Making Sense of Redistricting

I'm working on a cold beer at the Westville Pub (Wedge IPA if you're interested) while waiting for new tires to get put on my car.  While I've been here, I've gotten a call from a smart reporter from the Citizen-Times who is (like I am) trying to make sense of the new Census numbers and what they mean for redistricting--particularly in North Carolina.  Not a bad way for a politics and beer geek to spend a rainy day.  I'll hopefully have something more intelligent to say soon, but in the meantime, here are a few random notes about, or inspired by redistricting and the new Census :

• North Carolina population increased by over 18% since 2000.  That's the largest increase of any southern state (unless you consider Texas the South.  And let's be honest--no state that considers beef to be bar-b-que can be the South).
• We're now the 10th biggest state in the country (up from the 11th).  Despite this gain, we don't pick up any congressional seats.  
• Most of the pick-ups were in the sunbelt.  In the traditional South, South Carolina and Georgia picked up seats.  In the questionable South, Texas and Florida picked up seats as well.
• Despite this increase in population, North Carolina is unlikely to make any move to increase the General Assembly's capacity to govern.  I'll have more on this soon.
• Although we're not picking up seats in NC, the General Assembly can still redraw district lines.  I don't expect this to have too much influence on the 11th Congressional District (I don't see how it could look substantially different), but it could on districts throughout the rest of the state.  It will probably have the greatest impact on state legislative seats.
• Mike McDonald at George Mason has a nice set of resources on redistricting here.
• As this Census suggests, the country's population keeps climbing, but despite this >200% increase in population since the early 1900s, we've held at 435 seats in Congress for over a century.  See this book for a case that we should increase the size of the House.

A Few Quick Links

• The battle for the 2012 Presidential election calendar has begun.  Check out the always interesting Frontloading blog for more.
• Nice article by Rob Christensen about the talk about independent redistricting commissions,  Turns out the Republicans supported it when the Democrats were in charge.  Now that the Republicans are taking over, guess who thinks they are a great idea?
• The Tuckreader reports on an NC Policy watch piece showing that the poorest counties are the ones investing the most money in the lottery. 

Just in time for The Holidays (2 years ago)

If you are interested in North Carolina politics and you don't have it already, I hope you'll check out our edited book on NC politics, which came out a couple of years ago.  If you can get past the stuff Gibbs and I wrote, then there are some really terrific chapters in this book covering all of the major parts of North Carolina politics.  The book really is a who's who of Political Scientists who cover the state so I hope you'll pick up a copy.

You might also want to check out Rob Christensen's excellent historical treatment of the same subject.  Rob's the preeminent journalist covering Tar Heel politics and after reading this book, you'll see why.

I think the two books complement each other quite nicely. 

Moving Towards Purple, Again

I wrote before on the decline of Democrats on County Commissions in NC here and here.  The story there was that all of the focus on the national level has missed the real story in North Carolina: that North Carolina is a divided state that has been moving towards purple for a while.  It therefore shouldn't be surprising when Obama wins the state, but the Republicans keep a number of important seats in Congress.  I have been wanting to simulate this analysis with the General Assembly for a while, but state legislative historical data are surprisingly hard to come by.  Fortunately, Jeff Stonecash at Syracuse was kind enough to send me his data from 1901-2008.  I cleaned it up in a few places, added the last few elections and here's the result:

The bumpy lines are the actual percent Democrat in the state Senate and House.  The smoothed lines with the long title (sorry--I meant to change it, but didn't have time) represent a fancy way to smooth the data so it's easier to see the trend.  This is interesting in isolation, but it's even more interesting when we compare it to other states around us.  Consider South Carolina:

How's that for a nosedive?  I think the Palmetto state left purple behind a while ago.  Looks pretty red to me.  Also, for historical comparison, look how strongly Democratic it was until the 1960s--we're talking 100% Democratic. 

Most of the South looks similar, but check out this weird outlier: West Virginia:

Somehow the Democrats have gained ground in WV.  I know WV isn't really the South, but I still find this surprising. My friend and colleague Roger suggested that it has to do with the lack of race being an issue in WV, or perhaps with the industrial (instead of agricultural) base of the state's economy.  Gibbs wondered about the influence of the Byrd machine.  I'm not sure the reason, but I sure find it interesting.

Looking Forward to 2012

Sure it's too early to be looking forward to 2012 and sure we have a lot of actual, you know, policy to be talking about, but the folks at PPP have recently completed two interesting polls in North Carolina that I'd like to speculate about. Of course this makes me part of the problem of highlighting horse-race coverage at the expense of policy, but what the hell?

First, PPP polled GOP voters in North Carolina about who they plan to vote for in 2012.  I graphed the results below:

My guess is that every Democrat who's reading this blog is licking their lips at the prospect of a Palin candidacy and, at least early on, Republican voters might be up to the challenge.  I don't know about you, but I'll be watching Sarah Palin's Alaska with a slightly different eye now.

PPP also polled general election voters (without a partisanship screen) and found tested a few match-ups:
- Obama vs. Romney (tied)
- Obama vs. Huckabee (Huackabee up, but barely within the margin of error)
- Obama vs. Palin (Obama up)
- Obama vs. Gingrich (tied)

Perhaps more interesting, PPP also reports their results in bivariate form so we can see whether how many Obama voters are planning to defect to the Republican party and how many McCain voters are planning to defect to the Democrats. If we see of Obama voters planning to defect, then that would be a terrifically bad sign for the incumbent President.  Fortunately for Obama, however, that is not what the results suggest.  There are just as many McCain voters are planning to vote for Obama as there are Obama voters who are planning to vote for McCain. Simply put: Obama is not losing any ground in the Tar Heel state (at least thus far).

Of course, a lot can (and will) happen in the next months and years.  And a good deal of it will likely be bad for the incumbent

Another Way to Look at the 2010 Election Outcomes

Most of the pundit attention about the last election has focused on the changes in the partisan distribution in Congress and the state legislatures.  This, of course, makes sense.  Party identification is the single-most important variable in American politics.  If you want me to predict how a legislator votes and you can only give me one piece of information about that person, it would be party ID.

Most important variable doesn't mean the only important one, however.  A large literature in both American and comparative politics shows that male a female legislators act differently.  Women introduce different kinds of legislation, vote differently, and even relate to their constituents differently. Further, when there are more women in office, women voters tend to feel more efficacious towards government.

With this in mind, I called the folks at the Rutgers Center for American Women and Politics to see if they had updated their data on the number of female legislators to reflect last month's election returns.  Although they hadn't gotten it all on their web site, they had collected it and were nice enough to share it with me and I graphed the results below.

As you can see, female representation dipped a bit in 2011, both in NC and nationally.  Although I don't know for sure, I assume this is because female legislators are more likely to be Democrats and this was, as everyone including my basset hounds knows, a bad year for Democrats.

Two parts of this trend worry me.  (1) We weren't exactly knocking the top out before.  We were hovering around 25% representation for a group that makes up half the population.  Now we're decreasing.  That's not good.  (2) As I discussed above, men and women  act differently when in office.  A dip in the number of women in office suggests that we might see a small dip in bills of interest to women.  (3) State legislatures are like the AA ball of national politics.  Some people stay in AA forever, but almost everyone in the majors played in AA at some point.  State legislatures help train people for national politics, but if we don't have as many women in the farm system, very soon we won't have as many in the majors.

County Commission Article in the N&O

Today's N&O includes an op-ed by Gibbs and me based on the county commission data discussed here last week

Voter ID Laws in North Carolina?

The Raleigh News and Observer reports that the Republicans are planning to introduce a bill to require photo ID before voting.  The Republicans are arguing that this will reduce voter fraud and increase faith in the system.  The Democrats will likely argue that this will reduce voter turnout--particularly among less educated and minority populations.  So--what does the empirical evidence suggest about this debate?

There is very little evidence of voter fraud.  As the N&O article points out, there were 18 documented cases of double-voting in NC in 2008 and I've seen similarly small numbers reported in Georgia.  The NY Times reports very small incidents of voter fraud from a multiple-state study (see also here).  I have not seen any reputable study that documents widespread voter fraud in the last 20 years (see, in particular, this book).  Unless this research is missing something, it doesn't appear that voter fraud is a widespread problem.

There is some (very limited and contested) evidence that voter ID laws suppress turnout. Vercellotti and Anderson document a small decrease in turnout when voter ID laws are put into place.  In perhaps the most sophisticated study to date, researchers from Cal-Tech (can you say Baysian multi-level analysis?  Me neither) investigate the range of potential restrictions* and come to a similar conclusion: voter ID laws result in a small decrease in voter turnout.  These findings are not uniform, however.  A team of researchers from the University of Delaware found no effect of voter ID laws on turnout and Columbia's Robert Erikson and his colleague replicate this small (~2 point) effect, but note that it is "statistically inconclusive" and that "the data that have been analyzed to date do not allow a conclusive test."  Sometimes science is frustrating. 

There is some (very limited and contested) evidence that voter ID laws suppress turnout among people with lower socio-economic status.  Minorities and the elderly (two groups who traditionally vote Democratic) are the least likely to have DMV issued photo IDs.  From this, you would expect that these groups would be negatively influenced by photo ID laws.  And some studies have documented a larger turnout decline among minority and less educated voters, but others remain more skeptical.  The Cal-Tech group does not find that minorities are any more disadvantaged than other groups, but they do find that lower income folks and the elderly are comparatively more disadvanaged by voter ID laws. Again--a mixed bag of findings.


There is little evidence that voter ID laws increase trust in the electoral system.  Harvard's Steven Ansolabehere (say that three times fast) finds that there is no relationship between the frequency of electoral fraud and whether someone votes.  He also finds that "voters living in states with stricter identification laws did not report higher levels of confidence or higher rates of voting than those living in states with relatively weak identification rules." In other words, if we pass this law in NC, it probably won't give people more faith in the system or increase voter turnout.

Unfortunately the science here doesn't tell us unequivocally that voter ID laws are a good or bad thing.  I think if anything, the evidence suggests that we should be skeptical of both sides.  When Republicans say that fraud is rampant or that voter ID laws will increase trust in the system, be skeptical.  Likewise, when Democrats say that voter ID laws will depress turnout in large and meaningful ways and that turnout will go down particularly strongly among low SES voters, be skeptical. 


*Here are the possible rules, according to Alvarez and colleagues:
- Voter must state his/her name
- Voter must sign his/her name in a poll book
- Voter must sign his/her name in a poll book and it must match a signature on file
- Voter is requested to present proof of ID or voter registration card
- Voter must present proof of ID or voter registration card
- Voter must present proof of ID and his/her signature must match the signature
on the ID provided
- Voter is requested to present photo ID
- Voter is required to present photo ID.

Feeding our Growing Appetite for Apps

This holiday season millions of shoppers downloaded apps to stay at the top of their game on Black Friday. 2010 saw a dramatic increase in the role of cell phones as consumers used them to find the latest deals, scan bar codes to compare prices and products, and make purchases.  Thrifty shoppers were standing in line with product in one store while trying to get a better deal online from the store’s competitor.  At most major chains, just walking in the door activated a text message with special discounts and stores instantly added rewards to your account for visiting the store even if you didn’t make a purchase.

Apps are not just for holiday shoppers, they are for citizen consumers as well.  The explosion of users and uses for cell phones will soon be coming to a government near you.  With apps from the federal government <> , busy holiday shoppers can verify product safety before purchasing, pass time waiting in line by monitoring the status of their federal job application or touring the Library of Congress, then check the air quality before heading outside.

In Arkansas , holiday hunters can use Game Check for real-time game tagging or pay their property taxes without having to go to town. Government Technology reports that after only five months, the suite of mobile apps has become one of the Arkansas website’s most popular features.  Sportswomen in California can use Department of Fish and Game apps to locate nearby fishing spots or the closest place to buy a fishing license, while car buyers can use iSmog  to be sure the car they are purchasing will meet state standards.

City and county governments are tailoring apps to visitor and resident experiences as well.  Some are creating their own apps and many are using apps by other developers.  The Salem, MA Office of Tourism and Cultural Affairs <>  created its own Haunted Happenings - 2010 travel and event guide for brave Halloween visitors.  Funded by the Rockefeller Foundation in an effort to encourage use of public transit,  City-Go-Round </> provides real-time arrivals to waiting riders, but only if the transit service makes the data available.  Want more of a two-way conversation?  Now you can give feedback as well as get information.  Ten cities across the country are piloting and participating in development of YouTown, a relationship and communication app designed specifically for local governments.

So how are we doing in North Carolina?  It looks like local governments are leading the way in the Tar Heel state.  On City-Go-Round the site recognizes the high level of participation of transit authorities in the Triangle area.  MyCharlotte is widely viewed as a top-notch services and information product, and the app will soon be adding an avatar  and personal help-desk to guide residents and visitors. The Center for Digital Government and Government Technology just unveiled the 2010 list of digital cities , and I found a North Carolina city in the top ten for three categories.  Charlotte was number four in the nation for cities of 250,000+ population.  Winston-Salem showed up at number five on the mid-sized cities, and High Point was number five for smaller cities.  Here’s hoping for a top ten North Carolina City in the 30-75,000 category for 2011.

*Despite what the signature line says at the bottom, this post was written by Anne Cortes, Graduate Assistant at the Public Policy Institute at WCU  

Televised Attack Advertising on the Rise

It seems like every election people complain that attack advertising is increasing.  I'm always skeptical of these claims. After all, in 1964 we saw about the nastiest ad imaginable. 

So what's the answer?  Fortunately, the folks at the Wesleyan Media Project can help us answer this question--at least from 2000-2010.  They combined their 2010 data with previous data from the Wisconsin Advertising Project and compare the frequency of televised attack ads to promote ads to contrast ads.  Contrast ads "mention both a favored candidate and his or her opponent" while attack ads "mention only the candidate being attacked."  Promote ads just mention the candidate sponsoring the ad.

They conclude that "the 2010 House and Senate advertising is the most negative in the past decade." Further, although the Republicans sponsor slightly more attack ads than the Democrats, attack ads are on the rise in both parties.  Below, I've graphed the Wesleyan/Wisconsin data and separated it out by party.  The message here is pretty clear.

Here are the data for the Democratic sponsored ads:

And the Republicans:

 

Brace yourself for more negativity in 2012.

Couple of quick links

Despite the fact that they quote me, both of these articles are worth checking out:

• Article by John Boyle in the AC-T about the most recent iteration of Charles Taylor's annual fundraiser:

• Good analysis of the rise of Heath Shuler by Quintin Ellison in the Smoky Mountain News:

New Senate Leadership in North Carolina

Now that the election is over (or, at least mostly over) it's a good time to examine how the political landscape is likely to shift.  A good place to begin is in the North Carolina Senate where we will soon have a Republican controlled legislature with Republican-controlled leadership.  So--how different will this leadership be? 

Thanks to the hard work of Phillip Ardoin at Appalachian State University, we can get a good sense of how liberal or conservative various members of the state legislature are.  Below, I've plotted the 2007 NC Senate by their "Nominate scores."  Phillip has borrowed this concept from work done at the Congressional level and, using a bunch of fancy statistics, estimates the ideology of every NC Senator based on their voting record on non-unanimous votes during that session.  He's got a terrific explanation here, but all you really need to know to understand this graph is that a lower Nominate score means a more liberal senator, a score of 0 indicates a moderate voting pattern and a higher scores means that senator has generally voted in a more conservative direction.  The higher peaks in this graph mean that there are more legislators in that particular range and you can see that in 2007*, we had far more liberal than conservative legislators.  If you're curious, the mean score was -.27.

That is interesting (well, at least to me) by itself, but perhaps even more interesting is the relative position of the two names on the chart.  As most readers know, Marc Basnight is the outgoing President Pro-Tempore of the Senate and Phil Berger is the incoming leader.  According to their Nominate scores, Basnight was the 12th most liberal Senator in 2007 with a score of -.785.  Berger, on the other hand was the 4th most conservative with a score of .688. 

The bottom line here is that we will not not seeing a more moderate leadership in the next legislative session and if anything, the leadership will be slightly more ideologically extreme than it was before.
----------

*Obviously some things might have changed since 2007, but compiling these votes and computing the Nominate scores is a tremendous amount of work, so Ardoin understandably doesn't have newer scores up yet.   When he does, however, we'll be sure to link to them here.