Skip to main content
Log in

The Elusive Law of One Retail Chain Price

  • Published:
Journal of Industry, Competition and Trade Aims and scope Submit manuscript

Abstract

We investigate whether prices for identical products differ across more than 1600 French supermarkets and find non-trivial price dispersion, albeit more limited than in other countries. We determine that more than 80% of the total variance of the observed dispersion of relative prices across stores and time is explained by the spatial permanent component (stores consistently sell products at relatively high or low prices), essentially driven by persistent heterogeneity in retail chain pricing. The analysis of between and within retail chain price dispersion also provides evidence consistent with a rather centralized multi-stage price setting, in which local stores play a much smaller role than retail buying groups and local branches.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11

Similar content being viewed by others

Data Availability

Price data used in this paper were made available for free by the CEO of Prixing to the author under a non-transferable licence, which prevents from directly sharing the data with other parties.

Notes

  1. While CPI data are not well suited to assessing price dispersion, they made it possible to study price dynamics in the USA and in Europe (see, for instance, Álvarez et al. (2006) and Dhyne et al. (2006)).

  2. The original price data are the same as described in greater detail in Berardi et al. (2017). However, prices examined in that study were weekly prices, whereas in this paper they are daily ones. Also, part of the store data we use were unavailable at that time.

  3. Allain et al. (2017) consistently find that pricing strategies are mostly centralized at the retail chain level in France.

  4. Throughout this paper, retail chain refers to a banner. Several retail chains negotiate with producers together, as a buying group. Internally, a retail chain is organized geographically into local branches. Finally, each outlet is called a store.

  5. The original data refers to more than one hundred thousand products. For more details about the cleaning of the price data, please refer to Berardi et al. (2017) where the same source of data was exploited at the weekly frequency. Note that focusing on the one thousand most widely sold products is likely to proxy weighting by sales. Ghose and Yao (2011) find that using transaction prices, rather than posted prices, results in lower price dispersion. Consistently, we find comparatively lower price dispersion both nationally and internationally.

  6. For more details concerning product categories, brands and manufacturers, refer to Table 15.1 in Berardi et al. (2017).

  7. One year of daily prices constitutes a big data set for evaluating price dispersion. However, it is a short time-series period to assess, for instance, the evolution over time of price dispersion, or the impact on price dispersion of cyclical variations in market conditions.

  8. Medium and large supermarkets represent in France more than 80% of grocery sales (see Anderton et al. (2011)).

  9. These data were collected by Prixing, a start-up company that provides consumers with a free mobile price comparator (http://www.prixing.fr/). The crucial feature is that prices are exactly the same as those of the brick-and-mortar supermarket associated with the “click &collect” service, which is almost always linked to a physical store (although a few standalone drive-throughs, known in French as “drives-entrepôt,” exist. For more details on the characteristics and evolution of “click &collect” (known as “drives” in France), please refer to Berardi et al. (2017). Note that more stores are available for the current analysis, thanks to subsequent data mining.

  10. See, for instance, Cavallo (2017) for a comparison between prices online and in brick-and-mortar stores.

  11. See, for instance, Berardi et al. (2020), Kaplan and Menzio (2015), or DellaVigna and Gentzkow (2019).

  12. Scanner data have a different comparative advantage in that they convey information about purchased quantities. This feature allows for demand models and consumer welfare considerations (e.g., DellaVigna and Gentzkow (2019)).

  13. Unweighted results (available upon request) are not qualitatively different.

  14. Indeed, not all retailers have developed click &collect at the same pace. In particular, one of the major retail groups in France has lagged behind, whereas a smaller player offered this option in most of its supermarkets even in smaller ones. Therefore, our sample of click &collect (although almost exhaustive) did not necessarily provide a representative picture of supermarket sales at the aggregate level.

  15. Retail market shares come from Kantar Worldpanel-LSA and refer to April–May 2012. The share of product categories (COICOP-level4) in the French consumption basket were obtained from the French National Statistical Office (INSEE).

  16. The data was bought from LSA (Libre Service Actualités http://expert.lsa-conso.fr/)

  17. Please refer to Berardi et al. (2017) for more details concerning the geographical distribution and representativeness of supermarkets for which we have price records (see Figs. 15.11 and 15.12) and the market share of the retail chains to which they belong (see Table 15.2).

  18. Instead of defining reference prices over a quarter, it would of course be possible to fix the reference price over a 1-year period, or rather to choose a more flexible reference price based on a month, a week, or even a day. We computed, for comparison purposes, the yearly, monthly, weekly and daily modal price for each item. However, on the one hand, yearly reference prices prove unsuitable because some grocery products are seasonal. Indeed, only 4.8% of items have the same quarterly modal price over the year. On the other hand, monthly, weekly, and daily modal prices would hinder the stability embedded within the concept of reference price, which is meant to abstract from higher frequency price changes like promotions. Consistently, reference prices were also defined over a quarter by Eichenbaum et al. (2011), who popularized the concept within price setting literature.

  19. This definition was adopted in Berardi et al. (2017). It is proposed again in this article to allow a direct comparison with our preferred alternative measure. Note that in this paper, prices are daily (and not weekly as in Berardi et al. (2017)).

  20. In our data, the mean and median price of the most expensive products are 23 euros and for the average product are slightly above 3 euros.

  21. Results (available upon request) are robust whether the product reference price is the mean or the modal price.

  22. For instance, the Limousin and Upper Normandy regions are characterized by average relative prices below 3%, while the average absolute deviation of relative prices from quarterly modal prices at the national level is well above 5%.

  23. However, the comparison with the latter two studies is not straightforward, because they rely only on online prices.

  24. Periodic sales could also be used to discriminate between different customer types. In order to investigate price discrimination, however, household scanner data would be more suitable.

  25. Berardi (2018) presents a visual intuition of the conclusions about temporal price dispersion in France that are reached in a more sophisticated way in this section.

  26. The spirit is similar to several variants of specifications in the literature, decomposing price dispersion into different sources of variation often through fixed effects (e.g., Gorodnichenko et al. (2018), Sheremirov (2019), Kaplan and Menzio (2015), Hitsch et al. (2021)).

  27. Since the data cover 1 year only, longer term changes in temporal price dispersion, whether common to all stores or related to retail chain sales, can’t be captured in this paper.

  28. It thus doesn’t seem necessary to include fixed effects combining days and local branches, which would require having to estimate too many parameters. Berardi et al. (2017) run a similar fixed effect model but on weekly percentage deviations from quarterly mean prices, whereas here price deviations are daily and relative to modal prices.

  29. Model I explains on average 87.2% of observed price dispersion (and 50.7% in the case of the product for which the explanatory power is the lowest) and model II 91.3% (and 62% of price dispersion in the worst case).

  30. For instance, in the UK four main retail chains made a public commitment in 2004 to offer the same prices in all stores.

  31. We define urban density categories based on a combination of population density and absolute population of INSEE “canton-ou-ville.”

  32. Berardi et al. (2017) runs a similar analysis on the spatial component of weekly percentage deviations from quarterly mean prices. However, here it concerns the spatial components of daily price deviations relative to modal prices estimated in model I.

  33. Allain et al. (2017) use the same variables as a proxy for local demand. Our results are robust to adopting a smaller administrative unit than the French “arrondissement” as local market (available upon request).

  34. We follow the definition of competitors adopted by the French Competition Authority (2010) and assume that large supermarkets are only in competition with other large ones, while medium-size supermarkets and discounters are also in competition with large supermarkets. Therefore, not all supermarkets are competing with a given medium-sized competitor, but for all supermarkets we can compute the distance with respect to their closest large-sized competitor. Distances and driving time to the closest competitor are calculated using two internet applications: GoogleMap and YourNavigation. Since the computed distances are similar, in what follows, we only present results based on measures calculated from YourNavigation.

  35. Anania et al. (2012) consistently find that prices tend to be higher in urban areas than rural ones in their analysis of price dispersion for one Italian region. Note that, although prices are higher in cities, groceries exhibit less geographic price variation than other goods (reflecting the fact that grocery items are necessities), as stressed by Diamond and Moretti (2021).

  36. DellaVigna and Gentzkow (2019) also find that prices are positively correlated with income, and Handbury (2021) that they are higher in wealthy cities relative to poor cities. However, he shows that high-income consumers are more than compensated for this price difference by the fact that more of the products they prefer to consume are available to them in these locations. Handbury and Weinstein (2015) also conclude that, once variety differences across locations are accounted for, the price level is actually lower in larger cities.

  37. Not surprisingly, product fixed effects account for most of the variance explained by the model (55%).

  38. The other factors rank as follows, in terms of decreasing explained variance: urban density, number of local stores selling the same item, local per capita income and population, and distance to the closest large supermarket.

  39. Buying groups (“centrale d’achat” in French) are entities created to leverage the purchasing power of a group of retail chains to obtain discounts from producers. In France, at the time the data was collected, the majority of buying groups in the retailing industry would negotiate prices with manufacturers for several retail chains, typically characterized by different size of supermarkets. For instance, two Auchan retail chains (“Auchan hypermarché” that is the banner of very large supermarkets, and “Simply Market supermarché” medium-sized supermarkets) would buy from Coca-Cola at the same price.

  40. However, in some cases the same retail chain has very different distributions of relative prices for Coke cans and Coeur Lion camembert. For instance, stores belonging to retail chains 2 H and 2 H/S sell Coke cans at very similar prices, suggesting national pricing at the retail chain level for that product. This is not the case for Coeur Lion camembert.

  41. Note that dispersion is markedly less than in Fig. 5 where reference prices are computed at the national level.

  42. Note that some retail chains (belonging to retailing groups 4, 5, and 6) have exclusively or mainly independent stores, while the opposite is true for other retailers (retailing groups 1 and 3). In the case of retail chains with both kinds of stores, Table 5 shows that independent stores align less with retail chain modal prices than integrated ones (that is, those belonging to the same owner).

  43. For some details about buying groups in France, please refer to last footnote of section 2.3.

  44. If a retail chain owns a store, then that store is labeled as integrated. Sometimes, however, a store may have the banner of a retail chain but an independent owner. In this case the store is labeled as independent. In France, a few retail chains have exclusively or mainly independent stores, while the opposite is true for other retailers that own all stores exhibiting their banner.

References

  • Allain M-L, Chambolle C, Turolla S, Villas-Boas SB (2017) Retail mergers and food prices: evidence from France. J Ind Econ 65(3):469–509

    Article  Google Scholar 

  • Álvarez LJ, Dhyne E, Hoeberichts M, Kwapil C, Bihan HL, Lunnemann P, Martins F, Sabbatini R, Stahl H, Vermeulen P, Vilmunen J (2006, 04-05). Sticky prices in the Euro area: a summary of new micro-evidence. J Eur Econ Assoc 4(2-3):575–84

  • Anania G, Nistico R (2012) Price dispersion, search costs and consumers and sellers heterogeneity in retail food markets. International Association of Agricultural Economists 2012 Conference

  • Anderton R, Meyler A, Gattini L, Izquierdo M, Jarvis V, Kaarup R, Komzakova M, Landau B, Mohr M, Page A, Sondermann D, Vermeulen P, Cornille D, Tsalinski T, Vladova Z, Hartmann C, Stahl H, Linehan S, Balfoussia C, Panagiotou S, de los Llanos Matea M, Alvarez L, Bardet-Fremann P, Berardi N, Sevestre P, Ciapanna E, Rondinelli C, Kapatais D, Walch E, Lunnemann P, Zerafa S, Pace C, Kieft J, Fritzer F, Cardoso F, Gabrijelcic M, Karma B, Kivisto J (2011) Structural features of distributive trades and their impact on prices in the Euro area. European Central Bank Occasional Paper Series 128

  • Asplund M, Friberg R (2002) Food prices and market structure in Sweden. Scand J Econ 104(4):547–566

    Article  Google Scholar 

  • Berardi N (2018) Is price dispersion a concern for French consumers? Rue de la Banque 54:1–5

    Google Scholar 

  • Berardi N, Gautier E, Le Bihan H (2015) More facts about prices: France before and during the great recession. J Money, Credit, Bank 47:1465–502

    Article  Google Scholar 

  • Berardi N, Ravenna F, Samano M (2020) Everyday regular prices. Banque de France Working papers 746

  • Berardi N, Sevestre P, Thébault J (2017) The econometrics of multi-dimensional panels, chapter on the determinants of consumer price dispersion: evidence from French supermarkets. Springer, pp 427–448

  • Cavallo A (2017) Are online and offline prices similar? Evidence from large multi-channel retailers. Am Econ Rev 107(1):283–303

    Article  Google Scholar 

  • DellaVigna S, Gentzkow M (2019) Uniform pricing in US retail chains. Q J Econ 134(4):2011–84

    Article  Google Scholar 

  • Dhyne E, Álvarez LJ, Bihan HL, Veronese G, Dias D, Hoffmann J, Jonker N, Lunnemann P, Rumler F, Vilmunen J (2006) Price changes in the Euro area and the United States: some facts from individual consumer price data. J Econ Perspect 20(2):171–92

    Article  Google Scholar 

  • Diamond R, Moretti E (2021) Where is standard of living the highest? Local prices and the geography of consumption. Working Paper NBER No. 29533

  • Dubois P, Perrone H (2015, October) Price dispersion and informational frictions: evidence from supermarket purchases. CEPR Discussion Papers 10906, C.E.P.R. Discussion Papers

  • Eichenbaum M, Jaimovich N, Rebelo S (2011) Reference prices, costs and nominal rigidities. Am Econ Rev 101(1):234–62

    Article  Google Scholar 

  • French Competition Authority (2010) La concentration des zones de chalandise

  • Ghose A, Yao Y (2011) Using transaction prices to re-examine price dispersion in electronic markets. Inf Syst Res 22(2):269–288

    Article  Google Scholar 

  • Gorodnichenko Y, Sheremirov V, Talavera O (2018) Price setting in online markets: does IT click? J Eur Econ Assoc 16(6):1764–811

    Article  Google Scholar 

  • Gorodnichenko Y, Talavera O (2017) Price setting in online markets: basic facts, international comparisons, and cross-border integration. Am Econ Rev 107(1):249–82

    Article  Google Scholar 

  • Handbury J (2021) Are poor cities cheap for everyone? Non-homotheticity and the cost of living across US cities. Econometrica 89(6):2679–2715

    Article  Google Scholar 

  • Handbury J, Weinstein DE (2015) Goods prices and availability in cities. Rev Econ Stud 82(1):258–296

    Article  Google Scholar 

  • Hitsch GJ, Hortacsu A, Lin X (2021, 12) Prices and promotions in US retail markets. Quant Mark Econ 19

  • Hong H, Shum M (2006) Using price distributions to estimate search costs. Rand J Econ 37:257–275

    Article  Google Scholar 

  • Kaplan G, Menzio G (2015) The morphology of price dispersion. Int Econ Rev 56(4):1165–206

    Article  Google Scholar 

  • Moraga-González JL, Wildenbeest MR (2008) Maximum likelihood estimation of search costs. Eur Econ Rev 52(5):820–848

    Article  Google Scholar 

  • Sheremirov V (2019) Price dispersion and inflation: new facts and theoretical implications. J Monet Econ 24:59–70

    Google Scholar 

  • Sorensen AT (2000) Equilibrium price dispersion in retail markets for prescription drugs. J Polit Econ 108(4):833–850

    Article  Google Scholar 

  • Woodward SE, Hall RE (2012) Diagnosing consumer confusion and sub-optimal shopping effort: theory and mortgage-market evidence. Am Econ Rev 102(7):3249–3276

    Article  Google Scholar 

Download references

Acknowledgements

The author would like to thank Patrick Sevestre and Jonathan Thébault, coauthors of previous work on which this paper builds on, and Prixing, the start-up that provided the price data used in this paper, and in particular Eric Larchevêque, the owner and CEO of the company at the time data were released. I also wish to thank Aix-Marseille School of Economics and the Banque de France for jointly funding the LSA database. I am extremely grateful to Fernando Alvarez, Huw Dixon, Etienne Gagnon, Yuriy Gorodnichenko, Peter Karadi, Oleksiy Kryvtsov, and Georg Strasser for very helpful comments, as well as participants at the Panel Data, AFSE, and JMA conferences, and Roberto Berardi for his insights into the retail industry. I wish to address special thanks to Sherly Jean-Charles, Sophie Saleh, Marine Tépaut, Sylvie Tarrieu, Alexandre Vigneron, and Yue Zhu for their precious assistance with complementary data collection and treatment. Finally, I would like to thank Vincent Guegan and Victor Yammouni for IT assistance. The views expressed in this article are those of the author and do not necessarily reflect those of her institution.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Nicoletta Berardi.

Ethics declarations

Ethical Approval

N/A.

Informed Consent

N/A.

Conflict of Interest

The author declares no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Berardi, N. The Elusive Law of One Retail Chain Price. J Ind Compet Trade 23, 261–281 (2023). https://doi.org/10.1007/s10842-023-00404-3

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10842-023-00404-3

Keywords

JEL Classification

Navigation