| By Matthew Struhar - Jul 31st, 2008 at 11:09 am EDT |
Analysis of polls usually doesn't give voters very much credit. Perhaps there is a recent poll in Ohio has Obama up two points, 48 to 46. Let's assume that that is a poll of likely voters and that it is accurate. Polls are usually conducted with 95% certainty -- that means we can be 95% certain that public opinion falls within the margin of error -- so, Obama could be ahead 51 to 43, if the margin of error is plus or minus three percentage points, or he could be losing 49 to 45. Most political reporters see the margin of error as something added to the overall poll, that you can only add or subtract up to three points from one number, or that you can only add or subtract three points from the total. Rather, you can add or subtract up to three points on both numbers, but anything you add to one you subtract from the other because you can't really have more than 100% of one thing.
For example: most political reporters would say that Obama could be winning 51 to 46, but actually he could be winning 51 to 43.
Back to our example. A conclusion political reporters, even opinion-making elites working at publications such as The Atlantic, will often make is that Obama has not experienced a post-Eurotrip bounce in Ohio, and that it shows the trip was ineffective. That's all kinds of dumb. 6% of likely Ohio voters have not made up their mind in our hypothetical (that's a number also subject to the margin of error but to go into all the possibilities that implies would take too much time). The assumption might be that the trip didn't work, but actually it might worked quite well with these voters. Perhaps a lot of the undecideds now have a more positive opinion about Obama but also have a positive opinion about McCain, or vice/versa.
Also, the fact that this has been a close race has caused political reporters to suggest that the Obama hype is overrated. Overrated among who? Again, voters aren't given enough credit. A lot of voters could have a very positive opinion about Obama but simply agree more with John McCain on the issues or believe that John McCain better reflects their interests.
This also demonstrates how the media sets the agenda regarding expectations. Because Obama has such high expectations, even if he wins a close election he'll lose.
Of course, as far as the media are concerned, John McCain never loses.
(Both links via Matthew Yglesias.)

Comments are closed for this post.