Domestic Market-Determined
Inflation
Bo William Hansen and Dan Knudsen, Economics.
INTRODUCTION
Inflation, measured as the annual change in consumer prices, is one of the key macroeconomic indicators. It is also among those that attract most attention.
However, the development in the overall consumer-price index cannot stand alone as an expression of the inflationary process. It can sometimes be useful to supplement it with a narrower definition of inflation. The overall consumer-price index weighs together the fluctuations in prices for many different goods and services, which are weighted according to their share of consumer spending. For a considerable proportion of these components, the price development is governed by international market conditions, by special factors such as harvest yields and weather conditions, or by government regulation in the case of administered prices that do not reflect the price pressure in the domestic market. Neither do changes in indirect taxation and price movements for imported goods reflect the inflationary process, by which is meant the price pressure from Danish business enterprises' payroll costs and profit margins.
Since 1984 Danmarks Nationalbank has on a regular basis published a measure of domestic market-determined inflation (IMI), for the first time in the Monetary Review, May 1984. This article describes and updates the basis for calculation of IMI.[1] IMI reflects the development in the market-determined Danish prices for goods and services for private consumption and can be perceived as an expression of underlying inflation in
For all EU member states, Eurostat publishes a measure that does not indicate the development in the entire consumer-price index, but in certain of its sub-aggregates. Eurostat compiles more than one expression of core inflation, calculated by "stripping" the consumer-price index. One of the expressions describes the development in consumer prices excluding energy and food. Energy prices typically match oil price trends in the global market, while food prices are sensitive to weather conditions and seasonal variations, just as prices for a number of food products are affected by various subsidy and regulation schemes. In the short term, the prices of these two components can vary considerably and conceal the true picture of the underlying development in prices.
While stripping energy and food from consumer prices to calculate core inflation is a simple matter, setting up an expression of IMI requires more calculation work. To achieve IMI, further adjustment is made for administered prices, indirect taxes, and indirect energy and import content.
CALCULATING THE PRICE INDEX BEHIND IMI
Table 1 presents the structure of the price index for IMI (IMI index). The starting point is HICP (the EU's Harmonised Index of Consumer Prices). Formulating the transition from the better-known HICP to the IMI index makes it e.g. easier to relate IMI to core inflation. There are three calculation stages in the transition: 1) "exogenous" sub-indices are eliminated from HICP; 2) indirect taxes are eliminated; and 3) the energy and import content are eliminated, based on input-output calculations.
| FROM HICP TO IMI INDEX |
Tabel 1
|
| Harmonised index of consumer prices (overall inflation) | |
| - Energy and food | |
| HICP excluding energy and food (core inflation) | |
| - Products with administered prices | |
| HICP excluding energy, food and administered prices | |
| - Indirect taxes | |
| Index of net retail prices excluding energy, food and administered prices | |
| - Input-output calculated effect of energy price | |
| - Input-output calculated effect of import price excluding energy | |
| IMI index (domestic market-determined inflation) |
Eliminating sub-indices from HICP
A number of "exogenously" determined prices that do not reflect the development in the domestic market are deducted from HICP. These are energy, food (including beverages and tobacco) and certain services with administered prices. The elimination or stripping of HICP takes place by removing the relevant sub-indices using the weights at which they are included in HICP. This is a calculation method that is generally used, including by Eurostat when it e.g. compiles HICP excluding energy and food to calculate core inflation.
The delineation of energy and food including beverages and tobacco adheres to Eurostat's standard definition of core inflation. There are no equivalent benchmarks for which products are subject to administered prices. Government authorities can choose to decouple the price of e.g. a transport service from costs and market conditions to a greater or lesser extent, and furthermore there can be both administered and non-administered prices within one category, e.g. the rental housing market. A delineation of administered prices does not remain constant over time either. For example, today the telecom market is subject to free price formation, whereas previously there was significant price regulation. The service items rent, public transport, education and childcare institutions have been isolated as products with administered prices.[2]For a number of services within the consumer groups health, recreation and culture, and miscellaneous goods and services (e.g. hunting licences and passport fees), the prices are also subject to a greater or lesser degree of government regulation.[3] However, only price indices with an estimated significant proportion of administered prices have been eliminated.
The sub-indices that are eliminated from HICP are shown in Table 2. Statistics
| FROM TOTAL HICP TO MARKET-DETERMINED HICP |
Table 2
|
| Per cent |
Weights
|
| HICP total |
100.0
|
| Energy: |
10.6
|
| - 04.5 Electricity, gas and other fuels |
6.9
|
| - 07.2.2 Fuels and lubricants for personal transport equipment |
3.7
|
| Food: |
20.7
|
| - 01. Food and non-alcoholic beverages |
15.2
|
| - 02. Alcoholic beverages and tobacco |
5.5
|
| Administered prices: |
11.9
|
| - 04.1-2. Rentals for housing |
7.7
|
| - 07.3.1 Passenger transport by railway |
0.6
|
| - 07.3.2 Passenger transport by road |
0.5
|
| - 07.3.5 Combined passenger transport |
0.2
|
| - 10. Education |
0.9
|
| - 12.4 Social protection |
1.9
|
| HICP excluding energy, food and administered prices |
56.9
|
| Note: The weights reflect the weighting basis as of January 2003. The names of the sub-indices shown refer to COICOP. Combined passenger transport indicates transport services for which the ticket permits combining means of transport, e.g. a change from train to bus. Source: Statistics |
|
In the most recent weighting basis the groups energy, food and administered prices account for around 43 per cent of consumer spending. Chart 1 shows the increase in HICP and in two stripped HICP indices. The increase in HICP excluding energy and food is Eurostat's core inflation. Generally, core inflation is subject to less fluctuation than the overall increase in HICP. Core inflation excluding administered prices is lower than core inflation in the period shown due to relatively high rates of increase in administered prices, including rent, and is also more volatile than core inflation.
| ELIMINATION OF ENERGY, FOOD AND ADMINISTERED PRICES |
Chart 1
|
 |
|
| Source: Statistics
|
|
Elimination of indirect taxes
Like administered prices, the level of indirect taxes (VAT and other excise duties) is not market-determined, and an increase in e.g. VAT cannot in itself be characterised as inflationary. On the contrary, an increase in indirect taxation can dampen demand and price pressure in the economy. There is an indirect contribution to inflation (second-round effect) only if the increase in indirect taxation is transmitted via real-wage claims that are accommodated by business enterprises and passed on to consumers.
The contribution from indirect taxes is eliminated from consumer prices by switching from HICP excluding energy, food and administered prices to the equivalent segment in the index of net retail prices, where the prices are compiled excluding indirect taxes and subsidies. The two indices are classified according to COICOP, so that the content of goods and services in the stripped index of net retail prices corresponds to the content in the stripped HICP. As the indirect taxation rates have remained by and large unchanged in recent years, this adjustment does not significantly affect the development in prices, cf. Chart 2.
| ELIMINATION OF INDIRECT TAXES |
Chart 2
|
 |
|
| Source: Statistics
|
|
Goods and services with domestically determined prices (IMI products) constitute 57 per cent of HICP's weight basis, cf. Table 2, and 51 per cent of the index of net retail prices, cf. Table 3. The difference reflects that the same goods and services do not count equally in the two indices, and does not express the indirect taxation content of the consumer prices. For example, highly taxed cars will account for less in the index of net retail prices, and it is significant that housing, which is stripped out, accounts for more in the index of net retail prices, where the housing weight includes owner-occupied homes. With a weight of 20 per cent (10.3/51.0), the consumption group recreation and culture is greatest in the stripped index of net retail prices, and it can be noted that goods and services have around the same weight in the stripped index of net retail prices.
| WEIGHTS IN TOTAL AND STRIPPED INDEX OF NET RETAIL PRICES |
Table 3
|
|
| Per cent |
Index of
net retail prices |
Stripped
index of net retail prices |
| Total |
100.0
|
51.0
|
| 01. Food and non-alcoholic beverages |
12.6
|
0
|
| 02. Alcoholic beverages and tobacco |
2.5
|
0
|
| 03. Clothing and footwear |
5.2
|
5.2
|
| 04. Housing, water, electricity, gas and other fuels |
31.6
|
3.6
|
| 05. Furnishings, household equipment and routine household maintenance |
6.2
|
6.2
|
| 06. Health |
4.2
|
4.2
|
| 07. Transport |
9.4
|
6.7
|
| 08. Communication |
2.1
|
2.1
|
| 09. Recreation and culture |
10.3
|
10.3
|
| 10. Education |
1.0
|
0
|
| 11. Restaurants and hotels |
5.7
|
5.7
|
| 12. Miscellaneous goods and services |
9.1
|
7.0
|
| Note: The weights reflect the weighting basis as of January 2003. Source: Statistics |
||
Elimination of energy and import content
All that remains in the calculation of IMI is to eliminate the impact of energy prices via the indirect content of energy, i.e. the change in the costs of e.g. shops and means of transport when energy prices fluctuate, and to eliminate the impact of import prices on imported goods. The index of net retail prices does not have separate sub-indices and weights for e.g. import prices, so that the import prices cannot be stripped out of the index, as was the case for e.g. food prices. In order to eliminate the price effect from the energy and import content in the stripped index of net retail prices certain assumptions have to be made and the necessary weights must be set up. In addition, the price indices to be used to calculate the contribution to prices from the energy and import content must be chosen.
The weights are based on input-output calculations with Statistics
The actual energy products are not included in the stripped index of net retail prices, and the input-output calculated coefficients for energy in Table 4 solely represent the indirect energy content that is included in IMI products as input to Danish production.[5] The price indices used for this energy content are the net retail sub-indices for electricity, gas, liquid fuels (fuel oil and petroleum), heat energy and fuels and lubricants for personal transport equipment. These sub-indices are thus stripped out of the net retail prices with a higher weight than the weight at which they are included in the index of net retail prices.
| INPUT-OUTPUT WEIGHTS AND RELATED PRICE INDEX |
Table 4
|
| Per cent |
Weights
|
| Index of net retail prices excluding energy, food, and administered prices |
51.0
|
| Indirect energy content: |
1.0
|
| - 04.5.1 Electricity |
0.3
|
| - 04.5.2 Gas |
0.1
|
| - 04.5.3 Liquid fuels |
0.1
|
| - 04.5.5 Heat energy |
0.2
|
| - 07.2.2 Fuels and lubricants for personal transport equipment |
0.3
|
| Direct and indirect import content excluding energy |
15.2
|
| - Raw materials for other industries (indirect import content) |
9.9
|
| - Other consumer goods (direct import content) |
5.3
|
| IMI index |
34.8
|
| Note: The weights reflect the weighting basis as of January 2003. Source: Statistics |
|
The import content excluding imported energy that is included in the imputed energy content can be divided into two. One part is the direct import content representing processed imported goods, e.g. cars, which via dealers are supplied to private consumers. The other part is the indirect import content that is processed in the Danish production sector before becoming goods or services for consumption. To represent the prices of the import content (excluding energy), two import price indices from the price statistics for the domestic supply of goods (the previous wholesale price statistics) are used. The direct import weight is applied to the import price index for other consumer goods. This index among other things excludes food, alcoholic beverages and tobacco. The indirect import weight is applied to the import price index for raw materials for other industries than agriculture. The indices reflect the price development in the first sales leg, i.e. importers' sales prices excluding VAT and excise duties. Typically some time elapses before a change in import prices is channelled to the retail leg. On this basis the price development in the import price indices is lagged in the calculation.
The price statistics for the domestic supply of goods and the consumer price statistics are not designed to make it possible to follow the same representative item from import leg to retail outlet, and the import price index for raw materials includes a number of commodities that are not used to produce consumer products. There is also a potential problem in the fact that the applied import prices include the importer's sales profit, while the import weights represent the value of the importer's purchases abroad. Generally, a certain degree of uncertainty must be accepted with regard to imputed calculations, and the uncertainty ends up in the residually calculated IMI. However, the result can still be useful, even if it in some months includes fluctuations that are difficult to interpret.
The elimination of the import content reduces the weight basis, so that the IMI index only represents just over 1/3 of consumer spending, of which the largest share comprises services, since import of goods is eliminated. For e.g. cars, in principle only the dealer's profit remains.
The deduction of the indirect energy and import content creates a negative short-term covariation between IMI on the one hand and energy and import price increases on the other. Chart 3 shows the difference between the stripped index of net retail prices and the IMI index. In addition to the fact that IMI fluctuates within a wider band than the rate of increase in the stripped index of net retail prices, IMI is the larger of the two throughout most of the period. This reflects the relatively large content of services in the IMI index and that the price increase for imported goods has been relatively moderate, with the exception of energy and other raw materials.
| ELIMINATION OF INDIRECT ENERGY AND IMPORT CONTENT |
Chart 3
|
 |
|
| Source: Statistics
|
|
INTERPRETATION OF IMI
The deduction of the import price content makes the IMI index a price index for domestic market-determined value added. IMI thus captures the price increases that can be attributed to business enterprises' payroll costs and profits. More specifically, IMI concerns the value added from the manufacture of private consumer goods apart from energy products, food, housing, transport services (train, bus), education (private schools, state-supported independent boarding schools and evening schools) and childcare institutions. The residually determined IMI sector is thus relatively broadly delineated.
The output of IMI products cannot be immediately identified in the national accounts. The IMI index is based on price indices that reflect the development in goods for consumption, while the national accounts' GVA (gross value added) deflators per structure not only reflect goods supplied for consumption, but also the price of deliveries for investment and export. As a rough approximation, the IMI index resembles the GVA deflator for the non-agricultural sector adjusted for services in foreign trade, which includes the volatile sea freight sector. This deflator is included in Danmarks Nationalbank's quarterly economic model, Mona.[6] Chart 4 shows the year-on-year increase in the deflator and IMI. To reduce the short-term volatility in especially the increase in the GVA deflator, annual averages are shown. In consideration of the different construction method and the different content there is good covariation in the development in the two series.
| IMI AND GVA DEFLATOR |
Chart 4
|
 |
|
| Note: Annual averages. The series for IMI is composed of the previous and updated IMI indices. | |
| Source: Mona databank and own calculations. | |
A price index for value added has a different reaction pattern to prices for finished consumer goods and services. There is e.g. greater fluctuation in IMI than in core inflation, cf. Chart 5. The IMI index is increased by price hikes in the shops, but reduced by price hikes for imported input, so that sometimes weak development in IMI indicates forthcoming consumer price increases from abroad. When IMI changes rapidly it is usually profits that have changed, rather than business enterprises' payroll costs. These profit changes can easily be redressed in the ensuing months.
Description of the interaction of the IMI index with other price and payroll indices and with cyclical movements will not be considered further here but may be reviewed in a separate article.
| CONSUMER PRICES AND DOMESTIC PRICES |
Chart 5
|
 |
|
| Source: Statistics
|
|
COMPARISON WITH THE PREVIOUS IMI INDEX
The updating of the IMI index does not significantly affect its development, cf. Chart 6. The weight basis for the various components has also only changed marginally, cf. Table 5.
| OPDATING OF IMI |
Chart 6
|
 |
|
| Source: Own calculations. | |
| UPDATING OF THE WEIGHT BASIS FOR IMI |
Table 5
|
|
| Per cent |
Previous
IMI |
Updated
IMI |
| Index of net retail prices |
100
|
100
|
| Energy |
5.2
|
5.3
|
| Food |
13.4
|
15.2
|
| Administered prices |
29.2
|
28.6
|
| Indirect energy content |
1.2
|
1.0
|
| Indirect import content excluding energy |
7.1
|
9.9
|
| Direct import content excluding energy |
4.7
|
5.3
|
| IMI index |
39.2
|
34.8
|
| Note: The weights reflect the weighting basis as of January 2003. Source: Statistics |
||
The IMI products' total share of consumer spending has fallen from 39.2 per cent to 34.8 per cent. This is partly due to a slightly different composition of the sub-index for energy, food and administered prices, and partly to the updating of the input-output weights. The updated IMI calculation eliminates solid fuel, which is a small item. All energy items have thus been stripped out. In the same way, all food and beverage items are eliminated, including beer and soft drinks, which were included in the previous IMI index. It can be argued that beer and soft drinks, and probably also certain other highly-processed food and beverage items, should be included in the IMI index. However, as described, all food and beverage items have been eliminated, as is the case for one of Eurostat's measures of core inflation, making this intermediate stage of the calculations comparable with the international standard. Previously, all transport services were eliminated from IMI, but today the prices of air tickets, ferry tickets and goods transport are no longer considered to be administered and are included in IMI.[7]
The previous input-output weights were from 1980 and had been adjusted to the current price level, while the new weights are from 2000. The new weights are also adjusted by the development in prices, so that goods subject to relatively high price increases are given a higher weight. The indirect energy content of the IMI products is lower with the new weights, while both the direct and indirect import content are higher. The lower energy weight has made the IMI index slightly less sensitive to the volatile energy prices.
SUMMARY
The IMI index can be seen as a price index for domestic market-determined value added and thus reflects the development in payroll costs and profit margins. It cannot be observed directly whether the IMI index is driven by changes in payroll costs or profit margins, although sudden fluctuations will normally reflect a change in profit margins that can be redressed at a later point in time.
On determining expressions of underlying or core inflation the objective is often to achieve a price increase that is less volatile and thus more indicative than the overall price increase. Eurostat's measures of core inflation also typically fluctuate less than the overall HICP index. The purpose of IMI is not to get a price increase with relatively little fluctuation, but a price increase that represents the domestic market development, and IMI often fluctuates more than the rate of increase in both overall HICP and the index of net retail prices. The index is difficult to understand without familiarity with the concepts used in the national accounts and an awareness of the uncertainty, but does give a different insight on the inflationary process compared to the more traditional expressions of core inflation.
In recent years IMI has become unusually low, cf. the account given in Recent Economic and Monetary Trends. On technical terms this reflects that the ongoing increase in prices for energy and certain other raw materials is not fully transmitted to prices. This may reflect that competition has intensified.
LITERATURE
Report no. 724 (1974). Price development 1971-74, its distribution on foreign and domestic factors, and calculation of the cumulative effect of cost-of-living adjustment, Annex 3 to the sub-committee's report, Report on cost-of-living adjustment (in Danish only).
Christensen, A. M. (1994). Price Increases and Inflation, Danmarks Nationalbank, Monetary Review, February.
Danmarks Nationalbank (1984). External Finance and Domestic Credit, Monetary Review, May.
Danmarks Nationalbank (2003), Mona – a quarterly model of the Danish economy.
Lauritzen, F. (1987). Continued Low Inflation, Samfundsøkonomen, 5/87 (in Danish only).
Sørensen, C. (2003). National Accounts, Balance of Payments and Foreign Trade, 4th edition, Systime Academic (in Danish only).
Thage, B. and A. Thomsen (2004). National Accounts, 6th edition, Handelshøjskolens Forlag (in Danish only).
[1] IMI has previously been described by Lauritzen (1987) and Christensen (1994).
[2] The HICP index for rent solely includes rented housing. The rent item for owner-occupied homes presents problems in international comparisons. The reason is that this is an imputed value compiled by various methods in the national consumer-price indices. Statistics
[3] Reference is made to the twelve main groups in COICOP (the Classification of Individual Consumption According to Purpose), cf. Table 3.
[4] The use of input-output calculations in order to split up the overall price increases into external and domestic factors has a long tradition in
[5] The indirect energy content of the consumption component is calculated for five sectors. These are Extraction of crude oil and natural gas, etc. (input-output code number 110000); Mineral oil industry (230000); Electricity supply (401000); Gas supply (402000); and Heat supply (403000).
[6] For a review of the Mona model reference is made to Danmarks Nationalbank (2003).
[7] The precise difference in the choice of sub-index is illustrated by Table 2. Under energy and food the sub-indices "04.5.4 Solid fuels", "01.2.2 Mineral waters, soft drinks, fruit and vegetable juices" and "02.1.3 Beer" were not previously eliminated. Under administered prices the index "07.3 Transport services" was previously included, while the update only strips out a sub-element of the transport services.