The growing use of the Internet and advances in information technologies enable firms to gather unprecedented volumes of consumer data. This has led to important changes in their pricing policies. For example, prices are often set by software algorithms instead of humans (Calvano et al. 2019) and firms can practice price discrimination at fine-tuned levels (Bounie et al. 2021). These possibilities increasingly raise concerns for competition policy. While the concern of algorithmic pricing often centres on whether algorithms can achieve collusive market outcomes in a dynamic environment (Ezrachi and Strucke 2017, Harrington 2018, Calvano et al. 2020), fine-tuned price discrimination can lead to a loss in consumer welfare even in a static environment, by allowing firms to extract a large part of consumer surplus (Varian 1989, Stole 2007, Waldfogel and Shiller 2007).
This concern, in conjunction with privacy issues has, for instance, led to the adoption of the General Data Protection Regulation (GDPR) in the EU in 2016 to protect consumers.
This column considers an explicit form of price discrimination, namely, personalised pricing. Personalised pricing refers to the practice of charging individual consumers (or small groups of consumers) prices that are based on their personal characteristics.
Although examples where firms use this practice remain fairly limited, the ability to do so is present in many industries, as was recently documented by Dubé and Misra (2021). In monopolistic markets, personalised pricing leads to a reduction in consumer surplus due to higher rent extraction. The impact is however less clear in competitive environments. For instance, in a highly influential paper, Thisse and Vives (1998) show that personalised pricing leads to fiercer competition compared to uniform pricing, as it enables firms to compete at the individual consumer level. This and several related articles have led to the view that personalised pricing benefits consumers in competitive markets.
In this column, we argue – based on Jullien et al. (2022) – that the trend in data collection has consequences not only for pricing, but also for long-term strategic decisions such as the choice of distribution channels used to reach out to consumers and vertical contracting. This is a particularly important issue in the digital age, as technological advances have led to the emergence of new online retail platforms. Whether to rely on these independent retailers or only on direct channels is a key question for manufacturers.
We are foremost interested in two questions: (1) Does personalised pricing change a manufacturer's incentives to distribute its products through an independent retailer? (2) What are the consequences of personalised pricing for consumer welfare in a context of intra-brand, rather than inter-brand, competition.
To answer these questions, we consider a simple setting in which a brand manufacturer sells directly to consumers and can also rely on an independent retailer. The retailer adds value to the industry but also competes with the manufacturer in the retail market. Contrary to the case of inter-brand competition, however, this intra-brand competition can be partly mitigated through a wholesale contract.
We find that the answers to the questions above depend on demand patterns, and particularly on the correlation between consumers' valuations across the two channels. Figure 1, Panel A illustrates the case where consumers who have a high valuation for one channel have instead a low valuation for the other channel, a pattern usually associated with horizontal differentiation between firms. Figure 1, Panel B illustrates instead the case where the same consumers have high – or low – valuations for the two channels, a pattern usually associated with vertical differentiation between firms (the figure further focuses on the case where high-end consumers value the manufacturer’s channel more than the retailer’s channel).
Figure 1 Correlation between manufacturer’s channel versus retailer’s channel valuations
When consumers' valuations are negatively correlated, as in the classic case of horizontal differentiation, selling through both channels (i.e. dual distribution) is always optimal, regardless of whether or not firms can set personalised retail prices. This is because the wholesale price paid by the independent retailer to the manufacturer provides a very effective tool for limiting intra-brand competition. In particular, a wholesale price equal to the willingness-to-pay of the consumer indifferent between the two channels enables the firms to fully segment the market. No firm can then profitably serve the other's core market (i.e. the market at which the other firm has a competitive advantage).
However, although personalised pricing does not affect the optimal distribution strategy, it has a negative effect on consumer surplus. Since the wholesale price can be used to segment the market, the firms no longer compete against each other. With personalised prices, this segmentation enables them to extract the entire consumer surplus, as in the case of monopoly.
These findings stress the importance of distinguishing between inter-brand competition and intra-brand competition. In the former case, firms are trapped in a prisoner's dilemma in which personalised pricing reduces industry profits and increases consumer surplus. By contrast, in the latter case, the wholesale contract can be used to limit the intensity of competition, which implies that the insights about the impact of personalised pricing are markedly different from those obtained for inter-brand competition. In the case of intra-brand competition and negative correlation of valuations, personalised pricing, together with an appropriate wholesale tariff, allows for full consumer surplus extraction. The competitive consequences of personalised pricing can thus be fully reversed when firms are no longer independent but connected through a wholesale contract.
When instead consumers' valuations for the two channels are positively correlated, as in the case of vertical differentiation, dual distribution remains optimal if firms compete in uniform prices. Setting an appropriate high wholesale price then suffices again to limit the intensity of intra-brand competition, while allowing the retailer to expand demand.
By contrast, under personalised pricing, mono distribution (i.e. selling only through the direct channel) may become the optimal strategy. Specifically, relying exclusively on direct distribution is optimal when the retailer does not substantially expand demand, as the effect of increased intra-brand competition then prevails. This is because, when charging personalised prices, the firms can price aggressively in each other's core segment without sacrificing margins in their own core business. As a result, it becomes difficult to control intra-brand competition without impeding market expansion.
The competitive consequence is again detrimental to consumers. Whenever personalised pricing leads to mono distribution, it deprives consumers from access to the product through the retailer’s channel. In addition, the absence of intra-brand competition enables the manufacturer to extract the full surplus, leaving no welfare to consumers.
These results show that wholesale contracting and the possibility of personalised pricing are crucial not only for determining the optimal distribution strategy, but also for the validity of the common view that personalised pricing tends to increase consumers surplus in competitive environments.
Bounie, D, A Dubus and P Waelbroeck (2021), “Selling Strategic Information in Digital Competitive Markets”, RAND Journal of Economics 52: 283-313.
Calvano, E, G Calzolari, V Denicolò and S Pastorello (2019), “Artificial Intelligence, Algorithmic Pricing, and Collusion”, VoxEU.org, 3 February.
Calvano, E, G Calzolari, V Denicolò and S Pastorello (2020), “Artificial Intelligence, Algorithmic Pricing, and Collusion”, American Economic Review 110: 3267-3297.
Dubé, J-P and S Misra (2021), “Personalized Pricing and Consumer Welfare”, Journal of Political Economy, forthcoming.
Ezrachi, A and M E Stucke (2017), “Artificial Intelligence & Collusion: When Computers Inhibit Competition”, University of Illinois Law Review, 1775-1810.
Harrington, J E, Jr (2018), “Developing Competition Law for Collusion by Autonomous Price-Setting Agents”, Journal of Competition Law & Economics 14: 331-363.
Herweg, F and D Müller (2014), “Price Discrimination in Input Markets: Quantity Discounts and Private Information”, Economic Journal 124: 776–804.
Herweg, F and D Müller (2015), “Price Discrimination by Big Manufacturers: Supporting Arguments for its Prohibition by European Competition Law”, VoxEU.org, 7 April.
Jullien, B, M Reisinger and P Rey (2022), “Personalized Pricing and Distribution Strategies”, Management Science, forthcoming.
Katz, M L (1987), “The Welfare Effects of Third-Degree Price Discrimination in Intermediate Good Markets”, American Economic Review 77: 154-167.
OECD (2022), Handbook of Competition in the Digital Age.
Rhodes, A and J Zhou (2022), “Personalized Pricing and Competition”, TSE Working Paper 22-1333.
Stole, L A (2007), “Price Discrimination and Competition”, In M Armstrong and R H Porter (eds.), Handbook of Industrial Organization III, North-Holland: Elsevier, 2221-2299.
Thisse, J-F and X Vives (1988), “On the Strategic Choice of Spatial Price Policy”, American Economic Review 78: 122-137.
Varian, H R (1989), “Price Discrimination”, In R Schmalensee and R D Willig (eds.), Handbook of Industrial Organization I, North-Holland: Elsevier, 597-654.
Waldfogel, J and B Shiller (2007), “Music for a Song: Can the Music Industry Price its Way out of Piracy Woes?”, VoxEU.org, 25 November.