« NC Trending Purple? | Main | So Where Are We? »

More on Labor Day Leads

I did a recent post arguing that Gallup's piece on labor day leads was misleading since they compared Bush's LV lead this year to (mostly) RV leads from previous races. They should have, I argued, compared apples-to-apples, RVs to RVs.

Frank Newport of Gallup was sporting enough to print an edited version of my comments on their editors' blog, along with his reply. In Newport's reply, his key rationale for conducting their analysis the way he did was he wished to "[use] Gallupís best available estimates at Labor Day for each year for which we have data." (But by all means read his argument in full through the link.)

To further this discussion, here are some additional remarks on the issue replying to Newport's argument. I should add that I don't believe that Gallup has any particular axe to grind in how they did this analysis--I just think in this case they got it wrong.

Thanks for your thoughtful reply. But I still don't buy it. You surely must see that it makes a difference when people read these analyses with "7 point deficit to overcome" in mind rather than "1 point deficit to overcome"? And in fact that's how your analysis was written, focusing reader attention on the 7 point LV deficit.

And the fact remains that apples-to-apples comparisons are far preferable to apples-to-oranges comparisons. Therefore the proper comparison is between this year's RV labor day results and previous years'. Otherwise, you are not analyzing the same change (RV labor day gap vs. final gap) across years. (By the way, thanks for drawing my attention to the national adults samples prior to 1952. In the same apples-to-apples spirit, I would drop these cases from the analysis.)

Using a consistent time series would make a difference to your analysis.

Instead of:

"In summary, the history of presidential elections since 1936 suggests that in about half of the cases, the type of gap change that would be necessary for Kerry to tie or move ahead of Bush has occurred. About half the time it has not. If a gap change does occur, the odds are higher than 50-50 that it would be in Kerry's direction (i.e., a shrinkage rather than an expansion of Bush's current lead). "

You would have:

"In summary, the history of presidential elections since 1952 suggests that in all cases, the type of gap change that would be necessary for Kerry to tie or move ahead of Bush has occurred. If a gap change does occur, the odds are very strong (11 out of 13) that it would be in Kerry's direction (i.e., a shrinkage rather than an expansion of Bush's current lead)."

This clearly sounds quite a bit different. And thinking Kerry is behind by one point, rather than 7 points, clearly makes a big difference when considering elections like 1960 and 1980, which loom large in your analysis. Kennedy was behind by a point in 1960 among RVs--the same as Kerry--and Reagan was behind by 4 points in 1980--more than Kerry. If you're thinking 7 points behind, those races look a lot different.

In short, lacking a consistent time series of any length on LVs, you just shouldn't use 'em in an analysis like this.

Comments

This is a critical issue, because the methodology used by Gallup (and some others) generates results that in turn DRIVE THE MEDIA SCRIPT. The campaign stories in the mdia have ALREADY turned from "too close to call" to "Bush surging, Kerry struggling." Those stories can and likely will affect some voters' intentions, pushing some "undecideds" to move into Bush's column because they perceive of him as a "winner." And that effect will in turn provide the pretext for stories along the lines of, "Can Kerry Come Back?" when the reality (according to your analysis, Ruy) is that he's really not all that far back to begin with. So the effect is not just cosmetic or an appropriate topic for academic debate; it's potentially pernicious and decisive. Given the stakes, I hope that you will continue to call Gallup and others out on their disingenuous use of data that may have a self-fulfilling-prophecy effect on this critical election.

I agree with Doofus; choosing to highlight one set of results drives the media narrative, which entrenches public opinion, thus having an effect on voter intention. It is really quite disturbing.

I think one other factor makes the Labor Day results less relevant this year than in other years. Usually by Labor Day, both parties have had their conventions sufficiently in the past that the bounce has had time to recede. This year, not so. Measurements on Labor Day reflected not only all the inherent problems Ruy has described, but there was the mask of the bright glow of the Republican's gauzy convention, too. It has occurred to me that Rove is just superstitious enough that he wanted the convention bounce to be on Labor Day to create numbers that weekend which would predict a Bush win.

It's also worth noting that lade deciders almost always break for the challenging party's candidate. In 1996, most of the later deciders picked Dole. In 1988, most of them picked Dukakis. The same happened in 1980, when a staggering 60% of late deciders picked Reagan (or defected from Anderson to Reagan). The only exception to this was in 2000, when many Nader supporters switched at the last minute to Gore (but given that their switch was from a 3rd party, an excpetion was not surprising).

That Bush is tued this late in the race is bad news for him. If his incumbency were secured, he could afford to just ignore his challenger (as Reagan did in '84 and as Clinton did in '96). That fact that he even has to campaign says something.