Voting to commit delegates for the Republican National Convention will commence in just over two weeks with the Iowa caucuses, or, as Rush deems them, the Hawkeye cauci.
The RealClearPolitics.com poll summaries for the first three contests in Iowa, New Hampshire, and South Carolina, reflect in differing measures those for the nation as a whole and the differing constituencies represented in those states. New Hampshire and South Carolina are for their regions favorite son, and Iowa is... well, Iowa.
The state Republican parties that moved their primaries before March 1 have chosen to risk that only half their delegates will be seated at the Tampa convention. That national party rule seems not to apply to caucus states, so Iowa's delegates are not in jeopardy. These early states have chosen early influence over the electorate over direct influence over the vote in Tampa. That choice makes sense, since they wouldn't have much of the latter, anyway.
The general election horse-race popular vote polls are amusing, but their conclusions must be tempered with some knowledge. It is certainly important to know what population the sample was drawn from. In increasing order of conservative voting tendency, common populations are all adults, registered voters, and likely voters, with about 2.5% change between each. For example, If a national poll of all adults shows Obama leading Mitt Romney by 50% to 40%, you can estimate that the two would be tied at 45% in a poll of likely voters.
Probably more important is that we don't elect the President by popular vote, but rather by a vote of electors in the electoral college with votes allocated to states, one for each member of the House of representatives and one for each senator. Since House seats are divvied up by population, but Senate seats are two per state, this has the effect of giving unequal weight to the votes of people in states with different populations, favoring voters in states with smaller population.
Though polling all fifty states individually is necessary for the best prediction of the electoral college result, that's too expensive to be done until next summer. Till then, the next best thing is a poll of voters in swing states. The polling organization separates states into three groups: those that are almost certain to vote for Obama, those that are similarly certain to vote for almost any breathing Republican, and those that aren't so clear, referred to as swing states. The organization then polls those swing states on the horse race question: If the election were held today, would you vote for Obama or for Old Stewball? You're right, John McCain is not running this time, but you get my drift. This approach allows the polling organization to take into account and, in effect, to ignore the large popular vote majorities racked up in the very blue states like New York, and California, similarly in the very red states like Texas.
Here is the first swing state poll of which I'm aware, done by Gallup for USA Today a couple of weeks ago. Gallup's choice of swing states, sort of counter clockwise on a map of the US from lower right, is Florida, New Hampshire, North Carolina, Virginia, Pennsylvania, Ohio, Michigan, Wisconsin, Iowa, Colorado, Nevada, and New Mexico. The poll used random digit dialing techniques to call registered voters with land lines and cell phones. You caught "registered voters" right?
In those states, Democrat party self-identification is down 4 percentage points from 2008, while that of Republicans is up 5 percentage points. You may note that while North Carolina and Virginia are traditional Republican strongholds, Pennsylvania, Michigan, Wisconsin are bigger Democrat strongholds.
The electoral vote totals for the sure Obama states and the sure Republican states are almost equal at196 and 191 respectively. The swing states hold the balance of power at 158 electoral votes out of 270 needed to win.
In this case, the swing state popular vote preference is Romney over Obama 48% to 43%, and Gingrich over Obama 48% to 45%. If we apply our 2.5% correction the "likely voter" popular vote preference would likely be for Mitt over Barack by... wait for it... 10%! Even the more generally conservative Newt would likely win it by 8%.
So, Obama's not toast, but he is in terrible shape for an incumbent. Without a major improvement in the economy -- something an overwhelming majority of economists think very unlikely -- he will have a great deal of difficulty getting reelected. And his unpopularity will likely drag down Democrat performance in Senate and House races across the country.
It couldn't happen to a nicer party.
