We investigate the correlation properties of different samples of galaxy clusters, which are obtained by cutting the Lick catalogue at different values of the galaxy surface density. This kind of cluster-finding algorithm is equivalent to the application of the biasing prescription to the Lick galaxy distribution. For all the samples, we find that the angular two-point correlation function can be well approximated by a power law, , with the same logarithmic slope , in the angular range . In addition, the correlation amplitude turns out to depend on the value of the limiting threshold, used to identify clusters, in exactly the same way as predicted by the biased model of galaxy formation. We also find that such a relation is reflected in a well-defined dependence of the clustering length, to the cluster richness which confirms previous results obtained in the literature from the analysis of the distribution of Abell clusters.