Structural Conditions for Collusion Generation: The “Coordinateable Space” Hypothesis and EU Manufacturing Evidence
This paper proposes the "Coordinateable Space" hypothesis, arguing that collusion depends on the number of effective competitors rather than just market concentration, and validates this through EU manufacturing data showing that a specific range of large firms (approximately 400 to 1,300) is a necessary structural condition for cartel generation.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
In the world of economics, there is a long-standing puzzle about how companies manage to secretly agree on prices without getting caught. This is known as collusion, and when it happens on a large scale, it forms a cartel. For decades, regulators and experts have looked at market concentration to find these groups. In simple terms, market concentration measures how much of the total sales in an industry is held by the biggest players. If a few companies control almost all the sales, it seems logical that they could easily sit down and fix prices together. This idea has guided antitrust laws and merger reviews for years. However, this traditional view focuses only on the size of the companies' sales, ignoring a more fundamental question: how many independent decision-makers are actually involved? A market where four giant companies share the sales is structurally very different from a market where forty companies share the same amount of sales, even if the total sales numbers look identical. The difficulty of getting forty different bosses to agree and keep a secret is far greater than getting four to do the same.
A new study by Shidong Li challenges the old way of thinking by shifting the focus from sales shares to the actual number of large companies. The research proposes a concept called "coordinateable space," which suggests that for a cartel to form, two opposing conditions must be met at the same time. First, the group of large companies must be big enough to cover a significant portion of the market; if they are too small, smaller competitors will undercut their prices and ruin the scheme. Second, the group must not be so large that the cost of organizing them becomes too high. If there are too many companies, the effort required to communicate, monitor each other, and punish anyone who deviates becomes more expensive than the profits they would make from the illegal agreement. The study argues that collusion is most likely to happen in the "sweet spot" between these two extremes, where the group is large enough to control the market but small enough to be manageable.
To test this idea, the researcher analyzed twenty years of data from 2005 to 2024, covering twenty-four different manufacturing industries across the European Union. The team looked at official government records of cartel cases and cross-referenced them with detailed business statistics that track the number of large enterprises in each sector. They specifically counted companies with more than 250 employees, treating these as the "effective competitors" who have the power to influence the market. The results were striking. When comparing different industries, the number of large companies was the strongest predictor of whether a cartel case would occur. In contrast, traditional measures of market share, such as the percentage of sales held by the top four firms, lost their ability to predict cartel activity once the number of companies was taken into account. This suggests that regulators have been looking at the wrong metric; knowing how much market share the top firms hold is less important than knowing how many of them there are.
The data also revealed a specific pattern in how the number of companies affects the risk of collusion. The relationship is not a straight line where more companies always mean more risk. Instead, it follows an inverted-U shape. As the number of large firms in an industry increases from a small number, the risk of collusion rises because the group becomes large enough to cover the market effectively. However, this risk peaks at a certain point and then begins to fall. The study found that this turning point occurs when an industry has roughly 1,300 large enterprises. Beyond this number, the coordination costs become so high that forming a cartel becomes too difficult and unstable. This means that industries with a very small number of large firms are not necessarily the most dangerous, nor are those with a massive number of firms. The highest risk lies in the middle range, where the group is substantial but still manageable.
Perhaps the most practical finding of the study is the identification of a specific threshold that acts as a safety zone. The analysis showed that if an industry has fewer than approximately 400 large companies, it is extremely rare for a cartel to form and be caught. In the data examined, no cartel cases occurred in industries where the number of large firms was below this level. This does not mean that collusion is impossible in these smaller industries, but rather that the structural conditions required to sustain a secret agreement are almost never met. This discovery offers a powerful tool for competition authorities. Instead of trying to monitor every industry equally, regulators can use this number as a filter. Industries with fewer than 400 large firms can be treated as a "safe harbor," allowing authorities to focus their limited resources on the sectors where the structural conditions for collusion actually exist.
The study also looked at price behavior to see if it matched the theory. They found that in the years leading up to a cartel being exposed, the prices in those industries were unusually stable and showed less volatility than in industries without cartels. Once the investigation began and the secret agreement was broken, price volatility returned. This pattern supports the idea that the companies were successfully coordinating their prices before they were caught. The research confirms that the structure of an industry—specifically the count of its major players—creates a long-term environment that either encourages or discourages illegal cooperation. This is not a temporary fluctuation but a deep-seated feature of the industry, determined by technology, product types, and how the market is built.
By moving away from complex share calculations and focusing on a simple count of large companies, this research provides a clearer map for fighting cartels. It suggests that the old rule of thumb, which assumed that high market share automatically meant high risk, is incomplete. The real danger lies in the specific number of decision-makers involved. For the first time, there is a data-driven way to define the exact range where collusion is most likely to thrive. This approach allows for a more efficient use of antitrust resources, ensuring that regulators are looking at the industries where the structural soil is right for cartels to grow, rather than wasting time on sectors where the conditions simply do not allow it to happen. The findings offer a concrete, evidence-based method to prioritize investigations, making the fight against price-fixing more targeted and effective.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.