Technical Discussion of Robo-Rasmussen Polls

[CARLHAYDEN]

From what I hear, Scott Rasmussen was mortified by the 2000 results and has recalibrated his polls to include a larger Democrat component than used in 2000.

I believe that this is a mistake.

In 2000 Gore had an unusually effective GOTV drive and Bush had essentialy none.

This year Bush appears to have an unsually effective GOTV drive in gear, and Kerry hasn't really got his act together on this one (its still five months till the election, plenty of time).

The bottom line is that if Kerry has a GOTV drive as effective as Gore had, due to the Bush efforts in this area this year, the 1-2 point advantage Gore accrued in 2000 will be eliminated, and if Kerry doesn't have such an effective drive then Bush may actually acrue a 2-4 point advantage (something Scott has not considered).

Also, Scott is deliberately omitting Nader from his results, which results in a slight skewing of the vote in Kerry's favor.

Finally, on a personal note, when I got home last night and played my messages on my answering machine, one was for a poll sponsored by a television station.  Wonder if I'll get a 'call-back.'  (Yes, I think Bush voters are being undersampled)

[CARLHAYDEN]

Actually, there is a reason Scott doesn't provided the weighting formula.

Its not merely partisan, but racial weighting.

In 1996 (the first experiments). 1998 (rather sucessful in the Congressional results), and 2000 (spectacularly sucessful results) the Democrats herded large numbers of blacks to the polls.

While Blacks are generally Democrats, they are more inflexible than other Democrat racial groups.

For example, Bush is typically taking a fifth of the white self-described Democrats, but only about a twentieth of the black self-described Democrats.

[CARLHAYDEN]

Frankly there is no hard evidence that early reporting of exit polls influences the results in the West.

Moreover, with the rapid recent rise in absentee voting, a high percentage of the vote is cast where it cann't be 'polled,' upon leaving (exiting) the 'polling place.'

Just ask "govenor" bradley (of California) about this.

My major complaint is that televison spends too much time with moronic anchors droning on with nonsensical hogwash (rather is the worst of this) and so-called expert analysts who frequently don't know what they're talking about.

Instead, I'd like to see more raw data, clearly labled as to source (absentee vs. precinct), by county, etc.

[CARLHAYDEN]

While you have established that the Democrat component of Scott's sample is slightly higher than you originally suspected, did you ever get the data on the black percentage of the sample?

If it is a high as I suspect, then Scott is overstating (compared to what my models suggest for the 2004 turnout) that Kerry is a point or two higher, and Bush a point or two higher.

If this is the case, then a Rasmussen tie, would actually mean a Bush lead of two to four points.

Also, as you have previously correctly noted, summer polls tend to favor Democrat candidates.  I have gotten home twice in the past couple of months and found messages on my answering machine for a robo poll from a local TV station.  

[CARLHAYDEN]

Thank you for the info.

It pretty much confirms what I was saying.

Scott's percentage of respondents who are black is slightly higher than the actual turnout in 2000, in which election (due to extraordinary get out the vote drive by Democrats) the percentage of black voters was far higher than usual.

If black turnout returns to the normal percentage of the electorate, and if we preseme a normal party distribution of the vote by race, then Scott will have overstated Kerry's support by just approximately two per cent, and understated Bush's support by the same extent.

[CARLHAYDEN]

First, I stongly suggest that you don't get the Japanese talking about the Koreans, or the Koreans talking about the Japanese.  They do get a little bit heated (personally, I sympathize with the Koreans).

Second, the voting patterns of Asians in 2000 were heavily influenced by two factors.  In Hawaii, Asians vote Democrat since the Republican party in that state has not treated them well historically (@#$% country club Republicans).  In California, a lot of Koreans vote Republican in Presidential elections because the Democrat party seems more interested in other racial groups.  Further, the 2000 election had a higher than normal Democrat vote for President as Gore's campaign worked Asian groups better than Bush's campaign did.

Third, Bush is picking up support (relative to 2000) among hispanics.  While Kerry will probably carry this group nationally (largely due to the overwhelming advantage among Puerto Ricans), I expect the overall advantage to be a historically low 60-40, and possibly as good as 55-45.  

Fourth, while Native Americans (indians) only account for about one per cent of the electorate, Gore did unusually well among this group in 2000 because (again), his campaign worked this group, and Bush did not.  Expect Bush will take a larger percentage of this group this year than he did in 2000.