A Look At The Polls Of Mid-March

Let's take a look at the polls for the two parties' nominations and for the general election as of mid-March, based on the compilations in realclearpolitics.com.

First, the Republicans. I've calculated the average percentages for Rudy Giuliani, John McCain, Mitt Romney, and Newt Gingrich for each month and for each have indicated the number of polls represented.

Polls Giuliani McCainRomneyGingrichOthers, undecided
March 2007 7 37 21 8 11 23
February 2007 13 35 20 8 11 26
January 2007 9 29 23 7 12 29
December 2006 6 30 25.5 6 10 28.5
November 2006 5 27 27 7 7 32

Conclusions: This confirms Giuliani's significant rise from the 27-to-30 percent level of November to January to the 35-to-37 percent level of February to March. Also, McCain's fall from 25.5 to 27 percent in November to December to 20 to 23 percent in January to March. Also, note that the percentages for others and undecideds have been declining, roughly in tandem with the move to Giuliani; Romney's numbers have not had a significant change, and neither, it seems, have those of Gingrich, who in any case is not a declared candidate. Yes, there's a market out there for another alternative. But it's not a market that has been expanding.

The Iowa and New Hampshire polls from December through March (though there have been none for Iowa in March yet) show a somewhat different picture.

Polls Giuliani McCainRomneyGingrichOthers, undecided
Iowa 6 26 22.5 8 13 30.5
New Hampshire 8 27 28 16 - 29

Giuliani's lead over McCain in Iowa is much less than nationally, which is probably a reflection of the cultural conservatism of Iowa Republican caucusgoers (although keep in mind that it is difficult to get a sample that is representative of the very small segment of the Iowa electorate that attends the caucuses). Nonetheless, Giuliani at this point seems capable of a credible enough showing to keep him very much in contention for the big state primaries February 5 (which now will definitely include California). McCain, too, has Iowa numbers that would keep him credible for later contests. New Hampshire, where McCain won a smashing victory in 2000, still shows him with a statistically insignificant lead over Giuliani; Romney, from next-door Massachusetts, does significantly better than in Iowa. Both of these candidates thus seem to be well positioned for the South Carolina primary and the February 5 contests. Romney needs to move up in New Hampshire to make himself credible for those contests. McCain needs to draw registered independents into the Republican primary; here his biggest competitor may be Barack Obama.

Now for the Democrats. I've included numbers for Al Gore, with a parenthesis indicating the number of polls per month in which his support has been measured. I haven't given an average for those undecided or others, because it varies significantly depending on whether or not Gore is included.

Polls Clinton ObamaEdwardsGore
March 2007 8 36.5 2 12 15 (4)
February 2007 13 36 22 12 12 (9)
January 2007 9 34 19 12 9 (9)
December 2006 6 35 16 11 12 (5)
November 2006 5 33 18 10 10.5 (4)

Conclusions: Hillary Rodham Clinton has a steady one third of Democratic primary voters or caucusgoers--a clear lead but not an overwhelming number, given her connection to the one successful (certainly in the opinion of Democratic voters, anyway) Democratic president of the past 40 years. Obama has clearly been rising, though not yet to parity with Clinton. John Edwards has shown no progress, Gore some oscillation.

Again, the Iowa and New Hampshire numbers look somewhat different:

Polls Clinton ObamaEdwardsGore
Iowa 7 21 17 23 14
New Hampshire 9 30 23 17 -

Edwards's nonstop campaigning in Iowa clearly gives him the potential for an upset victory here, which presumably would boost his later numbers upward. But remember that in 2004 his impressive second-place finish in Iowa (buoyed by a deal with Kucinich supporters for their second-choice support) failed to get him above fourth place in New Hampshire. Edwards is currently running above his national level in New Hampshire but still is a distinct third there. These numbers suggest that the biggest hurdle for the Clinton campaign is Iowa; finishing on a high note there will leave her well positioned for the Nevada caucus and the primaries in New Hampshire and South Carolina. But if she doesn't do well in Iowa, those other three early contests could be fraught with peril for her--and full of opportunity for Edwards, Obama, and perhaps another candidate to improve their standing in the February 5 contests.

Now let's look at the general election pairings by month, starting with Giuliani and McCain vs. Clinton. (There's not much point in examining pairings with Romney, who is little known and runs far behind the well-known Democrats.):

Polls Giuliani ClintonUndecidedNoMcCain ClintonUndecided
March 2007 3 48 42 10 1 43 44 13
February 2007 9 48 44 8 7 46 45 9
January 2007 6 49 44 7 9 47 45 8
December 2006 4 47 44.5 8.5 5 47 44 9
November 2006 4 47.5 45 7.5 4 48 44 8

The first thing to say is that while Giuliani and McCain both tend to run ahead of Clinton (except in the one March poll matching McCain and Clinton), they don't tend to run ahead by a robust margin (except in the three March polls matching Giuliani and Clinton). I wouldn't make much of the March results unless later polls show they were the start of a trend. In 2006, Giuliani and McCain ran ahead of Clinton by essentially the same margin, but in 2007, Giuliani's margins have been just a bit larger than McCain's. Giuliani's numbers have moved slightly upward, McCain's slightly downard. But there's not much difference in either direction.

Here are the data on Giuliani and McCain vs. Obama.

Polls Giuliani ObamaUndecidedNoMcCain ObamaUndecided
March 2007 3 45 42 13 3 43 48 9
February 2007 6 46 42 12 6 43.5 44 12.5
January 2007 7 47 42 11 8 46 43 11
December 2006 3 47 40 13 4 47 39 14
November 2006 2 49 40 11 1 47 39 14

One sees that Obama has become a closer competitor in 2007 than he was in 2006; starting in February, he leads McCain but still trails Giuliani. The number of undecideds is lower than in pairings against Clinton, and so are the percentages for the Republican candidate. This suggests that Clinton polarizes the electorate and that negative feelings toward her send some significant number of voters who might otherwise be undecided over to the Republican nominee.

Now for Edwards vs. Giuliani and McCain. There are fewer polls here, so I will collapse the months, so as not to place undue emphasis on the results of single polls.

Polls Giuliani Edwards Undecided Polls McCain Edwards Undecided
Feb.-Mar. 2007 6 46.5 42 11.5 5 44 43 13
January 2007 5 45 44 11 6 45 44 11
Nov.-Dec. 2006 2 49.5 41 9.5 2 46.5 42 11.5

I see a little more volatility in pairings involving Edwards than in those involving Clinton--which could be because Edwards is considerably less well known and also less polarizing. Also, perhaps, because of the smaller number of polls involved. Edwards supporters could argue that for the most part, the margins Giuliani and McCain have over him are smaller than those they show over Clinton and Obama. But note that Edwards is not topping 44 percent of the vote, at a time when the generic preference is for the candidate of the Democratic Party. That's true, also, of the other candidates. Giuliani and to a lesser extent McCain are running ahead of the generic Republican; Clinton, Obama, and Edwards even with or behind the generic Democrat.

Iranian Defectors

Here's some very interesting speculation by Beirut-based Michael Young on the defection of the Iranian former Deputy Defense Minister Ali Reza Asgari. And here's a report from the Wall Street Journal on the investigation of the murder of Rafiq Hariri (via Instapundit).

By Michael Barone