[Atlas Sneezed]

I considered that at first, but ruled it out after looking at NM. The most populous county in NM is Bernalillo, by a HUGE margin, yet it's only slightly raised and actually shows as lower than some of the other counties. (it went just SLIGHTLY Kerry)

---

I still think it is height=population, but I agree that Albuquerque (Bernalillo County) and Reno (Washoe County) appear to be missing. You shouldn't be able to hide 500,000-1,000,000+ people that easily.

[PrtzlLogic]

If it were height=population, Bernalillo would be the tallest in NM -- because it's not only the most populous, it's also the smallest and most densely-populated. I'm pretty sure it's the margin of winning votes (i.e.- the "density" of the winning vote margin), which is what I was attempting to say, albeit not as clearly as I could have. I don't think it has anything to do with population, in and of itself -- other than in how that population contributes to winning vote margins. It's almost easier to draw it:

EX (line represents height on graph):

2 million votes, 1 million vote margin for K/E: ==============================

2 million votes, 34,000 vote margin for K/E: =

[Aggie Mama]

I don't think height=population. Look at the Houston area (4th largest city in the country)-- it should be very tall. Conservatives live on the outskirts and Shelia Jackson Lee's constitutents live in the city. I think there was a good sized vote for Kerry in her district.

[puppetz]

Maybe its highest percentages for a particular candidate: if Kerry got 90% of the vote in that county then it would be very high.

[ThePythonicCow]

After looking at the following map, I suspect that the bars are population per county. The map is found at http://www.esri.com/censuswatch/graphics/census2000_p9_lg.gif.