Many antitrust theories of harm relate to price effects. Governments regulate retail prices, for example, to protect smaller shops from more cost-efficient rivals. But for online shopping, shipping costs must also be taken into account. Looking at Amazon, this column quantifies the role of shipping costs in the effective prices for consumers when shopping online. Larger retailers may enjoy economies of scale in shipping costs and outperform smaller sellers, even if the latter offer lower prices before adding shipping. Thus, shipping costs must be considered when measuring or remedying competitive effects for online retail.
Prices of goods are a central variable in measuring the degree of competition and gauging consumer welfare. Many antitrust theories of harm relate to price effects. For example, the US Federal Trade Commission recently alleged that Amazon effectively ties the use of the company’s fulfilment services to the use of its Marketplace platform and, in addition, punishes sellers on Amazon Marketplace when they offer lower prices for their goods elsewhere (Federal Trade Commission 2024). This is said to result in an overall increase in price.
Another prime example of retail price regulation is the German Book Resale Price Maintenance Act (Buchpreisbindungsgesetz, henceforth ‘book price law’), which obliges publishers in Germany to set a fixed retail price for new books in order to protect smaller bookshops from price competition by more cost-efficient rivals and to support the supply of commercially less successful niche titles. According to the German Federal Cartel Office (Bundeskartellamt), this book price law helps small bookshops prevail in competition against larger, mostly digital, retail competitors, which is supposed to bear on the assessment of mergers between larger book-retail chains. 1 Shipping costs, however, play no role in the book price law, even though they can affect the total price faced by consumers. 2
In our paper (Gado et al. 2023), we advance the hypothesis and bring forth evidence that a ‘price’ of a good in a digital retail environment is more complex than usually understood in traditional competition or price regulation. In an online retail environment, the price tag becomes less important because it is overshadowed by the costs of shipping; 3 what ultimately matters for the consumer is the effective price inclusive of shipping costs. Including shipping costs is particularly important because these shipping costs may relate to scale economies in the logistics of shipping that only some (likely large) competitors enjoy but not others (likely small).
Through the lens of Amazon prices, our study quantifies the role of shipping costs in the effective prices that different consumers face when shopping in a digital retail environment. 4 We find that taking shipping costs into account is crucial in determining whether Amazon is cheaper or more expensive than its competitors, whether on its own platform or in the universe of online competitors.
To establish these results, our analysis creates a novel dataset that includes prices of products that were Amazon bestsellers and of identical counterparts from online sellers on Amazon Marketplace and from sellers outside the Amazon platform, as represented by data from the leading German price-comparison website Idealo. Our analysis centres on Germany as Amazon’s second-largest market, with $34 billion of revenue versus $356 billion in the US (Coppola (2023).
In a first step, we compile lists of Amazon bestselling products for all 24 bestselling categories using data from Helium 10. In a second step, we collect price-comparison data for the Amazon bestsellers from Idealo, made available by PriceAPI. 5 We recorded prices both including and excluding shipping costs. Our baseline comparison consists of 1,726 distinct products, all of which were new and not used.
To expand the scope of our analysis, we also considered prices of 2,172 bestsellers in the US rather than Germany and their identical Amazon Marketplace counterparts. 6 We likewise analysed data of 1,889 former, rather than current, bestsellers in Germany to assess the specificity of our results to holding bestseller status.
Key findings
Amazon’s prices relative to its competitors are widely dispersed; sometimes Amazon is more expensive, sometimes less. Crucially, the sign of the average percentage price difference tends to reverse depending on whether shipping costs are included or excluded in prices. 7 These findings hold true regardless of whether we compare Amazon’s prices to all of its competitors or only to competitors selling through Amazon Marketplace.
Notably, when shipping costs are included in both Amazon and competitor prices, Amazon is on average less expensive. Comparing only prices of Amazon Marketplace sellers and outside online competitors, moreover, the median percentage difference is zero and substantially less dispersed.
Regardless of whether sellers pay shipping costs for both sources or neither source, there is now substantially more mass at zero relative to the other comparisons. At the same time, when shipping costs are included for both sources, the mass of products with no price difference shrinks. Figure 1 illustrates the role of shipping costs in the comparison of Amazon to all of its competitors.
Figure 1 Distribution of the price difference between Amazon and competitors (Amazon Marketplace and outside)
Notes: The distribution of the price difference between Amazon and Amazon Marketplace and outside competitors, expressed as a percentage. The figures differ in their treatment of shipping costs. The price differences are relative to the competitor with the lowest price. A value of 50 corresponds to a 50% higher price by Amazon relative to the cheapest competitor. The spike at 100 is due to binning price differences over 100% into that point for representation purposes.
In its totality, these findings suggest that shipping costs can make up an important factor for consumers when assessing the benefits derived from a purchase. For example, when a large retailer enjoys economies of scale in shipping costs, it can pass the associated lower costs through to consumers and outperform rival sellers, even if the rivals have lower price tags before adding shipping costs.
An extension of our analysis confirms our findings in a broader context. A first extension shows current and former German bestsellers on Amazon have a median competitor relative price difference that is the same, as long as all sellers are subject to shipping (or no shipping) costs. While the sample used is limited, this finding suggests that our results are not necessarily subject to bestseller status.
A second extension shows a similar pattern with respect to shipping costs in the US data that are available to us. Using data on US Amazon bestsellers sold by Amazon and Amazon Marketplace sellers only, we obtain the same results as for the German data: inclusion of shipping costs is crucial for determining price differentials. However, while the relevant distributions have similar shapes, more relative prices have a 0-price difference in the US sample. Depending on the inclusion of shipping costs, this share is between 30% to 40% rather than 15% to 20% in the German data, in line with the findings of Cavallo (2017 2018).
Policy considerations
Our findings lead us to several policy considerations. First, it is essential to assess the importance of shipping costs when measuring or remedying competitive effects. Finding supra-competitively high product prices is not necessarily evidence of market power, as total prices that include shipping costs may be lower than the competition. These considerations gain importance in a digital retail environment where economies of scale enjoyed only by a few large retailers can strongly impact effective prices through lower shipping costs.
Second, allegations about tying and punishment schemes with respect to the use of shipping services and pricing implemented on platforms should be checked against actual pricing data for full counterfactual analysis. Competition from outside the platform might be sufficiently strong to counter a platform operator’s attempts to establish a supra-competitive pricing floor (see Ennis et al. 2000 on price parity in the hotel booking context, and Crawford et al. 2022 on the effects of Amazon’s entry into competition with third-party sellers).
Third, retail price regulation, such as the German book price law, might fall short of the intended effects if confined to product prices absent shipping costs, as is currently the case. The intended protection of smaller businesses from price competition can be undermined if larger rivals can effectively lower total consumer prices by offering cheaper or free shipping.
Fourth, the role of shipping costs most likely negatively correlates with the product price. In a low-price segment of products, shipping costs can become as relevant or even more relevant than the naked product price.
Fifth, consumers benefit from lower effective prices due to lower shipping costs (compared to buying at the same price tag but adding higher shipping costs). This may seem to be a trivial point at first sight, but it illustrates the intricacy it poses to competition policy: large platforms may use cost advantages of proprietary shipping services to outperform rivals in product sales. Yet, such efficiency-based gains in market share should not be confused with the imposition of anticompetitive entry barriers. Otherwise, antitrust law and other types of market regulation, such as the book price law, would punish efficiency in the shipping service sector despite its beneficial effects on consumers’ rents (on measuring the impact of online platforms on efficiency and productivity, see Sorbe et al. 2019. For further deliberations on the burden of proof for objective justifications in abuse of dominance cases in digital platform environments from a competition policy standpoint, see Kaeseberg 2019).
Sixth, the role of shipping costs highlights the importance of competition in shipping logistics. Interoperability of sales platforms with regards to different types of shipping services may foster competition related to shipping cost and should be a consideration for competition regulation (on the role of contestability for unleashing the efficiency potential of platforms, see Sorbe et al. 2019).
Ultimately, our analysis underscores the necessity of interpreting price data through the lens of a competitive model. A competitive counterfactual is needed to evaluate theories of harm accurately, as well as the effects of any regulatory interventions. Such a comprehensive model-based exercise would navigate the intricate relationship between Amazon’s strategies, pricing dynamics, and competitive implications while revealing the nuanced balance between static and dynamic efficiencies that may feature in antitrust considerations. Our data provide a starting point for further investigations in this area.
Source : VOXeu
