One thing public opinion polls can’t do is to predict turnout; you need to have an actual election to find out who is going to turn out and in what numbers. That’s why I’ve been looking at turnout in this year’s Democratic and Republican primaries, to see which side’s backers are motivated enough to vote. Comparative turnout is particularly helpful in states without party registration, since all voters there can choose to vote in either party’s primary.
Five of the states that voted June 8 have party registration, and in each registered Democrats significantly outnumber registered Republicans. Yet in three of the five it appears that more people voted in the Republican than in the Democratic primary and in the two others it appears that almost as many did. One reason is that in many of these states Republicans had more seriously contested races than Democrats did.
That was the case in New Jersey, where 247,658 votes were cast in Republican primaries for the U.S. House (there were no statewide offices on the ballot) and 164,300 were cast in the Democratic primaries. That’s almost a direct reversal from 2008, when 171,072 New Jerseyans voted in the Republican primaries for House and 278,053 voted in the Democratic House primaries. In Maine Republican turnout (87,115) was only slightly below Democratic turnout (87,764), as both parties had seriously contested primaries for governor. But this is not unusual there: Republican primary turnout in Maine was higher than Democratic turnout in 2006, 2002 and 2000 as well. In Iowa, where Democrats have gained a big party registration advantage over Republicans in the past decade, Republican turnout seems to have been higher, thanks to seriously contested primaries in the three Democratic-held House seats; Republicans had only one House primary and a not very seriously contested primary for U.S. Senate (the winner got 78% of the vote).
Nevada was an interesting case. The riproaring primary for the Republican nomination to oppose Senate Majority Leader Harry Reid drew 175,656 voters while the Democratic primary brought only 116,005. And one-quarter of those Democratic voters opted for Reid’s single unknown primary opponent or the Nevada option of “none of the above.” (Unfortunately, Nevada law does not provide that in the case “none of the above” gets a plurality no candidate is nominated.) That’s a stunningly bad result for Reid, and one notes with interest that pollster Scott Rasmussen shows Angle leading Reid 50%-39%; Democrats will try to disqualify Angle as an extremist but the fact that Reid is getting only 39% when his name and substantive identification is up around 100% is not a good sign for him, to say the least. It should be noted that Republican primaries have drawn more voters than Democratic primaries, despite the Democratic edge in party registration, in 2006, 2004 and 2002, but this year the margin was larger than in any of those years.
The final June 8 primary state with party registration is California. The current results do not include some 700,000 ballots that have not yet been counted; the state website says that 4,227,070 votes were cast, but the returns reported so far show significantly smaller numbers cast in the Democratic, Republican and minor party primaries for governor (3,539,613) and senator (3,493,311). What I find interesting here is that almost as many votes were cast in the Republican as in the Democratic primary, as indicated in the following table showing Republican and Democratic turnout and the Republican percentage of two-party turnout in this and other recent primaries.
Demo turnout Repub turnout % Repub
2010 Senate 1,774,464 1,687,087 49
2010 Governor 1,759,660 1,717,401 49
2008 President 5,066,993 2,932,811 37
2006 Governor 2,506,261 1,916,066 43
2004 President 3,107,629 2,216,351 42
2002 Governor 2,169,555 2,285,452 51
2000 Senator 3,940,664 3,021,699 43
Democratic turnout as currently reported is way below past showings. One reason is that there were no serious contests on the Democratic side for senator or governor, but that was true as well in 2002. If Democrats get the same proportion of the unreported vote as they have of the reported vote the Democratic primary turnout will still be slightly below that in 2002, although California’s population according to Census Bureau estimates increased from 34.9 million in July 2002 to 37.0 million in July 2009. And one might add further that in 2002, incumbent Democratic Governor Gray Davis was reelected by only a 47%-42% margin and then was recalled from office in October 2003 by a 55%-45% margin—not a good omen this year for California Democratic nominee Jerry Brown (who as governor employed Davis as chief of staff). But it’s possible that the uncounted vote may lean more Democratic; we’ll just have to wait and see.
Six of the states voting on June 8 do not have party registration, but only one of them provides really useful clues about the balance of enthusiasm. Virginia had only Republican primaries in five U.S. House seats and no Democratic contests; as I noted on election night, Republican turnout in the one district that had other recent contests was up sharply, double that of 2008 and triple that of 2006. That’s the heavily Democratic 8th district across the Potomac from Washington, D.C. Arkansas had a statewide Democratic runoff for U.S. senator, Democratic runoffs in the two open 1st and 2nd district House seats and a Republican runoff in the open 3rd district House seat: not a basis for comparing party turnout. But it is interesting to note that the latest reported turnout in the Democratic Senate runoff (257,512) is, despite the high attention to and heavy advertising in the race, below the Democratic turnout in the 2008 presidential primary (314,234), the 2004 presidential primary (266,848), the 2004 Senate primary in which Blanche Lincoln had an little known opponent (278,047) and the 2002 governor primary (279,097)–not a good sign for Arkansas Democrats in November.
South Dakota, North Dakota and Montana have only one U.S. House seat each and in each of those races and the North Dakota Senate race there was no competition in one or both parties.
That leaves South Carolina, which had a riproaring race for the Republican nomination for governor and a much quieter race for the Democratic nomination. There Republican turnout was much higher than in recent state primaries, while Democratic turnout was much lower. The numbers and the Republican percentage of two-party primary turnout:
Demo turnout Repub turnout % Repub
2010 Governor 188,765 421,256 69
2008 President 532,151 445,499 46
2008 Senator 147,312 280,861 66
2006 Governor 138,343 247,821 64
2004 President 293,843
2004 Senator 167,790 294,669 64
2002 Governor 316,255
2000 President 573,101
That shows a pretty strong Republican advantage, comparable to or larger than in other election years when Republicans have run strong in South Carolina.
The bottom line: Robust Republican turnout in South Carolina, Nevada and California leaves Republicans in relatively strong position in the balance of enthusiasm in those states in comparison with where they were in the 2006 and 2008 cycles. And the robust Republican turnout in New Jersey might say something positive about the strength of feeling on the policies of first-year Governor Chris Christie.
