Carol, if you start with the premise that there are no differences there is no need for arguing that it would be difficult to conduct good research on that field. But since you discuss that I chose to take your 'premise' statement not verbatim.

Any research on such questions would be quasiexperimental and/or correlational. Many variables would be confounded and the effect of one single variable would be difficult to detect. But it can be done and it has been done. The way it has been done is of course open to critique for that is the only way in science to separate the chaff from the wheat.

Take social background (or: class) for example, a very interesting variable which has often been used as a predictor. This variable too is confounded with gender, ethnic background, religion etc. Should we not do research on the influence of social background because it is difficult to isolate this variable? Should we not allow gender studies for that variable is confounded with so many others? A lot of very good and interesting findings would have been prevented to be found.

I disagree that such research would only promote hatred. It also can do a lot of good. I don't think it is pure chance that you mention explicitely only the variables race and religion. I see a deep rooted reluctance among USAmericans to consider these variables as input variables whereas they have no similar reluctance to consider social background, gender, even ethnic background as variables in a complex environment. The real reasons for that seem to me more ideological than methodological for the reasons actually given are reasons that are equally valid for each other variable considered.

Wolfgang