The difference-in-difference method is a very effective but demanding econometric tool to estimate cartel damages. It can avoid the shortcomings of cross-sectional and time-series approaches by combining comparisons over time and across unaffected markets. But it requires that the comparative market is explicitly cartel-free.
Difference-in-difference estimators are capable of controlling for what would have happened absent the infringement by examining what changed over time for the infringed and non-infringed markets, followed by a comparison of those differences. Factors such as differences in size, macroeconomic conditions, market characteristics, and exogenous shocks can be controlled in order to improve the comparability of the basis and the calculation of the damage. Also, the control of those factors is a demanding task.