A study by Frontier Economics, commissioned by METRO AG, suggests that participating in a buying cooperation led to lower resale prices for customers and end-consumers
Buying cooperations between retailers and/or wholesalers, especially with regard to the procurement of branded products have become increasingly common in the food product value chain allowing retailers and wholesalers to create a counterbalance to the strong bargaining power of major suppliers of the food industry. While there has been and still is a controversial, partially emotional discussion in Europe regarding their potential impact and benefit for customers, empirical evidence for their effects are limited. METRO AG commissioned a study to contribute to the debate by providing a quantitative analysis of the price effects of buying cooperations using actual price data, spanning several years.
1.1 Retailers and wholesalers sit in the middle of the grocery product value chain. Upstream, they negotiate with suppliers and fix the purchasing conditions for products that are eventually sold on to their customers. Downstream, they compete with other retailers and wholesalers to attract customers to their stores and websites.
1.2 Purchasing conditions are an essential input for the value that retailers and wholesalers can offer their customers. They govern how products are marketed, when and where discounts can or have to be applied and, above all, the price that a reseller (i.e. a retailer or wholesaler) can sustainably charge to its customers.
1.3 Downstream markets are characterised by intense competition between rival retailers (or between rival wholesalers) with their customers constantly seeking the “best deal” of product choice and value for money. As a result, any improvements in purchasing conditions are typically shared with customers as retailers and wholesalers use improved terms to gain a competitive advantage over their rivals.
1.4 Buying groups affect this supply chain by concentrating the power of retailers and wholesalers in the procurement market: strengthening their negotiating power versus suppliers. Below we summarise why:
a. economic theory predicts that this should lead to lower prices for customers and, ultimately, end consumers in all but exceptional circumstances (circumstances which are already proscribed by EU competition law); and
b. the new empirical evidence put forward in this report supports this prediction, showing that Metro’s participation in a buying group in France has lowered prices for customers by around 5%-7%. This is in line with another recent empirical study of buying groups in France (Molina 2019), which also finds a price reduction of 5%-7%.
Economic theory suggests that buying groups lower prices for end-consumers
1.5 The first order effect of a buying group is to lower end prices for customers. This is because:
a. buying groups only concentrate reseller power in procurement – this doesn’t reduce competitive tension in the negotiation (in fact, it increases the pressure on suppliers to offer the best terms) and it doesn’t reduce competition tension between resellers, who continue to compete against each other downstream;
b. retailers’ and wholesalers’ incentives to use their buyer power in procurement are fundamentally aligned with the interests of customers and end consumers – who also want retailers and wholesalers to obtain the lowest input prices; and
c. because buying groups do not concentrate competition in the downstream retail and wholesale markets, retailers and wholesalers who gain a stronger bargaining position continue to have strong incentives to share terms improvements with their customers.
1.6 It is therefore fundamentally misleading to characterise buying groups as similar in their effects to other forms of horizontal concentration – the economic effects are fundamentally different. They are pro-customer and pro-consumer by default.
1.7 There are, in theory, circumstances where buying groups can be harmful to competition in the upstream or downstream markets. This arises when the concentration of buyer power is so great as to give buying group members the power to exclude competition – either by foreclosing other market participants, or by facilitating coordination.
1.8 However, such exceptional circumstances already sit fully within the ambit of existing EU competition law provisions – designed to prohibit exactly these types of horizontal concentration. There can be no credible suggestion of an enforcement gap in EU competition law, precisely because buying groups remain pro customer and pro-consumer at levels of horizontal concentration significantly above the threshold at which EU competition law would already intervene in other circumstances.
The empirical evidence suggests buying groups lower prices for customers and end consumers
1.9 This report presents the results of a detailed quantitative analysis of the price effects of buying group participation. Our study uses a large dataset of article-level monthly selling price data provided by METRO AG (hereafter ‘Metro’) for six EU countries over the period from 2013-2018, of which ultimately four countries were used for the econometric analysis
1.10 This data enabled us to identify Metro’s buying group in France as suitable for econometric regression analysis – enabling us to make statistically robust pricing comparisons that control for product mix effects and general inflation thereby isolating the resale price effect of buying group participation.
1.11 Our analysis shows that Metro’s participation in the buying group in France led to lower selling prices for its customers of around 5%-7%, and that:
- The results are robust to a number of sensitivity tests such as variations in the econometric specification, time period covered or when using a dataset with an alternative SKU identification.
- The effect broadly holds in the within-country analysis, when the data satisfies the more stringent criteria under a smaller sample size.
- The fall in prices was found to also apply in most product categories covered by buying group participation within our sample.