HICP at constant tax rates
- Data extracted in January 2018. Most recent data: Further Eurostat information, Main tables and Database. Next planned article update: January 2019
The Harmonised Index of Consumer Prices at constant tax rates (HICP-CT) is a variant of the HICP that measures inflation without the impact of changes in taxes on products, such as value added tax (VAT), excise duties, etc. It thus aims to assess the contribution to inflation of changes in taxes on products. This article analyses HICP-CT data for the European Union (EU) and outlines the methodology underlying these statistics.
- 1 Main statistical findings
- 2 Data sources and availability
- 3 See also
- 4 Further Eurostat information
- 5 Notes
Main statistical findings
Figure 1 shows euro area HICP and HICP-CT December annual rates (% change of the index value in December of the observation year y compared to December of the previous year y-1) from 2003 to 2017, since HICP-CT is calculated by keeping the tax rates from the previous year’s December constant. In the event of a tax change, the difference between HICP and HICP-CT rates indicates its effect on prices, assuming that the full tax change is passed on to the consumer instantaneously. Consequently, if HICP rates are higher than HICP-CT rates it would mean that, overall, taxes on products have increased. Conversely, HICP rates that are lower than HICP-CT rates would imply, on average, a decrease in taxes on products (see Limitations).
Trends in the euro area
Headline HICP and HICP-CT
When looking at the period for which HICP-CT data are available (December 2002 onwards) at euro area level, the annual rate of change in the HICP has been on average 0.2 percentage points (pp) higher than the HICP-CT annual rate. That indicates an increase of taxes on products over time. However, there were sub-periods with clearly discernible patterns: there was not much difference between HICP and HICP-CT annual rates until 2009 (with the exception of 2007). During 2010-2012, the annual rate of change of the HICP-CT turned out to be significantly lower than the general index. Since 2013 the difference has narrowed.
By groups of products
However, the impact of changes in taxes on products varies across groups of products. Four such groups (also called special aggregates) are differentiated as follows: Food, alcohol and tobacco; Energy; Industrial goods excluding energy; and Services. Analysis of Figure 2 shows that in the case of Energy the annual rate of HICP, with the notable exception of 2006, has always been higher than the HICP-CT by the end of every year, In 2007, 2011 and 2012 the difference was above 1.0 pp.
In the case of Food, Alcohol and Tobacco, HICP annual rates by the end of the year have also always been higher than HICP-CT, with the highest values in the years 2004-2006 and 2010-2012 (above 0.3 pp).
In the case of Industrial Goods excluding Energy, the difference between HICP and HICP-CT annual rates was greatest (0.6 pp) in 2007. Additionally, HICP rates have been higher than HICP-CT by the end of every year except 2006, 2008 and 2009 (this would indicate a slight decrease in taxes for that aggregate in those two years).
Differences between HICP and HICP-CT annual rates in December are smaller in the case of Services (never more than 0.4 pp). In 2009 and 2016 HICP-CT annual rates were higher (signalling a broad decrease in taxes on products). There is more information about other groups of products (special aggregates) in the Eurostat database (prc_hicp_cind).
Trends at Member State level
Every country recorded on average an increase of taxes on products (see Table 1), except Denmark, whose rates of HICP and HICP-CT have been similar during last 15 years. Additionally, the rates of HICP and HICP-CT for Belgium, Italy, Austria and Sweden have been nearly identical (only 0.1 pp difference). In contrast, Greece and Latvia had the largest difference between HICP and HICP-CT rates, at 0.7 pp on average. Looking at individual years, the differences in terms of HICP annual rates being higher than HICP-CT rates were greatest in Romania at 5.7 pp in 2010, Lithuania at 5.2 pp (2009), Latvia at 5.1 pp (2009) and Greece at 4.3 pp (2010). Conversely, the only occasions where HICP-CT was higher than HICP by more than 1 pp were recorded in Romania, at 3.0 and 1.9 pp in 2015 and 2016 and the United Kingdom, at 1.1 pp in 2008.
Data sources and availability
Starting from December 2002, HICP-CT all-items and special aggregates are available for most Member States. From January 2013 on, detailed HICP-CT data by COICOP categories are available for all Member States and Norway in the Eurostat database.
The HICP-CT is defined as an HICP index where tax rates are kept constant in the observation period compared to the reference period, i.e. through time. Hence, in the event of a tax rate change, the difference between the current HICP-CT and HICP theoretically would indicate the effect of the tax rate change on price changes. The tax rate may be a certain percentage of the price, or an absolute tax amount levied on a physical unit.
The HICP-CT is a Laspeyres-type price index which reflects the average change of g item prices pi,t from the price reference period t0 (previous year’s December) to month tn). The expenditure shares are denoted as wi. HICP-CT uses HICP expenditure shares. All product-related taxes, which may be merged in vectors τi,t, are part of the prices observed (and refer to the same periods as the prices). The notation LCTP (Laspeyres-type Constant Tax Price index)
indicates that prices in t0 and tn refer to tax rates from the same period t0. Whereas the prices pi,t0 (τi,t0) are observable in the base period t0, the prices pi,t0 (τi,t0) are not directly observable. They have to be calculated using the reporting period’s purchaser prices pi,t, the product-related tax rates τi,tn as well as the tax rates of the base period τi,t0. For more details on the methodology applied, please refer to HICP-CT manual.
As explained previously, the HICP-CT is useful for analysis only if it is compared with the HICP, e.g. where tax rate changes occur in the observation period, the effect of these changes on prices will be seen by comparing the HICP-CT with the HICP.
The taxes falling within the scope of the HICP-CT are taxes on products relating to final monetary consumption expenditure and taxes directly linked to the level of final consumption (see Table 2). The European System of Accounts (ESA) 2010 classifies taxes on products in economic terms.
However, some taxes in the scope of the HICP-CT are small in terms of tax revenue and hence a tax rate change would be unlikely to influence the HICP. Therefore only relevant taxes are kept constant for calculating the HICP-CT (for more information, see HICP-CT reference methodology.
The HICP-CT assumes that changes in tax rates are passed on to the consumer instantaneously and fully. Since most of the time that is not the case, the HICP-CT tends to provide an indication of the upper limit of the impact of tax changes on inflation. In real life, taxes are not always passed on to consumers at the same time as they are introduced. Retailers can adjust the prices of their products to reflect tax changes later on or they can anticipate the rate change and adjust their prices before the new tax rate is applied. It is also possible that retailers will change their prices less than expected or even keep prices constant. Thus the HICP-CT might overestimate a price increase, e.g. if stock is sold at the ‘old’ price.
The price elasticity of supply and demand determines to some extent how taxes will influence prices. The less price-elastic the demand, the bigger the share of tax that consumers have to pay. In the case of more price-elastic demand, the retailers have to bear a bigger part of the tax burden (see Figure 3a).
Hence, the assumption that changes in tax rates are passed on to the consumer fully is true only in the case of perfectly inelastic demand, while in other cases the burden of the tax rate change is shared between consumer and retailer (see Figure 3b).
- D – demand,
- S1 – supply function before tax change,
- S2 – supply function after tax change,
- q1 – quantity purchased before tax change,
- q2 – quantity purchased after tax change
- p1 – price before tax change,
- p2 – price after tax change,
- p3 – price recorded by HICP-CT.
As depicted in Figure 3, whenever a tax rate changes, the supply curve shifts upwards (or downwards, depending on whether the change is positive or negative) by the distance equal to the difference between the new and old tax rates (τ2 - τ1), assuming that changes in tax rates are passed on instantaneously and fully. There is then a new equilibrium point with a price p2 and a quantity q2. Only in the case of perfectly inelastic demand is the consumer’s tax incidence equal to the tax rate change, i.e., p1 equals p3. In other cases, although the HICP-CT records the price p3, the actual price with constant taxes would be p1 (ceteris paribus: assuming there are no other changes). Since cases of perfectly inelastic demand are rare in real life, generally the HICP-CT just gives an indication of an upper limit of the changing tax rate’s influence on inflation.
In case the change of taxes creates second round effects, that effect is not captured by the HICP-CT. Additionally, changes to indirect tax policy that affect the base of the tax and not the rate are not covered by HICP-CT.
Therefore, the HICP-CT is an important tool for analysing the causes of inflation, but it cannot be used as an accurate assessment of the impact of tax policy changes on consumer prices.
Further Eurostat information
- HICP at constant tax rates (prc_hicp_ct)
- HICP at constant taxes — monthly data (index) (prc_hicp_cind)
- HICP at constant taxes — monthly data (annual rate of change) (prc_hicp_cann)
- HICP at constant taxes — monthly data (monthly rate of change) (prc_hicp_cmon)
Methodology / Metadata
- Harmonised indices of consumer prices (HICP (prc_hicp)
- Responsiveness of the quantity demanded of a product to a price variation.