Changing public attitudes toward federalism [Updated with correction about continuity in public attitudes towards federalism and education policy]
In this fascinating recent Cato Institute study, political scientists John Samples and Emily McClintock Ekins compile extensive survey data indicating that public opinion has shifted in favor of political decentralization over the last several decades. In a poll conducted last year, Samples and Ekins replicated a series of questions on what level of government should make "the major decisions" on various issues, that were first asked in a 1973 Harris poll. With the exception of national defense and education (where there has been no statistically significant change), there has been a substantial increase in preference for state and local control relative to federal.
On several important issues, majority opinion has actually flipped over the last forty years, shifting from a majority in favor of federal dominance to a majority against it. For example, the percentage of Americans who believe that state or local government should make the major decisions on drug policy has increased from 39% in 1973 to 61% in 2013. On health care, it has risen from 40% to 62%; on environmental protection, it has gone from 36% to 56%. On prison reform, the proportion supporting state and local primacy has increased from 43% to 68%.
In both 1973 and 2013, substantial majorities favored federal primacy on national defense, Social Security, and cancer research. But in the last two cases, the minority preferring state or local control has substantially increased. Similarly, in both 1973 and 2013, large majorities favored state or local control of education, transportation, housing, and welfare policy. But on all three of the four issues, those anti-federal government majorities have grown substantially.
Samples and Ekins cite other data that also indicate a shift in favor of local and state control on many issues. They do, however, point out that large majorities still favor federal primacy on civil rights and immigration issues, as well as the few issues on which federal control retains majority support in their replication of the 1973 Harris study.
Overall, the data shows a major shift in favor of greater support for state and local control of many issue areas. As Samples and Ekins note, the change is probably due in part to growing public distrust of the federal government. I would add that another factor is the decline in the association between state and local governments and racism. In the 1960s and early 1970s, "states rights" was still commonly associated efforts to defend racial discrimination and desegregation. Today, that association is far weaker, and there is therefore less stigma attached to state and local control over various issue issue areas. Even the extreme states' rights idea of secession has lost much of its previous association with racism, as shown by a 2008 poll in which support for a right to secession was higher among African-Americans than whites. That changed by 2012, but probably only because of the shift in partisan control of the White House.
None of this proves that greater political decentralization is actually a good idea. Having done extensive work on political ignorance, I would be the last to deny that popularity is often a poor predictor of good policy. If you're a liberal policy wonk who supports greater federal power, you can argue that the public fails to appreciate various dangers of decentralization, such as as races to the bottom and oppression of local minorities. If you're a conservative centralizer, you can still claim that the public doesn't understand the spread of addiction and moral decay that occurs when individual states are allowed to legalize drugs or adopt same-sex marriage. My fellow libertarians, too, often have reasons for adopting a skeptical attitude towards federalism, at least on some policy areas.
But if you believe, as I do, that increased decentralization is desirable on many issues, then the survey data is somewhat reassuring; not because it somehow proves that our ideas are right, but because it suggests they are more politically feasible than in recent past.
UPDATE: In the original version of this post, I accidentally failed to note that support for federal primacy in education has grown slightly since 1973, according to Samples' and Ekins' data, rising from 23 to 25 percent (a difference which is not statistically significant). I apologize for the mistake, which I have corrected). But it is also worth noting that a large majority in 2013 (75%, as compared with 72% in the 1973 survey, which had an undecided option) still favored state or local dominance in this policy area.