Observations

The fall in sterling: who is hit by the rise in inflation?

Date: 15 November 2016
Authors:
Publisher: The IFS

This morning the Office for National Statistics announced that CPI inflation rose to 0.9% in the year to October, down from an inflation rate of 1.0% in the year to September but still substantially up from 0.6% in the year to August. The Bank of England expect inflation to rise further, to 2.4% in 2017 and 2.8% in 2018 – considerably higher than the 1.5% and 2.1% expected back in May. Most of this forecast increase is driven by the recent devaluation of the pound, which pushes up the price of imports. In this Observation we look at how the overall 2.5% increase in the price level which is likely to result from sterling’s decline since the June referendum will affect the prices of different goods. We then look at whether this is likely to have a bigger effect on poorer or richer households.

Since the eve of the EU referendum in June, the value of sterling has fallen by around 12%. The Bank of England estimate that 60% of a fall in sterling is passed through into higher prices faced by UK firms and consumers (this is less than 100% as foreign firms selling into the UK may cut their prices in response to reduced UK demand). UK firms can deal with this increase in costs by switching to (now comparatively cheaper) domestic producers, allowing their profits to fall, or raising their prices. The Bank estimate that in the longer run (around three years) the latter effect dominates, with all of an increase in import costs being passed through to UK consumers in the form of higher prices.

This affects not just goods bought directly from abroad, but also goods which are produced using imported inputs. The stories of British brands raising prices in the wake of the depreciation due to the increased cost of imported ingredients are possible examples of this indirect effect.

Those goods and services that have higher import contents – either through direct imports or inputs into products made in the UK – are likely to see larger price rises following a devaluation. Table 1 shows the expected long run effects on the prices of several broad categories of goods, covering around a third of average household spending. Bear in mind that the table simply isolates the expected impact of the fall in sterling, and is not an overall forecast of inflation for these goods which will also be affected by factors other than exchange rate movements. The Table also shows how richer and poorer households tend to spend different proportions of their budgets on different items.

Table 1: Budget shares by expenditure quintile and expected price increases as a result of the 12% fall in sterling since June 23rd, selected items

Item

Expected price increase (%)

Budget share by expenditure quintile (%)

Bottom

2

3

4

Top

All

Clothing and footwear

3.1

3.0

4.5

4.9

5.5

5.4

4.7

Education

0.2

0.1

0.3

0.3

0.8

2.5

0.8

Food

2.9

24.9

18.9

15.4

12.3

8.5

16.0

Household goods & domestic services

1.0

1.9

1.8

1.8

1.3

1.2

1.6

Petrol and other vehicle running costs

5.1

5.1

8.7

9.9

9.5

7.7

8.2

Tobacco

1.0

2.6

2.0

1.3

1.0

0.5

1.5

All (CPI 2016 weights) a

2.5

         

 

All (Budget share weights) a

2.7

         

 

Source: Authors’ calculations using Living Costs and Food Survey (LCFS) 2014, ONS, Import intensity for each COICOP class, Indirect import content of domestic final demand by product and component 2010, and Input-output supply and use tables 2010.

(a) For ”budget share” weights we use the average LCFS budget share for each item. CPI weights, which are appropriate for calculating the overall price level, are calculated using a different source and different methodology.

Note: See methodology section below for assumptions. The expected price increase of the ‘clothing and footwear’ item is calculated using CPI 2016 weights. Budget shares are calculated at the two digit Classification of Individual Consumption according to Purpose (COICOP) level. Items in the table correspond to COICOP groups of the same name, except “Household goods and domestic services”, which refers to the COICOP group “Goods and services for routine maintenance”, and “Petrol and other vehicle running costs”, which refers to the COICOP group “Operation of personal transport equipment”. Budget shares are calculated for out of spending on goods covered in the CPI (so excluding mortgage interest). Figures are for Great Britain.

Almost 40% of food is imported, meaning that it is expected to see a relatively large price increase of 2.9% as a result of sterling’s devaluation. Clothing and footwear – with import contents of 40% and 50% respectively – are also affected significantly. Petrol and other vehicle running costs see a substantial price rise, but for a slightly different reason. Although only a modest share of petrol is imported, oil is traded on a world market denominated in dollars, which makes even domestically-produced oil highly sensitive to the sterling-dollar exchange rate. Conversely, the prices of more service-oriented items such as education or household goods and domestic services, and more heavily taxed items such as tobacco, tend to be less affected by changes in import prices.

Because different households spend their money on different things, the pattern of price increases summarised above can mean that some households are hit harder than others. For example, households which spend especially high fractions of their budgets on petrol, food and clothing would tend to be hit harder than average.

Clearly the big increases in food prices will hit poorer households particularly hard – while on average 16% of household spending is on food, almost a quarter of the spending by poorer households goes on food. On the other hand increased petrol and clothing prices hit richer households harder. Put these effects together and we find that essentially all major demographic and income groups are affected similarly. Figure 1 shows what happens when we use our budget shares to weight the price increases of different goods. Each expenditure decile is expected to see costs rise by roughly 2.7% on average, though with a slightly bigger effect for richer households. Broadly, this is because those at the upper end of the expenditure distribution are more affected by rises in vehicle, holiday, and furniture costs, while those at the lower end are more affected by rises in food, telephone, and utility costs (because these account for a relatively large share of their budgets). In combination this gives the fairly flat profile seen in Figure 1.

Figure 1: Average expected increase in costs as a proportion of expenditure as a result of the devaluation since June 23rd, by expenditure decile

 

Source: As listed below Table 1. Note: See methodology section below for assumptions. Note: The ‘all’ figure here uses average LCFS budget shares as weights (see footnote to Table 1).

Looking at the effects by region, or by whether or not the household has children, or by whether the household includes a pensioner, all tell a similar story: these groups all see a similar rise in costs as a result of the devaluation. That said, there is a wide distribution of effects according to household spending patterns. The 5% of households most exposed to exchange rate changes are projected to experience a hit of at least 4.1%, while the 5% least exposed see their prices rise by less than 1.5%.

Of course how much individual households are affected by sterling's depreciation does not just depend on the goods they buy but also the degree to which their real incomes adjust as inflation rises. Some households are better protected than others, for example those whose incomes are indexed to inflation (or better), such as those receiving an occupational pension or state pension. Normally working age benefit recipients would also be at least partly protected as benefits usually rise in line with prices, but, as we have discussed before, their benefits have been largely frozen in cash terms, meaning that their income from this source is fully exposed to future inflation. Those in work will, unless they are able to negotiate a bigger pay rise, find that their earnings will stretch less far than they otherwise would have done.

Notes on methodology for estimating price impacts from exchange rate changes

Estimates for price impacts of each item are calculated on the following assumptions:

  • Non-energy imports increase by 60% of the fall in sterling.
  • Energy imports increase by 100% of the fall in sterling.
  • Increases in import prices are fully reflected in consumer prices.
  • UK package holidays have a 0% import intensity and overseas package holidays have a 100% import intensity
  • The relationship between indirect import content by Classification of Products by Activity (CPA) component and Classification of Individual Consumption according to Purpose (COICOP) group is the same as the relationship between household consumption by CPA component and COICOP group.
  • The Office for National Statistics provides the data on import intensity ratios at the two-digit COICOP level. These give the imports valued at pre-tax prices divided by the final gross value of goods sold in each category. To account for the fact that increasing import costs will lead to an increase in the value of taxes such as VAT which are charged on each good, we scale import intensities up according to the average rates of proportional taxes charged within each COICOP group. To calculate average VAT rates at the level of two-digit COICOP codes, we first sort goods at the three-digit level into VAT categories (standard rate, zero rate etc.), then take an expenditure weighted average of the applicable VAT rates within each of the two-digit spending groups.

 As a result of the level of detail in the data we use, one factor that we cannot account for here is whether richer or poorer households tend to buy more imported goods within a particular category (such as “food”). This is because we use the average import intensity for each category, and apply the associated price rise to all households. But if, for example, households further up the expenditure distribution tend to buy more Italian shoes or Wagu beef, they might have a higher than average import intensity in the clothes and food categories, and consequently would be more affected by the devaluation than shown in Figure 1. Conversely, if they tend to buy more British made suits or cheese, they will be less affected.