Do website users follow links equally to racial and non-racial sites?

If people followed web links randomly, we would expect different traffic statistics than the data actually show.