Numbers Game: How to Read the Polls Now
Barack Obama's poll numbers have been looking good for so long that it is easy for his supporters to assume a triumphalism stance as America's longest-ever presidential campaign enters its final week.
But be careful about that. The Democratic nominee for president, while he is currently ahead of Republican John McCain, stands perilously close to a dangerous threshold.
How so?
First, a little recent history: In the Democratic primaries last winter and spring, Obama rarely ran better than his poll numbers. He either hit the figure he was at in pre-election polls (in states such as Wisconsin, Ohio and Pennsylvania) or fell a little below it (in states such as New Hampshire and California).
That's worrying because, while the Democrat's poll numbers now look strong by comparison with those produced for a stumbling McCain campaign, they still hover around the 50 percent line nationally and in a number of current and former battleground states.
The McCain camp is betting that primary patterns will hold and that Obama will finish little or no better than his pre-election poll numbers. They see that as their opening, on the theory that McCain will get his base polling figure in any particular state and an overwhelming portion of supposedly "undecided" voters.
To understand how the theory works, let's put the variables introduced by third-party candidates and other factors on hold and simply consider the one-on-one competition in the hotly-contested state of Florida.
As of Monday in Florida, the polling averages had Obama up with 47.7 to 45.8 for McCain. That leaves 6.5 percent undecided. McCain strategists bet their man gets three quarters of the supposedly undecided voters, while Obama takes the remainder. Final result: McCain 50.6 to Obama 49.4.
If the Republicans are right, this could still be a close election -- perhaps even a "Dewey Defeats Truman" upset election.
So, how worried should Obama backers be at this point?
The first answer is: A lot less worried than McCain backers.
The second answer is: There is still some argument for disquiet on the Democratic side.
Let's begin with the numbers we've got.
Various "poll of polls" surveys give Obama a solid national popular vote lead of 7.3. points -- 50.4 for the Democrat to 43.1 for Republican John McCain.
Of course, the United States does not hold national elections. But, on the surface at least, the state-by-state results of races for Electoral College votes are equally encouraging for Democrats.
The latest analysis of polls from all 50 states by Real Clear Politics Obama could win as many as 375 electoral votes, to 163 for Republican John McCain. That's a 212 vote advantage for the Democrat, a certifiable landslide if it happens.
The popular www.fivethirtyeight.com website puts Obama at 351 electoral votes.
So why worry?
McCain currently leads in polling from 19 states: Alabama, Alaska, Arizona, Arkansas, Georgia, Idaho, Kansas, Kentucky, Louisiana, Mississippi, Nebraska, Oklahoma, South Carolina, South Dakota, Tennessee, Texas, Utah, West Virginia and Wyoming. That adds up to 157 electoral votes.
Obama is ahead in 24 states -- California, Colorado, Connecticut, Delaware, Illinois, Iowa, Maine, Maryland, Massachusetts, Michigan, Minnesota, New Hampshire, New Jersey, New Mexico, New York, Ohio, Oregon, Pennsylvania, Rhode Island, Vermont, Virginia, Washington and Wisconsin -- and the District of Columbia. Total electoral votes: 306.
Obama is, as well, looking strong in a number of the remaining battleground states of Florida, Indiana, Missouri, Montana, Nevada, North Carolina and North Dakota, which have 75 electoral votes.
So far, so good.
But... Obama's still polling just above 50 percent nationally and at or below that level in not just battleground states but a number of states -- Ohio (49.9 percent), for instance, and New Mexico (50.7 percent) and New Hampshire (51 percent) -- that have been moved into his column by a number of analysts.
Let's assume that the various third-party and independent candidates, credible and appealing as they may be, do not make much of a dent in this year of celebrity campaigning and hyper-partisanship.
Then let's consider this scenario: As the presidential race closes in this final week, the competition narrows a bit. Obama's numbers tick down a bit and McCain's tick up.
Obama might still have substantial leads in national and state-based surveys -- four, five, six or more points ahead of McCain. But he could fall below the magic 50 percent figure.
That's the point at which to begin worrying.
There has been much discussion this year about the so-called "Bradley effect" -- the phenomenon, most common in the 1980s and early 1990s, of white voters telling pollsters they would back an African-American candidate, such as 1982 California gubernatorial candidate Tom Bradley -- and the prospect that it might play against Obama.
But what if the voters who are uncomfortable with Obama, for whatever reason, aren't saying that they will vote for or against him? What if they are crowding the undecided column? And what if they break at historically disproportional numbers for McCain on November 4?
There are plenty of counter-arguments: Polling doesn't capture the universe of cell-phone users, polling doesn't accurately assess likely youth turnout, polling can't offer an accurate take of the impact of Obama's community-organizing model for political mobilization or the extent to which this is an "event" election that will draw dramatically higher numbers of voters to the polls.
This writer's bet is still that Obama prevails. Indeed, there is good reason to believe that the polls could "open up" in the next few days -- as has frequently been the case in the past -- and give Obama the expanding lead that frontrunners often accumulate as a "go-with-the-winner" mood takes over at the close of a long campaign.
But we are now a nation of poll watchers. (Any why not? Survey research data is now far more comforting to reflect upon than stock market tickers.) Traffic on poll-aggregating websites is astronomical. And in this final week, when we are awash in data, it is important to read the numbers right.
And the number that matter is not Barack Obama's five-, six- or seven-point lead over John McCain.
If experience is an indicator, the number that matters for Obama is 50.1 percent – or, to be on the safe side, 51- or 52 percent.
No comments:
Post a Comment