One common way of thinking about the possibility of federal reforms – in both the legislature and the Supreme Court – is that they are more likely to occur as the number of states that enact those reforms at the state level grow. For example, Ruth Bader Ginsburg once argued that the Supreme Court had stepped in too quickly in the abortion issue. When Roe v. Wade was decided in 1973, only 4 states allowed abortion “in nearly all cases before the fetus was viable.” But support for abortion was growing. Ginsburg’s point was that the Supreme Court’s early and decisive action had prevented the country from continuing to change its mind gradually on the issue. After such a development and a large number of states supporting abortion, a Supreme Court decision constitutionalizing abortion would have been less controversial. By contrast, Griswold v. Connecticut, which recognized a constitutional right for married couples to have contraception, was a far less controversial decision in part because it struck down laws in only one or two states.
Ginsburg’s analysis of Roe recently came up when the issue of gay marriage was being debated in the country and decided by the courts. It was commonly thought that the Supreme Court would wait until a large number of states actually had decided in favor of gay marriage before announcing it as a constitutional requirement. And in 2015, when Obergefell was decided, 36 states allowed same sex marriage (although the process had ended up moving more quickly than many people expected).
Although 36 states allowed same sex marriage, the great majority of these states did so only because of court decisions. A rough and quick count indicates that only 10 states legalized same sex marriage by legislative decision, with the remainder being required to do so based on judicial decisions. Thus, while a significant elite supported same sex marriage, legislatures and the voters were much less supportive.