The present study examines the impact of several industry characteristics on the propensity to collude using a dataset on the existence of collusion across Dutch industries during the late 1990s and early 2000s. The results of the Probit model with sample selection indicate that our sample of Dutch concerted practices is non-random in the sense that it only consists of anticompetitive agreements that were subject of an antitrust immunity behavior. Our bivariate probit model with sample selection indicates that concerted practices are less likely to be seen in service industries relative to manufacturing industries. The results also show that it is more likely that firms engaged in concerted practices in unconcentrated industries. Furthermore, we could not find a non-linear relationship between concentration and the presence of collusion. There is also strong evidence from all the regressions that concerted practices are less likely in industries where entry is more possible. Interestingly, our estimation results indicate that there is a positive correlation between cartel prevalence and import penetration, which implies that import competition did not discipline firm behavior and foreign importers joined the cartel paradise in the Netherlands. As to the role of measures of asymmetry on concerted practice prevalence, the association between patenting activity and propensity to engage in collusion is ambiguous in the current setting, while advertising intensity, as the second measure of asymmetry, is associated with increased likelihood of collusion. Contrary to the previous empirical findings, market growth has been found to have a negative effect on the probability of a concerted practice in an industry. Furthermore, our proposition that growing demand might attract new entrants, which, in turn, hampers collusion, has been falsified in the current context. cross-border cartels that involved private firms. If one takes the within-border cartel cases and undetected cartels into account, the number is enormous.More importantly than the number of cartel cases, the damage they give to the economy has made economists and policymakers worry about the phenomenon of collusion. A recent studyby Connor (2004) reveals that the median increase in price resulting from collusion was about 25 %, which shows how effective these collective agreements among firms are. Similarly, Griffin (1989) finds that cartels in his sample charged a 45 % markup over marginal costs on average.Given the enormity of cases, and the welfare transfers they have caused, the detection of cartels-which is a longstanding antitrust problem-is highly essential. In this respect, it is important for regulatory bodies to develop structural screens to forecast where and under which conditions cartels are more likely to be formed, which will help them allocate investigative resources where they would be most constructive. According to Harrington (2008), this screening process is the first stage of a multi-stage process of the detection of cartels, which...