Explaining marriage stats

I'm looking at an interesting graph on page 18 of Promises to Keep (ed. by Popenoe et al.) in an article called "Values, Attitudes, and the State of American Marriage" by Norval D. Glenn. The graph plots percentages of people 18 and older who are unmarried, happily married, and unhappily married between the years 1973 and 1993. I'm not sure what I expected, but the percentage of those unhappily married is a pretty straight line, with happily married decreasing and unmarried increasing. This could mean a couple things. First, it could mean marriages are no more or less happy than they ever were; just fewer people are trying it. This doesn't work in my mind because if that were the case the unhappily married line would be going down too. Second, marriages overall are unhappier (or unhappy marriage is easier to get out of with no-fault divorce), so more people are getting out of them. Once again, though, you'd think the unhappily married line would go down. Could it be that all the well-adjusted people who would be good candidate for happy marriages are finding other things to do (like pursue careers or higher education)? I don't know, just trying to figure out the graph. It would be helpful if they would break down the "unmarried" line into "never married" and "formerly married." That might shed some light on what's going on.