chrisishardcore.com

Comments: The Dreaded Flag Debate

Let's hear it.

Posted by Chris at May 30, 2003 05:41 AM

I'll feed your data to a logit model in a few weeks (out of country, wife finishing grad school, moving, life ...). I'm a bit worried about sample selection though. How'd you pick your counties?

Posted by kog at May 30, 2003 03:29 PM

Since I don't currently have precinct by precinct data (I don't want to pay $600 for it), I can only analyze data by county, as that is how it is presented on the secretary of state's website.

Therefor, I picked counties which were wholly represented by a single senator and/or representative. This precludes counties such as DeKalb, Fulton, and Gwinnett, because I can get a total state rep/senator vote by adding up all the votes for different districts in the county, but there's no way for me to tell which individual districts the voters in the statewide districts live in.

Basically, counties that are wholly represented by a single state senator have less than 145,000 people living in them and counties that are wholly represented by a single state rep have less than 45,000 or so. Plus or minus 5%, of course.

Posted by Chris at May 30, 2003 03:51 PM

I think impact of the flag issue had just as much to do with Barnes' loss as the impact of the "war" on terrorism. Georgians tend to be slightly more conservative than the rest of the country when it comes to war and such.

My point is that I don't believe you have presented conclusive evidence that the flag cost Barnes the election. I will grant you that it played a partial role, though. :-)

Posted by Paul at May 30, 2003 04:14 PM

Thanks for the clarification on sample selection. But why won't Cathy just give you the data for free? :-)

Posted by kog at May 30, 2003 08:03 PM

There was some trouble about people not paying recently. Or more specifically, about Democrats not paying for the same data Republicans shelled out for. I tell you what, I'll talk to some state party people I know (and some elected officials) and see if they won't let me have the precinct by precinct data.

If worst comes to worst, I can just call all the county election offices and get them to fax it to me!

I do know that the parties may not be able to give this data away, though. We'll see.

Posted by Chris at May 30, 2003 08:14 PM

Let me rephrase -- I think the flag issue is an effect, not a cause. There is something else, like Georgia voters' general political ideologies, or the general ideologies of the candidates, that correlates them along the flag issue. The flag issue, I would suggest, is no more of a cause of any particular election results than any other single issue, though.

Posted by Paul at May 31, 2003 09:06 AM

That begs the question of why Perdue would touch it ? Is he a neoconfederate himself ? And if so why wouldn't he come out for an actual flag design ?

My take is that the other conditions (like a popular president coming to ATL every weekend to campaign for Sonny and Saxby, teachers being really pissed, redistricting and the "King Roy" BS) would have given Sonny a chance without the flag, and if you watched his campaign they tried to drop it [flag] towards the end when they realized that fact. but the damage was done. They weaned themselves on the milk of the flag and by the time they were ready for solid food, they couldn't back away from it.

So I'm camp #2. If everything else went Barnes' way the flag itself wouldn't have done it. But flag plus anything else wins it for Perdue.

Nice weblog, Chris, thanks for running the numbers. I'll be back.

Posted by Wes at June 1, 2003 01:43 AM

I'm still in group #3, but the numbers shown here have definitely shifted me toward two (although I am an independent voter). I still believe that Max losing shows an overall rightward shift in the entire state, regardless of the flag. We shall have a better idea in a few years.

Nice post Chris.

Posted by Ron C at June 2, 2003 10:55 PM

Post a comment