In SPS vol. 19, no. 3, 1996 Professor Douglas Hibbs has responded to our critique in SPS vol. 17, no. 1, 1994 of his book Solidarity or Egoism from 1993, which in turn consists mainly of a critique of some work we were doing at that time, most of which had not yet been finalized, not to say published, when Hibbs' book came out.¹ Thus, by now we can look back at more than three years of debate. It has not been restricted to the pages of this journal: we have also had ample opportunity for discussing our differences at several scientific meetings.

As a matter of principle, we welcome academic debates and exchanges, even fierce ones. After all, it is disagreement rather than dull indifference or inattentiveness to the results of other researchers that ferments science and may give rise to new developments. As to ourselves, we gladly acknowledge that we have benefited from our ongoing discussions with Hibbs. They have made us go much further in testing and documenting the robustness of our main results across various designs, operationalizations of key variables, periodizations, etc. than was originally intended. In our view, the robustness of our results not only raises reasonable doubts as to the substantive significance of Hibbs' varying objections - it also represents an interesting finding in its own right.

Considering Hibbs' last contribution in the context of our previous exchanges, we cannot help feeling, however, that the debate has passed the point where it can be expected to yield something scientifically fruitful or constructive. We are obviously still in disagreement, and what we are into now - and maybe have been for some time - is at best "methodological

---

*Peter Nannestad, Department of Political Science, University of Aarhus. Martin Paldam, Institute of Economics, University of Aarhus, Universitetsparken, 8000 Aarhus C, Denmark.

sniping," as this activity has so aptly been characterized by Sanders, Marsh
& Ward (1990, 88).

In this rejoinder, we therefore restrict ourselves to one final comment. It concerns Hibbs' charge that in our review of his book we argue ad hominem. We totally agree with Hibbs that arguments ad hominem are inappropriate in scientific discourses. We think that this applies to all participants and to all types of exchange. Hence, we would not want to attack anybody, including Hibbs, in that way.

What we have done in our review is to raise the issue of a value bias in Hibbs' book. We have claimed that he argues from a preference for one particular result or interpretation. We have said nothing about why he may hold this preference - that might easily involve assertions ad hominem. The book's value bias, we have argued, reveals itself in the usual ways: nonneutral, evaluative terminology, selective use of material, and particular types of argumentation. We provide examples of all this in our review.

Hibbs seems to equate raising the issue of a value bias with making ad hominem assertions. We believed it to be a basic methodological tenet in the social sciences, especially in political science, since the days of Weber - or at least of Brecht, Myrdal & Easton - that value biases may be as damaging to the validity of analyses and results as, say, biased statistical estimators. Both types of bias result from some rule of scientific method having been violated. Raising the issue of a possible value bias is hence as scientifically legitimate as raising the issue of a possible estimation bias. Hibbs may disagree with us as to whether there really is a value bias in his book - that is an entirely different matter and does not make raising the issue illegitimate. Discussing a possible value bias is not any more ad hominem than discussing a possible estimation bias.

---

NOTES

1. Hibbs' historical account is so blatantly self-serving that it hardly calls for any detailed comment — one single correction will suffice. Contrary to Hibbs' account the quality of our intended contribution to the joint book project was never an issue in our dealings with the research director of the Rockwool Foundation. The decision not to publish it in the end was due to a conflict over intellectual rights. However, we found other outlets for our work on egotropic and sociotropic voting, including the parts Hibbs focuses on: at present, two pieces have been published, and two more have been accepted for publication, in refereed, international journals.

REFERENCES

Sanders, D., Marsh, D. & Ward, H. 1990. "A Reply to Clarke, Mishler and Whiteley," British
Journal of Political Science 20, 83-90.

As regards Scandinavian Political Studies, this discussion is hereby closed.