Monday, July 27, 2009

The missing female voter politician

The PG passed on some factoids from a state commission about how women in Pennsylvania actually make up more of the vote than men. 54% was the factoid. It's not anything surprising. In fact I am pretty sure that percentage would be even higher (as in more skewed toward women) for most local elections here. The companion facts all point out the low female participation among elected offices in the state; something that has been documented for a long time as well.

But are they facts that speak for themselves? Depends what you think they are saying. I am pretty sure in this case the answer is no. I am sure a lot of folks jump to the conclusion that women are just that more politically active than men. Of course that just isn’t the thing driving that 54% number. I think I’ve beaten to death the fact that we are a very old voting demographic. (my details are here, here, and here just to begin with )Far older than the overall population demographic would imply. Now add in the fact that women live longer than men. Put it together and what do you get, that a lot more of the people making it to vote happen to be women. Those supervoters in particular are going to be a female demographic. There just are more females than men among the folks far more likely to vote. What that 54% does NOT mean is that if you pick an average person of a given age out there that the probability a woman votes is necessarily different from the voting bahavior of comparable men.

While I have some longer parsing of the local electorate, you really can look at in much simpler terms that get you to the 80% solution. The center of gravity in most local elections is clearly women over age 60. Strong support in that one group and very little else matters.

Then there is the further implication of the factoid. Women are the ones voting, but women are not in public office. One might think that the entire issue is women not being able to get elected. Of course that appears to not be the case around here. Years ago I had some long comments on the decision by one-term state rep Lisa Bennington to not seek a 2nd term. That was after defeating long time incumbent Frank Pistella which then had implications for female representation in the state general assembly. More recently I am sure each had very solid reasons for their decisions, it remains amazingly rare events for politicians to walk away from office voluntarily, let alone with no scandal involved. To have two such local cases recently is almost impossible statistically. Yet in both cases it was the female candidate who won contested elections. In Bennington’s case a race against a fairly well liked long term incumbent, says it is not entirely a bias against women that is the issue.

No time to recompile it right now, but the pie chart I did breaking out voters in the spring primary by age would be an even stranger result if I did it just for women. Strange as in skewed even more toward older voters if that is even possible. I said 5 times as many voters age 60 cast a ballot in the spring primary compared to the voters under 30. I will lay good odds that the same ratio would be more like 10 to 1 if calculated just for women. Think about that just a bit.

But here is one thing. From the spring primary in the city, the ratio of female voters to male voters among the actual ballots cast, by age, looks like this:

and I have to say I have no idea what is going on with more men among the youngest voters. Never thought about that much and I will try and see what is causing that number and make sure it's correct. I was reading a book from a decade ago titled NonVoters, but I don't remember any discussion about that in particular. Anyone know what is causing that? or why it may be appearing that way in the data.


Blogger fester said...

Pulling a WAG on the older population ratio of female:male voting disparity --- my guess is lower mortality and better mobility for older women compared to older men --- just more older women who are somewhat mobile...

Now for the youngest voters, why do men out-vote women 2:1 at the age of 20-25ish --- I have no idea nor even a decent WAG.

Monday, July 27, 2009 2:28:00 PM  
Anonymous MH said...

Possibly more of the young women were students and they were gone for spring break? Matt H has lots of friends that are all male? Or the fact that you can get funny results from small samples, which could at least explain the .2 ratio for the very youngest.

Monday, July 27, 2009 2:40:00 PM  
Anonymous Anonymous said...

Is WAG Wild Ass Guess?

Monday, July 27, 2009 4:49:00 PM  
Blogger C. Briem said...

You're probably right. I do think Matt H. may be impacting those numbers all by himself.


a more precise number would be a SWAG

The mobility question separate from mortality is actually kind of an interesting question for someone to look at.

Monday, July 27, 2009 5:31:00 PM  
Blogger Gloria said...

I wonder if younger women aren't voting in larger numbers b/c they're in their child bearing years, add to that the fact that many have to work & (even in this day & age), they assume more of the domestic duties than their male husbands/companions do?

More interesting to me is whether the older, female voter is more likely to vote for a woman candidate than a man.

Monday, July 27, 2009 6:13:00 PM  
Blogger Unknown said...

I agree with Gloria. What percentage of female voters voted for the female candidate for Mayor?

Monday, July 27, 2009 9:57:00 PM  
Blogger C. Briem said...

This is entirely gut feel on my part.. I have no numbers to support it.. but I think older women are pretty tough on younger female candidates. Maybe a bit less harsh on older female candidates, but those are few and far between.

Monday, July 27, 2009 10:43:00 PM  
Anonymous Anonymous said...

Recent local races featuring women come to mind: Georgia Berner and Jason Altmire Congressional primary up in Beaver Co. a few cycles ago; Rudiak and another, older woman and a guy in Dist. 2 City Council; and 3 women just ran for school board in Dist. 1. Wonder if there's some way to suss out the womens' vote in these races, although there are probably serious confounding variables like race and low turnout(school board) and low turnout (city council).

Tuesday, July 28, 2009 8:56:00 AM  
Anonymous MH said...

I was just surveyed about the mayoral race. This also happened for the DeSantis election. So, somebody may be sharing lists or it may be the same person as last time (or maybe somebody wants a Republican in their sample and there aren't any others). The poll took at least 10 minutes, was definitely not a push pull, and was clearly designed by somebody with experience. Unfortunately, they also trained their interviewers not to drop hints as to who paid for the poll.

On the one hand, who has that kind of money besides Luke? But, it didn't seem like a poll designed for Luke. It is hard to see anybody in his camp thinking to ask how important it is they stop paving streets based on political connections. They ran down a list people with a feeling thermometer. In addition to the candidates and local politicians, they asked about Franco Harris and Mean Joe Green. I took this to be an attempt to get a baseline to use to check against the candidate Harris rating. The list of 'issues' I was asked to rate sounded like was developed by somebody reading Peduto's press releases (WiFi, green this, etc). And in the list of qualities you think a candidate should have, they asked about "having left Pittsburgh to go to college, but returned" and "Being married and having a family."

Tuesday, July 28, 2009 8:35:00 PM  
Anonymous MH said...

"push poll", not "push pull". I was thinking about doors.

Tuesday, July 28, 2009 8:53:00 PM  
Blogger C. Briem said...

curious. But if they really asked "how important it is they stop paving streets based on political connections" then I think it would come close to being defined as a push poll in a sense. My survey colleagues would never ask such a question even in political surveys... it's too loaded a way to phrase that.

as to who... assuming it was a push poll in some sense. As you describe it it does not sound like what LR campaign would spend money on. Both Acklin and Harris I think will or already have access to some amount of funds. For things like this I always suspect it's some indirect support. If I recall correctly, DeSantis said he wasn't polling. I accept what he said, but when he said it there was someone out there polling for him.

Since I didnt think the Harris campaign was organized to the point of this, by elimination that would leave Acklin although I find it hard to see why they really would go out of there way to mention Franco... I am curious how Mean Joe Greene was brought up? Could be they were looking for a baseline measure of the sheer Steeler impact.. but still you rarely want to help out your opponent too much. But who knows.

Tuesday, July 28, 2009 9:50:00 PM  
Anonymous MH said...

The street paving issue was probably phrased positively (i.e. "repairing streets based on need and not political connections"), but they did use 'political' in the wording. And, while the question was loaded, it doesn't seem out-of-line given that the issue has been in the news.

I have some experience (years ago) with polling on political issues, and it was clear the interviewer was well-trained. No awkwardness with names, handled "Don't Know" responses properly when it wasn't listed as an option, etc. And the survey used the common questions/prompts to get at party ID, demographics, likelihood of voting, etc.

I'm guessing the point of the survey was to test which messages resonate with which groups of voters for a non-incumbent candidate. In which case, using me as a test-audience probably isn't the best idea, but randomness happens.

I tried poking around the candidates' websites for hints as to who took the poll, but nothing on Acklin's page struck a chord and Harris's page doesn't have any information anymore. And it still could be Luke, trying to see where he is vulnerable.

Tuesday, July 28, 2009 10:24:00 PM  
Blogger C. Briem said...

push poll does not imply unprofessional. Nuanced it sounds like. But if it was a highly professional poll, then by all accounts I hear the Acklin campaign is highly professional in it's rank and file operation fwiw.

but who knows. As I said, for polling in particular I have suspected off-book stuff going on.

but what did they ask about Mean Joe.... just some general recognition questions? Everyone knows mean joe? Maybe they are signaling Mean Joe is going to jump in the fray to balance the Franco factor.

I better not jest lest someone take that as more than random musing. Less than a WAG even, I have no idea whether Mean Joe cares a whit about any local election. But does Lynn Swann make an appearance? I still have been meaning to look at the Swann election results when he ran for governor to see if there was any local steeler imapct.

Tuesday, July 28, 2009 10:38:00 PM  
Anonymous MH said...

I'm still kicking myself for not asking the company conducting the poll, because they are almost always upfront about that.

Green was just one of a list of people asked about in a feeling thermometer type question (on a zero to ten scale where ten is...). The other were the candidates, Harris Sr., Rendell, Onorato, Dowd, Robinson, Shields, Peduto, the city council, Obama, and Clinton (Hillary). I think Green is a control variable for somebody's model.

Tuesday, July 28, 2009 10:57:00 PM  

Post a Comment

<< Home