Home » all category posts » USA » USA: Campaign 2008

Explaining Why the Daily Kos Tracking Poll Tilts Obama


If you're among the poll obsessed like me and click for the latest polls several times a day at Pollster.com and Real Clear Politics, you may have noticed that the Daily Kos tracking poll appears the most favorable to Obama.

Today, the tracking poll, conducted by Research 2000, has Obama leading McCain 49-43, or by 6 points. In contrast Gallup Tracking has Obama up by four points, as does Diageo/the Hotline (.pdf file). The Rasmussan tracking poll has the race tied. One tracking poll even has McCain up by 2 points, though I can't tell if this is just for the Battleground states or not.

One might be quick to say that the difference between the Kos/R2K poll and the others is within the margin of error--and you'd be right. On the other hand, the Kos/R2K poll has consistently favored Obama over the past few days.

Kos laudably releases the internals of the poll which show the demographic composition of the poll. I was surprised to see that Latinos comprise 13 percent of the sample. Now, Latinos compose around 13 percent of the population, so you might say that makes sense.

However, many Latinos are non-citizens. The share of non-citizens is especially high among the voting-age population. A much higher share of Latinos not of voting age are citizens because they were born in the U.S., and thus are citizens by right under our Constitution even if their parents are not.

Latinos skew around 2-1 for Obama, so including too high a share of Latinos might skew the survey in favor of Obama and against McCain. I don't know what share Latinos compose in many other polls--it's 8% in the tracking poll which favors McCain. Two top experts in Latino politics tell me that Latinos will probably compose around 8 percent of the electorate, confirming my sense that the share of Latinos in the Daily Kos poll is too high.

I'd be thrilled if the Kos/R2K poll were on the money but the oversampling of Latinos may well explain Obama's edge on Big Orange.

(Crossposted on Maryland Politics Watch.)


David Lublin September 19, 2008 - 12:00am
David Lublin's diary | printer friendly version | digg
( categories: USA: Campaign 2008 )

Do some googling

Census numbers are based on legal entities.

If you want to include illegal immigrants, then add about 1% to the census numbers.

/me shakes head...

If it sounds mildly racist, perhaps do just a leeeeeetle bit more research hmmm?

shah8 September 19, 2008 - 12:21am
login to post comments
Let's Think About This

The Census does undercount people who aren't in the country legally though it does get more of them than you might think. It helps that the Census cannot be used for other purposes.

Let's do a little thought experiment. Assume that there are 8 billion illegal immigrants/undocumented people of non-citizen Latino origin in the country. How many more uncounted Latino voters does this add to the electorate?

As I said, some of the best experts in the country on Latino politics tell me that Latinos will compose around 8 percent of the electorate this year. My guess is that it's not racism but the facts which are driving the estimates of these (surprise!) Latino scholars.

David Lublin September 19, 2008 - 12:41am
login to post comments
????

If you want to shift the goalposts from size of population to size of electorate...fine, whatever.

I just protested at the eeeeeeek! Illegals Voting! angle of the post, which was, as I have said before, slightly racist. Thus, I decline to participate in your thought experiment, since I cannot tell if you actually being serious. Why would I think of illegal immigrants as voters? Do you think there is vote fraud going on?

There are plenty of posts out there on the intertubes about how latinos have trouble with vote suppression tactics--with one of the motives being that they must be illegal.

No, David, your attitude is more than a little racist, if it weren't, you'd have started the post off with Daily Kos poll has 13% of latino respondents while [cited] believes that only about 8% of the electorate is latino...

shah8 September 19, 2008 - 9:13am
login to post comments
Talking Past Each Other

I'm not trying to raise the specter of illegal immigrants voting. I'm merely pointing out that the share of eligible Latinos is much lower than the share of Latinos in the population. The problem with the Kos poll is that Latino voters form too high a proportion of the sample relative to the share that they will make up on Election Day which skews the poll--not that there are too many Latinos voting. While you can non-productively go on a witch-hunt for racism and ethnic bias (now I suppose you'll say I hate witches), it doesn't change the facts or that not a single one of the many Latino scholars who saw the same post had your reaction.

David Lublin September 19, 2008 - 10:20am
login to post comments
Dangerous stuff

I am suspicious that Daily Kos serves what the customers want to eat, even if it is unhealthy.

I recommend Intrade and FiveThirtyEight. Intrade reflects the opinions of insiders having access to better data and secret polls. FiveThirtyEight is run by a guy leaning toward democrats - he has an amateur model but no good data - I mean he has only public data. FiveThirtyEight has a good insight on unreliable and reliable pollsters.

If you see anything public information which is 1-2 day-old and disagrees with Intrade odds, the public information is wrong at the probability level of 98%-99%. A pollster which has a continuous bias is either a crook or an idiot.

I have got an education story about a crook pollster. Once upon a time in a faraway country a pollster always underestimated the numbers of a conservative party. Why? Because the conservative voters are known to be motivated to vote if they think that they are about to lose.

The pollster is still in business and nowadays it releases figures in which the conservative party is exactly at even with it opponents. I don't trust the pollster still.

And in another faraway country the self-defined smart leftie voter let herself be outsmarted by herself with early exit polls which predicted that some Kerry was about to win. Thus, why to bother to vote in the afternoon?

Singular September 20, 2008 - 9:45am
login to post comments

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.