In situations where competition could be unfair or consumer choice may be affected, the Competition Authorities in EU member states are responsible for investigating mergers. In cases where a merger might restrict competition in the EU single market and involves firms with significant turnover, the EU Commission may investigate the merger rather the national Competition Authority. These investigations, whether at national or EU level, take the form of a preliminary phase 1 investigation, following by a more in-depth phase 2 investigations if concerns remain following phase 1.
In reviewing proposed mergers and acquisitions between competitor firms, the key concern of competition authorities is whether the transaction is likely to reduce competition and thereby cause consumer harm. A review of the impact of the proposed merger on prices is a key part of the analysis. Economists use a variety of analytical tools to evaluate the likelihood of post-merger price increases, however, traditional tools do not consider the relationship between historic and future consumer behaviour and may therefore overstate potential price rises from mergers.
Analysing the price impact of mergers
Market concentration ratios, such as the Herfindahl-Hirschman Index (HHI), are often the starting point for the analysis, particularly during the initial phase of a merger review. Whilst simple to calculate, and sufficient for an initial consideration, they suffer from an inability to capture different degrees of competition amongst differentiated products.
In more controversial / complicated assessments (typically those that trigger a second phase of investigation), competition authorities usually look to the development of merger simulation models. In differentiated product markets, the go-to framework for the economic modelling of price effects from proposed mergers is the differentiated products demand system, building on the contribution of (Berry, Levinsohn, and Pakes 1995), popularised by (Nevo 2000). These models allow for flexible substitution patterns across goods by consumers in response to price changes. They incorporate strategies to deal with statistical issues arising from the fact that in real-world data, prices are driven by a number of potentially unobserved factors that can bias model estimates, such as changes in consumer taste over time.
Limitations of the standard techniques
Standard differentiated product models are broadly static with respect to consumer behaviour, such that individual’s decisions only depend on information about product characteristics and prices at the time of purchase, and effectively ignore the influence of prior purchases. While significantly simplifying the model, this assumption can be problematic. Issues that are effectively ignored could include, for example, shopping habits which become ingrained over time, the hassle involved in finding and evaluating available alternatives, or costs (either monetary or in terms of time and effort) of switching to a competitor. Ignoring such behaviours means that analysis may be overstating the potential for consumers to switch to an alternative firm following a price rise.
Consumer inertia in switching between firms therefore affects the optimal pricing decision, as it requires a firm to take into account the impact of today’s pricing decisions on its future customer base. The overall effect of switching costs on market prices depends on a number of other market characteristics, however, both theory and empirical results point towards switching costs exerting a dampening effect on competition, raising market prices over and above the competitive outcome that would result in the absence of switching costs.
The microeconomic literature has long recognised the importance of this consumer inertia in consumer choices for the pricing decisions of firms. (Klemperer 1995) It provides a survey of the theoretical models investigating the effects of this persistence, which more generally can be seen as a consequence of “switching costs” incurred by consumers choosing repeatedly between competitor products. Intuitively, if a firm faces demand from consumers who find it costly to switch to competitors, it gains the ability to raise prices without losing its customers. Such a price rise would be expected to raise profitability in this instance. Further, the increase in prices will potentially deter new customers entering the market, i.e. those who have not yet developed an attachment to a certain competitor.
Adapting standard models to take into account dynamic consumer behaviour
In a recent working paper, (MacKay and Remer 2019) the example of the U.S. gasoline market is used to demonstrate that a model without consumer inertia predicts price increases, following a hypothetical horizontal merger that is significantly higher than those predicted by a model that explicitly accounts for the formation of consumer habits in shopping for gasoline.
In addition to consumer inertia and switching costs, other features of markets or products can render the application of a static modelling approach inappropriate. Recent contributions in the academic literature have considered storable goods and consumer stockpiling (Hendel and Nevo 2013) or durable goods with forward-looking purchasing behaviour by consumers (Gowrisankaran and Rysman 2012). Again, optimal pricing behaviour of firms – and therefore the pricing response of a newly merged entity post-merger – is shown to be crucially dependent on past behaviour and future expectations of consumers.
These results show that it is crucial to look beyond the “standard” demand modelling framework and take account of the empirical realities of consumer behaviour to robustly estimate the expected effects of mergers on consumer welfare. In the majority of mergers involving B2C businesses, there is likely to be some degree of dynamic consumer behaviour, as it would perhaps be rare to find a purchasing decision by existing customers that it is not influenced by previous decisions.
In constructing market models to simulate merger effects, careful consideration should therefore be given to empirical evidence on consumer inertia – either from customer-level data on purchase histories, or from aggregate data on firm’s pricing decisions over time.
Models of the type utilised by MacKay and Romer provide a comprehensive framework for estimating demand for differentiated goods that account for dynamic consumer behaviour. These can minimise the overestimation of price effects that occur with traditional merger simulation models and may results in mergers being cleared by competition authorities that otherwise would have been.
Berry, Steven, James Levinsohn, and Ariel Pakes. 1995. ‘Automobile Prices in Market Equilibrium’. Econometrica 63 (4): 841. https://doi.org/10.2307/2171802.
Gowrisankaran, Gautam, and Marc Rysman. 2012. ‘Dynamics of Consumer Demand for New Durable Goods’. Journal of Political Economy 120 (6): 1173–1219. https://doi.org/10.1086/669540.
Hendel, Igal, and Aviv Nevo. 2013. ‘Intertemporal Price Discrimination in Storable Goods Markets’. American Economic Review 103 (7): 2722–51. https://doi.org/10.1257/aer.103.7.2722.
Klemperer, P. 1995. ‘Competition When Consumers Have Switching Costs: An Overview with Applications to Industrial Organization, Macroeconomics, and International Trade’. The Review of Economic Studies 62 (4): 515–39. https://doi.org/10.2307/2298075.
MacKay, Alexander, and Marc Remer. 2019. ‘Consumer Inertia and Market Power’. SSRN Scholarly Paper ID 3380390. Rochester, NY: Social Science Research Network. https://papers.ssrn.com/abstract=3380390.
Nevo, Aviv. 2000. ‘A Practitioner’s Guide to Estimation of Random-Coefficients Logit Models of Demand’. Journal of Economics & Management Strategy 9 (4): 513–48. https://doi.org/10.1111/j.1430-9134.2000.00513.x.