Cartels are a common feature in most economies. For instance, in the 2015-19 period, the European Commission imposed fines in excess of €8.3 billion in 27 cartel cases. It is well understood that cartels reduce social welfare by increasing prices and restricting output. Cartels may generate additional costs to society by increasing the market share of less efficient producers.

One way in which this may happen is if during cartel episodes low-productivity firms are tempted to enter the market by the artificially high profits generated by the cartel, raising productivity dispersion and reducing total surplus. This intuitive mechanism has so far not been quantified because of data limitations: A researcher dealing with this question needs to know with certainty when cartels started and ended, and to have detailed plant-level data. And, since cartels are illegal throughout the developed world, it is very difficult to access the necessary information in modern settings.

In his job market paper, “Cartels, Entry, and Productivity: Evidence from the Chilean Nitrate Cartels”, UCLA PhD students Felipe Carrera and Vitaly Titov quantify the effect of cartels on the quantity and quality of new firms in an industry with low barriers to entry by studying the Chilean nitrate cartels in the early 20 Century. This industry is attractive because cartels were legally enforceable, entry barriers were low, and, over 35 years, the industry switched between cartel and perfect competition multiple times.

Using a newly collected dataset, created from original handwritten archival records, they show that during cartel periods, average entry increased from 5 to 9 plants per year. Moreover, entrants during cartel periods were one-third less productive than average firms. Building on these findings, they build a model of firm entry and conduct two counterfactual simulations. First, using detailed accounts from historical records about these cartels’ inner workings, together with their structural estimates, they are able to show that low barriers to entry lowered incumbent profits by 40%. Second, they estimate a model of firm dynamics and show that had the cartel not existed, 25% of plants would have postponed or cancelled their entry to the industry.

These findings are important from a historical and present-day perspective. In the early 20 Century, Chile’s nitrate industry dominated the country’s economy, accounting for 65% of exports and 45% of government revenues. The excess entry lowered the mean productivity of the industry by 3% and had a measurable effect on tax revenue and GDP. This poses important lessons for other developing countries that are dominated by extractive industries. Moreover, the paper speaks to the literature on productivity dispersion by showing that market power caused by coordinated action by otherwise independent firms can generate substantial amounts of inefficient entry.