Monday, September 21, 2009

About that New Jersey Governors Poll, Part III

FHQ promised to revisit the recently released Neighborhood Research poll and its August predecessor and gauge the impact those polls would have on the race if they were included in our graduated weighted averages of the polling in the New Jersey governors race. I was going to go the whole nine yards and include a mock-up of the usual graphic I've been putting up with our updates of the New Jersey and Virginia contests. In the interest of clarity, though, I'll hold it to a simple numbers-to-numbers comparison.

I don't think anyone will be surprised by the fact that these polls hurt Chris Christie more than Jon Corzine. That the Republican has pulled in only 37% support among the likely voters in both polls is indicative of how accurate they are. Even when those polls are included amongst the others, they are barely within two standard deviations of the unweighted polling average. Corzine's distribution is much more tightly clustered. Every single one of the 39 polls conducted in this particular match up since the first of the year is within plus or minus one standard deviation of the unweighted average.

I don't want to bog you down with statistical gobbledygook, so this is just a long way of saying that both these Neighborhood Research polls are outliers and both polls put more of a drag on Christie's numbers than on Corzine's.

How much? Well, to be fair, the polls aren't helping Corzine either (other than to decrease Christie's average support). But the incumbent only loses four tenths of a percentage point (from 37.8% to 37.4%) when these polls are added to our averages. However, the comparable figure for Christie is a loss of one point (46.4% to 45.4%). In other words, the polls have twice the negative impact on Christie as on Corzine (at least in terms of FHQ's averaging of the race).

Is that a big deal? Honestly, it isn't, but in an environment where a decreasing margin is expected to some extent (Corzine catching up due to registered Democrats outweighing Republicans in the Garden state), these polls have the effect of inaccurately deflating that gap.

And that, in a nutshell, is why FHQ is looking the other way when these polls are released.

...well, sort of...

Recent Posts:
State of the Race: Virginia Governor (9/20/09)

About that New Jersey Governors Poll, Part II

Huckabee Takes 2009 Value Voters Straw Poll

No comments: