Brand price differentials in retail distribution: product quality and service quality

ABSTRACT

A theoretical model is proposed to disentangle the contribution of brand quality and retailer service quality in explaining brand price differentials across retailers. Two testable hypotheses emerge: (i) for each brand type, price differences across retailers are independent of brand quality differentials and (ii) at a given retailer, price differences between different brand qualities are independent of service quality differentials. Our empirical analysis, for a sample of the U.K. grocery retailer prices, discloses that retailers that offer higher service quality sell same quality brands at higher prices. In particular, service quality premia amount to 6% for national brands and are in the range of 9–15% for low-quality store brands. Besides, at a given retailer, the price premia paid for the national brand are very large: around 150% between national brands and low-quality store brands, and around 40% between national brands and high-quality store brands. Also, the price differential between the national brand and the low-quality store brand does not increase with service quality.

KEYWORDS:

• Store brands
• brand quality
• service quality

JEL CLASSIFICATION:

• C70
• L13
• L20
• J50
• J52

Acknowledgements

We would like to thank the comments and suggestions by a referee, participants at Strategy and Business Economics Division Seminars at the University of British Columbia, participants at Leicester University and participants at the University of Warwick Economics Department Seminars. The authors gratefully acknowledge the financial support from the Spanish Ministry of Economy and Competitiveness and FEDER under the projects ECO2011-25033, ECO2013-45045-R and ECO2013-46550-R, and support from Generalitat Valenciana under the project PROMETEOII/2014/054.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Source: Kantar Worldpanel, market share summary, 12 weeks to 8 November 2015.

² Responses to the consultation on the Office Fair Trade’s proposed decision, particularly from individuals, made clear that access to a range of products of different stores (including independents) is valued by a significant number of consumers.

³ In particular, focusing in the differences between supermarkets chains and discounters, it is common to observe that discounters concentrate on the so called value items, that is, low price and simple packaged goods. They offer less products and less service quality than the supermarket chains. Davies and Brito (2004) report that many discounters operate with as few as four full time equivalent staff per store compared with the 200 or more in a superstore. And also that a supermarket chain may offer 20,000 product lines while a discounter only 1,500.

⁴ Testing the effect of quality on price differentials in the banana market, Wiggins and Raboy (1996) find that price premia associated with containerization (as a quality attribute) are quite robust and explain the bulk of inter-company price differences, whereas brand name does not. Also, see Shepard (1991) for the gasoline retail market in Eastern Massachusetts. She finds price differentials consistent with the price discrimination hypothesis and inconsistent with cost-driven differentials. On average, the price differential at multiproduct stations is 9–11 cents higher than the differential across single-product stations. The higher differential comes largely from higher full-service prices. Finally, Delgado and Waterson (2003) find price differentials among retailers for tyres, basically a homogeneous product, on the basis of the scope of vertical integration.

⁵ When a firm belongs to a superior form of organization, its competitive advantage is explained not only by its own value to the chain but also by how it is within the organization. The relevant actor is the full organization not each member (see Porter (1985)).

⁶ It is assumed that own effects dominate cross effects.

⁷ The unit costs can be interpreted as wholesale prices paid by retailers to manufacturer producers. Then, linear contracts are implicitly assumed. Non-linear contracts could be introduced in the model. The analysis of contracts is beyond the scope of the present formulation; yet the properties of equilibrium prices we wish to note are not affected by this. For the sake of the exposition, any other type of costs are disregarded.

⁸ Evidence provided in Choi and Coughlan (2006) indicates that consumers routinely buy and hold both national brands and private labels in quite a number of product categories. Such preference for variety is captured by the proposed utility-maximizing framework.

⁹ A model with product line competition between sellers that considers these type of competition can be found in Dobson and Waterson (1996); yet their analysis does not include differences in brand quality.

¹⁰ The same sort of analysis can be carried out with five products. Retailer 1 carries three products and retailer 2 only two. One way to interpret such asymmetric product lines is that retailer 1 has two tiers of store brands, one of a higher quality than another, in addition to selling a national brand. Retailer 2 carries a national brand plus its store brand in that product category. We have to adapt the utility function to account for five products and the corresponding own and cross-effects. The same qualitative conclusions are obtained which confirms the findings with three products. Computations are available upon request.

¹¹ Cannon Park Shopping Centre included a few small shops, a pharmacy from a national chain, Tesco and Kwik Save.

¹² This two-tier store brand strategy is not unique to U.K. supermarkets, Steiner (2004) also identifies it for Wal-Mart in the U.S.

¹³ By construction ${\beta }_0$, the mean relative price of Kwik Save low-quality store-brands is equal to 1 as the reference price is always the price of Kwik Save low-quality store brand.

¹⁴ For a matter of completeness, we test for differences between Tesco and Sainsbury’s price premia both for national brands and low-quality store brands. Since we have assumed that both supermarkets offer similar service quality, we expect these premia to be non-significant. Thus, we test whether ${\hat{\mathrm{\Delta }}}_J^{S-T}$ for $j=N,L$ is significantly different from zero. It is shown in Table 2 that ${\hat{\mathrm{\Delta }}}_N^{S-T}$ and ${\hat{\mathrm{\Delta }}}_L^{S-T}$ are not significant at a conventional 5% level.

¹⁵ Dhar and Hoch (1997) report a price advantage for own brands that ranged from 24% to 44%. Also Davies and Brito (2004) find on average a 42.6% price premium paid for national brands when compared with the store brands sold by supermarkets. The source of this price premium being the internal costs of brand manufactures, in particular, marketing costs (i.e. consumer advertising and trade marketing), employment costs and R and D expenses.

Additional information

Funding

This work was supported by the Spanish Ministry of Economy and Competitiveness and FEDER under the projects [ECO2011-25033, ECO2013-45045-R and ECO2013-46550-R], and Generalitat Valenciana under the project [PROMETEOII/2014/054].