The Effect of No-Fault Divorce Legislation on Divorce Rates: A Response to a Reconsideration

Abstract

We would be very comfortable if our original article (Nakonezny, Shull, & Rodgers, 1995) were scrutinized by careful social scientists alongside accompanying reconsideration of our findings by Glenn in this issue. In that case, arguments would stand on their own, and two sets of findings could be blended by impartial experts in proportion their value. We are certain that each article would find its share of praise and criticism. Although we are clearly not an impartial audience, we have given careful consideration Glenn's reconsideration, and our conclusion has not changed. Glenn's article did not succeed in convincing us that his designs substantially improved upon ours, that we offered inappropriate, untentative that we should stay away from gatherings of student[s] of divorce in U.S.; or, especially, that we should in any fundamental sense revise our conclusions. In fact, we think Glenn bolstered our conclusions but with data and methods weaker than ones we used. We never did believe that all of increase in divorce during divorce boom resulted from enactment of no-fault provisions. We do believe that there is little evidence support Glenn's conclusion that the adoption of no-fault divorce had little direct, immediate effect on divorce rates, not even in Glenn's own results. We occupy a middle ground. For both empirical and logical reasons, we believe that when a no-fault divorce law was enacted in a typical state, divorce increased a higher level than they would have reached otherwise. It is hard for us imagine how anyone-a serious student of divorce or not-could expect otherwise. But we also believe that many postlegislation divorces would have occurred anyway, with or without no-fault laws. In this reply, we compare Glenn's results and conclusions with our results and conclusions. Our basic position is that our findings and Glenn's findings lead same conclusions, ones from our original article. We begin with problem of whether changes in divorce should be interpreted as raw changes or should be adjusted for period effects. Glenn criticized our comment that the switch from fault divorce law no-fault divorce law a measurable increase in divorce (Nakonezny et al., 1995, p. 485). He understood that our reference an increase of .80 divorces per year per 1,000 individuals in state population was simple difference between average divorce rate before and after implementation of no-fault law in each state, for background divorce rates. He was concerned, though, that a casual reader would interpret this figure as one adjusted for prevailing divorce patterns, and we share that concern. We agree with Glenn that our use of phrase led to is confusing and too strong if viewed in isolation. But it was not used in isolation, and we do not feel that there is much doubt about our methodology or conclusions from our original article. Glenn's results and additional results that we present in this reply make our original case even stronger. Neither Glenn nor we feel that unadjusted effect size is of primary interest. We restate one of our primary conclusions, which followed from several statistical tests: The change in divorce rate is different from what it would have been if divorce law had not changed (Nakonezny et al., 1995, p. 485). Compare that statement with one of Glenn's conclusions: This control procedure . leaves enough of an indicated increase be tentative evidence for modest positive effects of no-fault provisions on divorce rates (p. 1024). It sounds like we agree. Glenn stated that if the adoption of no-fault divorce had been a crucial reason for increase in divorce, increase should have been substantially greater for states that adopted no-fault divorce during boom years than for either states that adopted earlier or those that adopted later (p. …