I'm sorry, but this is obviously not true; there are clusters on exactly 3, 4, 5, and (faintly) 6 percent in the European market. And given that forcing to arbitrary whole numbers _happens_, it suddenly seems very possible that in a (larger) market like the US the forcing acts even stronger.
In other words, we need _evidence of collusion_ to determine for sure if there is a cartel, not just evidence that it is the same.
That said, however, the difference between actual collusion and accidental collusion isn't that thin.
I'm sorry, but this is obviously not true; there are clusters on exactly 3, 4, 5, and (faintly) 6 percent in the European market. And given that forcing to arbitrary whole numbers _happens_, it suddenly seems very possible that in a (larger) market like the US the forcing acts even stronger.
In other words, we need _evidence of collusion_ to determine for sure if there is a cartel, not just evidence that it is the same.
That said, however, the difference between actual collusion and accidental collusion isn't that thin.