Friday, August 17, 2007

Deconstructing Dowd Bodack

I think we are sufficiently past the election to deconstruct what happened in the District 7 race where Pat Dowd defeated Len Bodack in a close race. There isn't anyone else running in the fall election for the seat.... so barring something cataclysmic the race is over. There certainly has been a lot of talk about what that race meant, but also a lot of hypotheses out there about how it happened. Note I'd try to do this for District 3 where Kraus beat Koch, but there isn't as clear a story to tell The earlier race where Koch was first elected was one of those odd special elections with a slew of wanna-be councilpersons which makes a dissection like this a little problematic.

Here is what the Dowd-Bodack results looked like in the Spring. A few caveats: yes, the color is awful, but given that this was a primary I wanted to avoid the red/blue schema. Also I don't seem to have a shape for cemeteries so for those who know the area, the large Allegheny Cemetery gives a distorted picture of one of the Central Lawrenceville voting districts. and I picked Bodack percentages for the legend just because he is the common denominator in the two races I am comparing:

Why did Dowd win? You first have to go back and look at the results 4 years earlier. In 2003 there were 3 candidates. Len Bodack defeated Mitch Kates in a similarly close race with a third candidate, Nancy Noszka getting a sizable number of votes as well. Many thought the 2007 race was a race for the votes that had supported Nancy Noszka. To understand 2007 I think you first need to look at the 2003 results with the Noszka votes taken out. That gives you this map:

This map looks awfully similar to what happened 4 years later. If the Noszka votes had split one way or the other, the 2007 race would have been a blowout whichever way they leaned. To see if they did, I put together this map of the change in votes for Bodack between the two races (2003 vs. 2007), again with the Noszka results excluded. That gives you this map:

This map may be surprising and counterintuitive. It pretty much says that Bodack increased his result in Highland Park areas while Dowd improved upon what Kates did in a lot of Lawrenceville areas. Surprised? All that is showing is that the Noszka votes split pretty evenly between the two candidates in 2007. The extreme Kates districts were not as extreme for Dowd and the extreme Bodack districts were less so in 2007. I think there had been a belief that if Noszka had not been in the race, Kates would have beaten Bodack handily... this difference map would disputes that hypothesis pretty thoroughly. In other words, if Noszka had been in the race it clearly would still have been an amazingly close race and it isn't even clear if it would have benefited one candidate (Bodack or Kates) over the other.

So again, how did Dowd beat Bodack? It is pretty clear that the core Lawrenceville and Highland Park voters voted nearly identically in 2007 as they did in 2003 and the Noszka supporters split evenly. I have heard some people say it was a big bump up in turnout that produced the different results.... yet that just isn't in the data at all. Total votes cast in 2007 were all of 1.4% higher in 2007 compared to 4 years earlier (6,175 votes vs 6,085). That's it. More surprisingly is that the turnout was nearly identical ward by ward between the two elections. In all of the 11th ward (Highland Park for dist 7) there were a total of 24 more votes cast this time around. 24! No ward differed by more than a 100 votes, up or down between 03 and 07. If one side targeted voters better than the other, that could have made the difference, but there clearly was not any noticeable change in overall turnout whatsoever.

What seems clear is that over the course of 4 years, Bodack neither won any new supporters nor lost any old ones. So in the end I can't answer the question. Whichever side you are on, it really was democracy in action.... just a few votes really made the difference in both races. The lesson may be that Pittsburgh voting patterns are a lot more consistent than many think they are. The result for District 7 is considered a big change in Pittsburgh City Politics. I would agree, but it's not because of big changes in how people voted. 99% of the district voted just as could have been predicted years ago, but it was that last 1% that made the difference.

update: apologies in that I didnt realize I had hit the button that disallowed comments. I have corrected.


Anonymous Anonymous said...

Great breakdown Chris. I think about this a lot in my 5th district. Without studying the numbers, it seems fair to assume that there is a similar split between Squirrel Hill and Greenfield. How to break/bridge that split, especially in light of the long-time dominance of O'Connor/Shields, is an interesting challenge.

Friday, August 17, 2007 8:25:00 PM  
Blogger Mark Rauterkus said...

The race for the 3rd city council district seat is NOT over as is the race for the 7th. There is a general election for us in the 3rd. We do have a candidate for the 7th -- Libertarian chairman, Dave Powell. But he is scared a great deal about all the ballot hurdles we face. And, since P.D. won -- we can all relax.

Friday, August 17, 2007 9:47:00 PM  
Blogger Smitty said...

where is North Oakland..Oakland has Peduto,Koch and Tonna Pain as districts in the 4th ward are represented in Bodack/Dowd totals???

BTW a Dowd supporter told me that a last minute effort by Bodack to doorknock woulda won the race for him...your analysis seems to point in that direction

Saturday, August 18, 2007 12:46:00 AM  
Blogger C. Briem said...

I am reminded of Diana Nelson Jones’ article on imprecise neighborhood names in Pittsburgh.

But the North Oakland label comes from Ward 5 District 9 which is part of the North Oakland neighborhood (both technically and common usage I would argue), and also in city council district 7. There is a longer wonkish history on why that may be confusing, but basically you half correct, there are no 4th ward precincts in Council District 7, but North Oakland has at least that one precinct from the 5th ward.

I really dunno about the 2nd question. Maybe, though I am not so convinced doorknocking changes many minds about candidates.. or at least not as many as people think it does. I kind of thought the conclusion from those maps was that the divisions are pretty consistent. How many doors do you have to knock on to find an actual voter... what is the probability that person does not already have a clear idea of who they are voting for and then what is the probability that you or your minions can change their mind? Gets to be a low % for the work involved.

Ask a professional and they will tell you uou door knock doors more to identify the voters you want to bring out on election day than to convince them to vote for you. In that sense, no I don't think LB just deciding to go walk a few more blocks would have made any difference. If they had had a comprehensive system to identify and get voters out on election day in place from the beginning, it could have been even closer than it was..

Saturday, August 18, 2007 1:33:00 PM  
Blogger C. Briem said...

on the comment about District 5.. just fyi I have put up maps in the past for the 2003 District 5 results for Shields, Sciulli and Smith.

ok ok... sorry Mark. Yes there is an election coming up for District 3.. but even by the standards of 3rd party candidates, I am not sure what traction he will have. I can see a few angry Koch voters could produce something in the numbers, but that is about it.

but this brings up a question I always have had that maybe Mark will address. As much as political talk in town is usually about divisions within the local Democratic party, the truth is that the local Republican party is by more beset by internal divisions. You have to wonder if that is more the reason for the wide margin of Democratic victories in a lot of local races. Some of those Republican party divisions seems to be ideological (the personal ones are ubiquitous of course.. just the way it is in all parties)... but there is a clearly libertarian element in the local conservative base. Yet none of that support really translates into local libertarian party success inthe polls, either candidates, money or just plain votes. That's not meant to be too negative there on you Mark, but is there a reason that is not obvious?

Saturday, August 18, 2007 8:52:00 PM  
Blogger yanmaneee said...

pg 1
yeezy boost 350 v2
nike off white
curry shoes
cheap jordans
golden goose
supreme new york
golden goose sneakers
yeezy 350
jordan 6

Friday, June 25, 2021 1:53:00 PM  

Post a Comment

<< Home