As the New York Times reports, even before the Citizens United decision, 26 states didn’t restrict independent corporate (and, I take it, union) speech about state-level political candidates. This means two things:
(1) We should take with a grain of salt claims that Citizens United will substantially transfigure American political life. Unless there’s some reason to think that elections in, say, Idaho, Indiana, or Maryland (three states that didn’t restrict independent corporate speech) are much more dominated by corporate or union speech than elections on Montana, Ohio, or Pennsylvania (three states that did have such restrictions), we shouldn’t lightly assume that shifting the other 24 states and federal elections to the Indiana model from the Ohio model would change things dramatically.
Of course, federal elections are in some measure different, because Congress is more important to corporations and unions (and others) than the states. But plenty of corporations and unions have special interests in particular states, or are even predominantly present just in one state or a few states. If political life in the corporate/union-speech-permitting states is basically much the same as political life in the corporate/union-speech-restricting states, there’s little reason to think that allowing such speech at the federal level would dramatically affect (as opposed to just moderately affecting) federal campaigns.
(2) What I just said, though, is based on my tentative sense of things; naturally, it would be great to have more real data on it. Fortunately, the 26-to-24-state split would provide a dataset that could let scholars figure out just how much of an important the allowing of corporate and union speech in political campaigns has had, on the spending on those campaigns and perhaps even on legislative decisionmaking. I asked some election law scholars about this, and didn’t get any pointers to such studies — but if any of you do have pointers to such studies, I’d love to see them. And of course such studies could also be done retroactively; it would be interesting to see what data scholars can extract out of this natural experiment.
None of this bears, of course, on moral, constitutional, or political theory arguments that focus on the theoretical right and wrong of restricting or protecting such speech. But to the extent that the arguments have included empirical claims about the likely effect of protecting corporate and union speech about candidates, it seems to me that the states’ experience would provide a helpful perspective.