I will show with Monte Carlo analysis that the test I propose is robust to unobserved heterogeneity, in particular to unobserved bidder and contract characteristics, unlike the existing method that is developed for the same purpose. I will also discuss how the old and the new test complement each other. I apply both these tests to school yard snow removal auctions in the City of Helsinki. Second, this empirical application is important in itself because it is the first empirical study of a territorial allocation scheme. Third, this essay has a practical economic policy implication. Since it points out two of the participating bidders suspect of collusion, this analysis should validate closer legal study to support the prosecution of these companies.