I read the article & the summary of the book, and it makes an interesting argument. Can we predict whether gay marriage will have an effect on the rate of marriage among heterosexuals?
The idea that marriage is about coupling, not about children, economics or family alliances is a change I've seen several sociologists comment upon, citing popular attitudes toward the president's marital status and faithfulness as indicators. Presidents FRD and Kennedy both had affairs, yet were not publicly criticized because they were able to maintain their marriages. OTOH, Nelson Rockefeller could not get his party's nomination because he had been divorced and remarried. 20-30 years later, Reagan's previous divorce was never a matter of public comment, but Clinton's affair almost got him impeached. (And lest you say it was his lying about it, even putting the question to him while his wife continued to maintain the marriage would have been considered unseemly a few years earlier, and if anyone had been so impertinent to ask, of course he would have been expected to lie, in no small part out of consideration for his marriage.)
The causation/correlation question posed by Blankenhorn still is not answered, though. ISTM, the trend toward viewing marriage as about coupling alone, and not about the other social, political and economic aspects previously associated with marriage (not just children, as he asserts) has been on a trajectory for at least 50 years or more. I'm not sure whether permitting or prohibiting gay marriage can be expected to change that trend in any way. Perhaps instead, gay marriage is an issue that may predictably be expected to arise at a particular point along that trajectory, but without having any effect at all on the trend itself.
I don't know. I'm just speculating here. It's an interesting hypothesis, but I think other hypotheses explain the correlation with equal plausibility.