am currently giving an annual lecture on Modern Monetary Theory (MMT) and the global economy at the University of Helsinki.
We’ve reached Lecture 3 in a 6-part series.
The teaching plan will be:
- Tuesday, January 25, 2022 – Streaming public lecture (YouTube) starting at 10:15 Helsinki time.
- Wednesday, January 26th – First Zoom Class Lecture – 08:15-09:45 Helsinki time.
- Thursday, January 27th – 2nd Zoom Lecture – 10:15-11:45 Helsinki time.
- Tuesday, February 1 – 3rd Zoom Lecture – 10:15-11:45 Helsinki time.
- Wednesday, February 2 – 4th Zoom Lecture – 08:15-09:45 Helsinki time.
- Thursday 3 February – Final Zoom Lecture – 10:15-11:45 Helsinki time.
The Zoom link to the lecture is:
https://helsinki.zoom.us/j/5354174274?pwd=OHdTdWJzSHNndHpyVkV2Y0lJUExRZz09
Conference number: 535 417 4274
Password: ETZhk9
This is an MMTed initiative.
MMTed MOOC – Modern Monetary Theory: Economics in the 21st Century
MMTed – You are invited to sign up for the edX MOOC – Modern Monetary Theory: Economics in the 21st Century.
It is free and the 4-week course starts on February 9, 2022 (note that edX adjusts start dates for time zones, so some start dates will be listed as February 8, 2022).
This course is offered through the Newcastle University edX programme.
The course is self-paced, so you can learn and participate as you choose. New material is released at the beginning of every four weeks.
Get a proper understanding of MMT with tons of videos, discussions, and more.
for – More details.
Enough for today!
(c) Copyright 2022 William Mitchell. all rights reserved.
