Measuring inflation

Measuring changes in average price levels requires the use of a device called an index. It is impossible to keep an accurate record of every price change for every good and service in the economy at all times. In 1914, the UK Government began to monitor food prices to help protect workers during the First World War. In 1916, price checks on clothing, fuel and a few other items were added to generate a simple cost of living index.

(Source: ONS).

Today, the UK uses a number of indices to track price changes, including the Consumer Price Index (CPI), which was introduced in 2003, and the much older Retail Price Index (RPI) which was introduced in 1947.  Using an index allows a general picture to develop to show the average price change for a sample of goods and services, measured at monthly intervals. In 2013 two new indices were introduced - the CPIH and the RPIJ.

 

The CPI is based on the European Harmonised Index of Consumer Prices (HICP) and its introduction in the UK allowed for more accurate inflation comparisons between the UK and Europe.

Finding 'average' inflation

Inflation is a rise in the level of ‘average’ prices. However, there are hundreds of thousands of goods and services,  and thousands of stores and outlets in an economy, meaning that recording the price changes of all goods and services in all stores would need an army of bureaucrats and would be highly inefficient. To solve this problem statisticians look at samples of products and outlets to create a general index of inflation. To be of use, the measurement of inflation must be representative of what typical households spend.

Simple example

Imagine an economy, with just 4 products (A = Apples; B = Bananas; C = Carrots and D = Doughnuts. Also, imagine that, over a year, apples increase in price by 20%, bananas increase by 10%; carrots fall by 10% and doughnuts fall by 20%. You would think that these cancel each other out so that inflation is zero. But, what if consumers do not allocate their incomes evenly between the four goods? Then the burden of inflation might actually be ‘positive’ or ‘negative’ rather than neutral. For example, households might choose to spend 40% of their income on apples, 30% of their income on bananas, 20% of their income on carrots and only 10% of their income on doughnuts – in this case it is easy to see that ‘average’ prices (in terms of the effects of these price rises) have risen.

We can put this information into a table:

Good

% Y spent

Price change

Price index*

Weights**

Weighted index

A

 40

 + 20%

 120

 4

 480

B

 30

 + 10%

 110

 3

 330

C

 20

 - 10%

 90

 2

 180

D

 10

 - 20%

 80

 1

 80

 

 

 *The index for each good is based on the number 100. A 5% price rise is an index of 105, and a price drop of 5% is an index of 95.

** The weights are based on the relative importance of the product to households –  for example, if a household spends 8% of their income on chocolate, and 25% on transport, the weights would be 8 and 25 respectively.

To find the average inflation rate, we multiply the weights by the individual price index for inflation (or deflation), and then sum these and divide by the weights, as follows:

Good

% Y spent

Price change

Price index

Weights

Weighted index

A

 40

 + 20%

 120

 4

 480

B

 30

 + 10%

 110

 3

 330

C

 20

 - 10%

 90

 2

 180

D

 10

 - 20%

 80

 1

 80

 

 1070/10

Which gives an inflation index of 107, and an inflation rate of 7%.

Indices

All indices, like the CPI and RPI, have certain key features in common, including:

  1. The use of a sample of typical goods and services bought by ‘average’ households.

  2. The use of a sample of different retail outlets, such as corner shops, supermarkets, and specialist stores taken from across the country.

  3. The tracking of changes in prices from a given starting point, a base year.

  4. The allocation of different types of good with different weights to reflect their varying importance in the consumer’s shopping basket.

  5. Changes are expressed in terms of the number 100 . An index of 110 means 10% inflation since the base year, and an index of 92 means 8% deflation since the base year.

The CPI

The Consumer Price Index (CPI) is calculated by tracking the price movements of 650 items, which represents a basket of goods and services typically bought by the ‘average’ UK household. The basket is updated annually to keep it as representative as possible, and prices are checked on a monthly basis by recording prices at outlets across the UK.  

Goods and services are put into one of 12 categories, as shown:

CATEGORY OVERALL WEIGHT % SUB CATEGORIES
Food and beverages 11.8 22
Alcohol and tobacco 4.4 4
Clothing and footwear 5.7 11
Housing and household services 12.6 5
Furniture and household goods 6.6 11
Health 2.2 3
Transport 15.1 6
Communication 2.3 1
Recreation and culture 14.5 17
Education 2.1 1
Restaurants and hotels 12.8 8
Miscellaneous 9.9 11
Source: ONS

RPIx and RPIy

Until 2003, the RPI was the main indicator of price changes and provided what was called the ‘headline rate’ of inflation. The RPI was then adjusted to include or exclude particular items, including the RPIx and RPIy adjustments. The RPI is a broader measure of inflation than the CPI because it includes costs associated with housing, which the RPI does not.

RPIx is the headline RPI index, minus changes in mortgage interest payments. Taking out mortgage repayments is considered a useful adjustment because the UK housing market plays a significant role in the wider macro-economy. Given that over 60% of householders are owner occupiers, many of whom are repaying mortgages, changes in interest rates, and mortgage rates, can have a considerable impact on spending and the rest of the economy.

Interest rates, which affect mortgage rates, are part of anti-inflationary policy, so it is argued there is a good reason to exclude mortgage costs. A rise in interest rates, designed to reduce inflationary pressure, would push up the RPI, but not the RPIx, so the effects of the policy can be better monitored by looking at RPIx.

Hence, monitoring changes in the RPIx allows policy makers to see the underlying trend in inflation. RPIy is the RPIx minus changes in indirect taxes, such as VAT. Changes in VAT distort inflation data, and make the index less accurate in terms of measuring underlying inflationary pressure, hence it may be useful, at times of increasing VAT rates, to exclude the effects of these changes on retail prices.

The RPIx and RPIy are measures of inflation still recorded and used in the UK, despite the introduction of the CPI in 2003. A large number of wage negotiations are based on the RPI, rather than CPI, which is one reason why the RPI is likely to be used in the future.

Comparing the CPI and RPI

Comparing the Retail Price Index (RPI) and the Consumer Price Index (CPI) raises the following issues:

Mathematical technique of calculation

The RPI uses an arithmetic average of price changes whereas the CPI uses a geometric average, which makes the CPI mathematically more precise. This is because it can continually capture the effects of changes in consumer spending patterns in response to inflation or deflation.

Adjustment

A potential problem with price indices is that they may not adjust quickly enough to reflect changes in spending. Indices are based on a sample of goods and services which are weighted according to how important the good is to the consumer. The importance of a good is based on how much of household income is spent on a product. For example, a typical household may spend 10% of their income on holidays, and therefore holidays will be given 10% of the weighting. But what happens if the cost of a holiday rises by 20%, as a result of a fall in Sterling? Consumers are likely to respond, and reduce their holiday spending. If they now spend only 5% of their income on holidays, the weighting used in the CPI index can be quickly adjusted to 5%. However, the older RPI could not be adjusted so quickly, and could not resolve the problem of changing spending patterns. 

Because of this, and because the CPI does not include housing costs, or council taxes, the RPI gives a slightly higher rate than does the CPI. The CPI gives a higher weighting to energy costs, so change in oil prices have a bigger impact on the CPI inflation rate.

This means that the Bank of England, using the CPI, can set a target of 2% inflation, and not 2.5%. Despite this, the UK authorities still track the RPIx and RPIy.

The CPIH and RPIJ

In recognition of weaknesses in the CPI and RPI two new indices were in introduced in 2013.

The CPIH ('H' for housing costs) was introduced in the UK in March 2013 to reflect the importance of housing costs to UK households. The new CPIH includes a measure of owner occupiers’ housing costs (OOH). These costs include mortgage costs, home insurance, estate agents’ fees and the costs of house renovation, and collectively account for around 10% of total household spending.

In recognition of the need to upgrade the RPI index, a new measure of the RPI was also introduced, called the RPIJ, which uses the same ‘advanced’ mathematical techniques used to calculate the CPI.

Evaluation of all indices

In general terms all indices can be criticised for a number of reasons, including:

Are the samples representative?

If we look at specific types of household, we can get quite different measures of inflation to the general index. For example, if in a given year, the prices of textbooks and rented accommodation for students rise above the average inflation rate, a household made up of students may face a relatively high inflation rate compared with a more typical household. In addition, there is likely to be a regional variation from the average. Therefore, it is quite possible that a married doctor in Manchester experiences a personal inflation rate of 2% whereas a single bus driver living in London, experiences a personal inflation rate of 7%.

Do goods stay the same over time?

A motor vehicle may have ‘inflated’ in price by 20% over a five-year period, but is this actually price inflation? The vehicle may be faster, more efficient on petrol, more comfortable, and safer - so much of the price increase is due to improvements in the vehicle and not to inflation.

When products are fairly standardised, like a litre of milk, or a loaf of bread, quality changes will be small, and the price index will give a more accurate reading of genuine price inflation. With non-standardised products, indices are far less useful.

How up-to-date is the basket?

Indices are usually out of date because the basket used does do not always change quickly enough to reflect current fashions and spending trends. Improvements have been made in terms of adding new goods to the basket, but it still takes up to three years to include new products. Given that new technology products are initially sold at premium prices, the implication is that the current basket always understates true inflation because of the time lag in introducing new technology products.

Why not measure capital goods prices?

Only consumer goods tend to be considered in price indices because the focus of inflation measurement is on households, and not on firms. There is an argument that capital goods prices should be included in a general inflation index.

The 'two-speed' economy

On index may not be good enough if two different sectors of the economy are inflating at different rates. In the UK goods sector, inflation has fallen steadily over the last 15 years, with many goods actually deflating in price. In contrast, service sector inflation has continued at around 4% to 6% per year. An index will average out these two sectors, and it is this average rate that forms the basis of policymaking decisions. This certainly creates a dilemma for policy makers; should they be more conscious of service sector inflation or of goods sector deflation?

Targeting inflation

It is generally recognised that a small amount of inflation is acceptable, with the objective of monetary policy being low and predictable inflation rates. Certainly, given a choice between mild inflation and mild deflation, mild inflation would be the chosen option. Between 1997, when the Bank of England was made independent, and 2004 the (RPI) target rate for inflation was 2.5%, which was an acknowledgement that a little inflation was acceptable.

The CPI target

Since 2003, with the adoption of the CPI, the target has been 2%. The Bank of England must act by increasing or reducing interest rates to achieve this target.

CPI and RPI 2013 - 2016

Inflation monthly | amCharts

It can clearly be seen that the CPI and RPI give a broadly similar picture of retail inflation. Following poor crop harvests, rising food prices contributed to a surge in inflation during 2010, which extended into 2011. During 2012 the rate of inflation fell back as the costs shocks worked their way out.

The RPI often gives a higher value for inflation when mortgage costs are high. This is down to the different weighting of housing and housing costs in the two indices. The RPI is often more volatile.  The RPI is a broader measure than the CPI, and, unlike the CPI, includes a number of housing costs, such as council tax, mortgage repayments, and buildings insurance. A higher and more volatile RPI can be explained because many housing costs, such as council tax and buildings insurance, have risen consistently over the last 15 years. In addition, the volatility of interest rates has had a significant impact on those households with variable rate mortgages.

Long run trend in inflation

Go to: Causes of inflation and deflation

Inflation quarterly | amCharts