Thank's for the link. I was planning on digging one of them out for Albion today but you saved me the trouble. If Albion is interested, she can find more references in Wade, N., Before the Dawn (2007).
One additional factoid. The earliest study I know on this subject was accidental. It was conducted by a hospital in Chicago in the 50's. They were studying something about heritability of blood type. In the course of that study, it became apparent that about 15% of the babies could not possibly have been sired by the supposed father because of differences of blood-type. That means the number was almost certainly greater than 15% as some of the women had probably philandered with men who had the same bloodtype as the supposed dad. (They terminated the study as soon as they discovered this because, in the 50's, this was an unacceptable fact. N.B., it's sad that it is no longer unacceptable--it's just another marker on the long, moral decline of America.) I don't have the time to look this study up. But I'm sure a google search project would provide it.
One could easily look at this fact as an easy explainer of why otherwise conservative women are so adamant about legal abortion. For every women who philanders and gets pregnant, there are certainly more who manage to avoid pregnancy and more who would like to have the option of philandering without having a baby. Altogether, if you start at 15% and add on, this amounts to a fairly substantial portion of married women. That makes abortion the ultimate, fix-it-up-after-you-break-it solution for a substantial group in the population. It would explain the odd voting pattern of married women on this issue.
I objected to that figure, because you attributed this behavior to married couples. I read your link and none of the studies gave that high a figure for married couples, only for all pregnancies. Also, the studies were conducted in ultra-liberal European countries as well as the U.S.
Given today's promiscuity among young people and Europe's wholesale rejection of religion except among its Muslim immigrants, I do believe this figure is possible when applied to all pregnancies in Europe and the U.S., certainly including poverty communities such as "trailer trash" or "urban youth", as well as those marriages that took place when, or because, the bride was pregnant.
But I do not believe it could be accurate for couples who marry and wait to have children after marriage. One of your sources also cited the lower incidence of "cuckolding" among the more educated couples.