Nordic Food Markets - Appendix 3. Estimation of the promotion effect

Face - Contents - Bottom - Previous/Next

"Nordic Food Markets"

Appendix 3. Estimation of the promotion effect

The EuroStat PPP price indices are based on general prices. It is the Competition Authorities experience that these prices do not at least in some cases take fully into account the impact from short term price cuts. In some countries, characterized by a relatively large promotion activity, the Euro-Stat PPP may therefore overestimate the price level. This appendix describe the approach which the Competition Authorities find most appropriate to estimate the impact from short term price cut, where these are not captured by the Eurostat indices. However, it is important to notice that it is complicated to provide an accurate estimate of the effect of short term price cuts on PPP-price indices.

AC Nielsen has collected data regarding the promotion activity in a number of countries in 2004. Data have been collected on the food products: beer, butter, soft drinks, cold cuts and milk. There has only been very limited data on other food products. The AC Nielsen data show how large a share of the total volume sold that was sold under promotion and how large the average price reductions were.

Using these data (taken into account the limitation of data), the effect of promotion on net prices faced by the consumers has been calculated via the steps outlined in the tables below.

Table A.3.1. Share sold under promotion, percent (O)

Table A.3.2. Weight in food index (S)

Table A.3.3 Average discount (Λ)

Table A.3.4 Net price index 2003, EU15=100 (P)

Table A.3.5 New net price index (Π)

Table A.3.6 Adjusted food price index (Ψ t+1)

Table A.3.7 Stepwise change in net food price index

Method of calculation:

The new net price index in table A.3.5 is calculated using the formula below:

For an example (Denmark, beer):

The adjustment in the overall net food price index in table A.3.6 is gradually calculated using the following formula, where Ψ is continuously changing as adjustments are made for each of the food products:

For example (Denmark, beer):

The change in the overall food price index caused by short term price reductions on beer in Denmark is thus 110.6-100.0 = 0.7 (decimals cause a deviation), c.f. table A.3.7.

Result

After having adjusted the overall food price index for the change in the price index of all 5 food products, the change in the Danish food price index is 1.5 per cent point, c.f. table A.3.7. This expresses the overestimation of the Danish price level, given that no adjustment should be made for all other countries.

Finland, Sweden and Norway take short term price reductions into account when prices are reported to EuroStat. Therefore, the estimated effects of promotion in these countries are irrelevant. However, the effect on the Danish price level must be interpreted in the view of the effect in all other EU15 countries. It seems reasonable to assume that the effect in the countries not included in the survey is approximately the same as in Germany and France, or somewhere between these two countries. France and Germany represent two very different retail structure. Therefore, the net effect on the Danish price level ends up being approximately a good 1 per cent point, which is the extent to which promotion has a larger effect on the Danish price level than the price levels of other countries.

Assumptions and uncertainties

It is important to notice that it is complicated to calculate the impact on price indices from promo-tion, and the approach applied by the Competition Authorities have some methodically weaknesses.

A number of assumptions have been required in order to estimate the effect of promotion as it is done in this appendix. First of all, the gradual adjustments in table A.3.6 are based on an assumption that there are no changes in the price indices in other countries, which would potentially change the EU15=100 point of reference.

Next, the AC Nielsen survey does not include all of the countries belonging to EU15. Therefore, when estimating the effect on the Danish price level to approximately 1 per cent point, it is assumed that the promotion activity and the effect thereof is similar to that of Germany and France in all countries not included in the survey.

There are also a number of uncertainties related to the calculations. First of all, only 5 product groups are included in the AC Nielsen data. However, these are expected to be some of the food products characterized by the highest degree of promotion activity.

Second, the discount in table 3 is based on average prices. Thereby, a situation with negative aver-age discount may arise, for instance when a relatively expensive product like ecologic butter is on sale. Even though it is on sale it may still be more expensive than regular butter and this could cause a negative discount. In the calculations above, negative average discounts have not been taken into account. However, this uncertainty is properly almost the same across the countries which make it likely that they are almost evened out when comparing across the countries.

Version 1.0 December 2005 • © Danish Competition Authority.
Published by the Danish Competition Authority, http://www.ks.dk/
Publication produced according to the standard for electronic publication set by the Government