Professor Daniel Lowenstein is the Director of the Center for the Liberal Arts and Free Institutions at UCLA Law School. He writes to comment on Paul Mirengoff’s “A sense of the Senate.”
Mr. Mirengoff’s state-by-state analysis seems correct to me, but for a complete assessment of the probabilities, another important variable needs to be considered. Congressional elections are decided by a combination of national and local (meaning state, in the case of the Senate) considerations. That in itself is no problem for a state-by-state analysis. Mr. Mirengoff’s assessment in each state and the evidence on which it is based (e.g., polls) themselves reflect national as well as local considerations.
But the national considerations are not necessarily fixed. They have been changing continuously since anyone started seriously doping out the Senate elections. In general, they have been changing in the direction of the Republicans. Continuing response to developments, positive and negative, in the economy, and other events, or simply a further drift in the direction of the national mood, can still change the national picture. Because the continuing trend has favored the Republicans it is probably the case that any additional changes are also likely to do so, but there is no guarantee of that. On election day, it is possible that things will break more favorably for one party or the other than seems likely today.
Your present state-by-state assessment is much more optimistic than seemed likely or even possible a few months ago. I am not predicting it will be much different on election day than now, much less which party it will favor if it is different. But it may be different and if so, then the list of states that seem to be in play or that seem safe for one party or the other will change. If you want to assess the likelihood of the Republicans gaining a net of ten seats or more, you need to supplement your state-by-state analysis–hold the currently Republican seats and win 6 of 6 of the close ones or most of those plus one or more of Connecticut, Delaware and West Virginia–by the possibility that the national tide against the Democrats gets even stronger, so that some or all of those six become safe and some other states–New York, perhaps?–that do not presently seem to be in play become so.
Again, I am making no predictions. Only trying to complete the picture of possible ways for the Republicans to gain control.
PAUL adds: Professor Lowenstein makes a good point. My analysis attempted to account for the prospect of continued improvement in Republican fortunes when I gave the Republican candidate the edge in a five of six “toss-up” races, including several where the Democrat holds the lead, and rated Washington as 50-50 even though the Democrat leads in most polls. It also was the main reason why I claimed that Republican chances of capturing the Senate are better than the odds quoted by Nate Silver based on his statistical model.
But it’s certainly possible that the national tide — which I agree is driving these elections in favor of Republican candidates– will prove to be stronger than I projected. It could also be weaker, but that seems less likely to me.