By Davide Del Prete[1]
This paper studies the distribution and consequences of market power along the global value chain (GVC) of coffee. Motivated by a growing academic and policy debate on market power and concentration, the analysis addresses two central research questions: (i) whether market power differs systematically across stages of a global value chain, and (ii) whether there is a trade-off between local (stage-specific) and global market power when firms operate across countries and stages.
The coffee sector provides an ideal setting for this analysis. It involves more than 12.5 million producers worldwide and approximately one billion consumers, while international prices are highly sensitive to short-run supply and demand conditions. At the same time, the sector has experienced substantial consolidation, particularly among international traders and downstream firms, with more than 70 major mergers and acquisitions (M&A) since 2014. These features make coffee a natural laboratory to study how concentration and firm size shape prices and market outcomes along a GVC.
This paper constructs a novel and comprehensive global dataset that traces coffee transactions across all stages of the value chain: upstream production, midstream international trade, and downstream retail. Upstream data include intra-national flows for Brazil and Colombia from Transparency for Sustainable Economies (TRASE), complemented by farm-gate price data from the International Coffee Organization (ICO). Midstream data consist of transaction-level customs records with identified exporters and importers for 14 major producing countries, covering roughly 87 percent of global production between 2014 and 2020. Downstream data are drawn from Euromonitor’s Global Market Information Dataset (GMID), which reports brand-level retail sales and volumes across about 100 countries from 2013 to 2022. In addition, the paper assembles a detailed dataset on 314 M&A deals in the coffee sector since 1974, allowing to study consolidation dynamics explicitly.
The descriptive analysis uncovers three main stylized facts. First, there is substantial price dispersion at all stages of the coffee GVC. Even after controlling for origin, destination, year, and product characteristics, prices vary significantly within narrowly defined markets, suggesting the presence of imperfect competition rather than purely competitive pricing. Second, the coffee GVC is highly concentrated, particularly at the midstream and downstream stages. The top 100 importers—approximately 1 percent of all firms—account for more than 90 percent of global coffee purchases. Third, the growth of large importing firms has occurred primarily through M&As, highlighting consolidation as a key driver of increasing market power.
The empirical analysis formalizes the relationship between market shares and prices at each stage of the value chain. Using regressions of log prices on firm-level market shares and using a rich sets of fixed effects, we document stark differences across stages. Upstream, higher exporter or importer market shares are associated with lower prices paid to producers, consistent with buyer market power. In the midstream, bilateral market shares matter for prices in asymmetric ways: exporters with larger shares tend to obtain higher prices, while importers with larger shares pay lower prices, reflecting bilateral bargaining power in international transactions. Downstream, higher retailer market shares are strongly associated with lower retail prices, consistent with dominant firms leveraging scale economies.
Overall, the paper makes three main contributions. First, it provides a uniquely detailed empirical picture of concentration and pricing across all stages of a major global value chain. Second, it documents that market power is present both in input markets (affecting producers) and output markets (affecting consumers), with systematic differences across stages. Third, it shows that mergers and acquisitions play a central role in shaping global market structure, particularly among large buyers. These findings have important implications for competition policy and development, suggesting that assessments of market power should adopt a global and multi-stage perspective rather than focusing narrowly on single markets or stages of production.
[1] Associate Professor of Economics at the University of Naples Parthenope, Italy.
