Your premise that “social stigma” correlates (or has ever correlated) with curbing human behavior is charming, though inaccurate. Americans may be defaulting on their mortgages, however there is no compelling relationship between the “decline of stigma” for homosexuality, divorce, or other behaviors and defaulting on loans. What people say they think and what they actually do are rather different. Social science research has repeatedly found this to be the case.

Shifting trends in what is considered “stigmatized” indicates changes in larger socio-political systems. For example, Stephanie Coontz’s research in her book “Marriage: A History” indicates that throughout the history of European cultures, the stigma of homosexuality has changed depending on the larger socio-political context. In the past 30 plus years in the U.S., many factors have changed for all members of society including increased economic opportunities for women, increased participation in parenting for men, legislation banning discrimination based on sexual orientation and other factors. (Incidentally, 1973 is the year that the Diagnostic Statistic Manual (version III) removed “homosexuality” as a mental illness.)

For another look at how individual opinions on “stigmatized behaviors” does not influence behavior, consider the evidence on affairs in heterosexual marriages. Check out research conducted by the late Shirley Glass: Two out of five heterosexual married couples are having affairs while almost all self-report that they believe that extra-marital affairs are wrong. The region in the U.S. with the highest rate of extra-marital affairs? The Bible Belt. While the rates of men having affairs has remained the same over the past 50 years, there is an increase in women having affairs, not because of a decrease in social stigma, but because of an increase in women’s access to affair partners which is nearly equal to that of their male partners.

When examining changes in stigmatized behaviors, self-report measures of individual’s opinions are rarely an indicator of actual behavior. More revealing questions on this issue may be:

1) What changes occurred within the profession and practice of mortgage lending?
2) Was there legislation that eased access to mortgages as well as easing consequences for defaulting?
3) What other socio-economic trends occurred as a population within the last thirty to fifty years in the U.S. that might create a climate for mortgage defaulting?
4) Are more Americans really defaulting on mortgages or are more Americans taking out mortgages and defaulting at the same rate as they always have?

Your last graph is actually evidence that norms have a dramatic change (e.g. whatever happened with opinions on homosexuals in the early 90′s) and then beliefs slowly asymptote to a new norm.