November 26th, 2015, 3:30pm by Sam Wang
I know it’s a time for people to come together, eat, and get into conversation. Sometimes things can get contentious. Here’s a primer on What To Say To Your Frequentist Uncle. Lots of good stuff there, including a question on what “never” means.
Happy Thanksgiving, everyone!
November 2nd, 2015, 2:58pm by Sam Wang
Today, I filed a friend-of-the-court brief in the Harris v. Arizona Independent Redistricting Commission case (S. Ct. 14-232). The brief can be found here (for a summary of other briefs, see the Arizona Eagletarian blog). In it, I argue that the Supreme Court should reject Harris’s case on the grounds that there was no partisan injury.
Claims of partisan gerrymandering are sometimes wielded fairly loosely. Harris et al. claim that voters of one party (in their case, Republicans) were packed into districts by the Commission to impair their representation. My brief describes how such a claim can be put to an impartial mathematical test. In fact, there was no partisan asymmetry to what the Commission did. Ironically, my analysis finds a slight (but statistically nonsignificant) partisan tilt favoring Republicans, who are the complainants in this case.
The test I proposed is a simple one: the difference between the average support for a party in a state, and the median district-by-district outcome. The average-median difference* has well-known statistical properties, and can be applied to any statewide districting scheme. It might be useful in the future as a general standard for statewide partisan gerrymandering. A need for such a standard for partisan asymmetry has been expressed in the Vieth v. Jubelirer and LULAC v. Perry cases.
I have written more generally on rigorous measures of partisan symmetry. A current version of my article, “A Three-Prong Standard for Partisan Gerrymandering,” is available at SSRN.
* The average-median difference has also been proposed by Michael McDonald (SUNY Binghamton) and collaborators. My own contribution is to point out that this measure, which was formulated around 1897 by the pioneering statistician Karl Pearson, has well-behaved statistical properties which can be used easily by a court.
Tags: House · Politics
October 8th, 2015, 11:25pm by Sam Wang
In Davis v. Bandemer and Vieth v. Jubelirer, the Supreme Court has held that partisan gerrymandering is justiciable (i.e. within their scope to regulate), but that a manageable standard does not, in their view, yet exist. Here is a draft of my paper on how to define such a standard. Prong #1 is the subject of my Gerrymandering Theorem post below (and my specific mathematical question is described here). Prongs #2 and #3 have well-defined statistical properties, and are in that sense complete.
Tags: House · Politics
October 7th, 2015, 7:05pm by Sam Wang
To me, today’s news that Gallup is sitting out the primaries, and maybe even the general election, is not all that notable. The primaries are a hard-to-poll question; any race with more than two candidates seems to have issues (see the UK and Israel as examples). And the general election? It’s such a well-populated space, so there’s not that much publicity value in polling it. Besides, Gallup says they’re tweaking their methods, which is a good thing for them to be doing after their performance in 2012.
Also, I am considerably more preoccupied with gerrymandering and redistricting. I am developing statistical approaches for detecting a partisan gerrymander that can be used as a standard to be used by federal courts. This builds on work I published in the New York Times and here at PEC. I have done new calculations demonstrating that in the House, the likely total nationwide effect of gerrymandering is larger than the effect of other factors, including the clustering of Democrats in populated areas. So if a standard (mine, or anyone’s) is adopted by courts, the resulting reform can have a rather large effect on fairness of representation.
For law aficionados, the goal is to come up with a manageable standard for partisan gerrymandering, which is a justiciable question due to Davis v. Bandemer in 1986, followed by Vieth v. Jubelirer in 2004. I will be writing about the math of this question in the weeks and months ahead.
In addition to the legal battle, as an offshoot this project leads to a need for what I’ll call “gerrymandering theorems”: ways to derive the relationship between voting and seats from a small number of assumptions. For that, I’m looking for someone who is good with probability distributions, the Central Limit Theorem, and stuff like that. That person would ideally be not far from Princeton. Any takers?
Update: I have now spelled out the problem in the comment section. If I don’t find a proof, I will just end up graphing the outputs of the simulations I did previously. These days people don’t care so much about rigor. It just feels like it should have a closed-form solution, and I hate the thought of somebody pointing that out in the future!
Tags: 2016 Election · House · Politics
September 30th, 2015, 10:42pm by Sam Wang
The U.S. Congress has reached the point where not shutting down the federal government at the start of a new budget year is considered an accomplishment. Wow.
The other day, Josh Marshall at Talking Points Memo asked his readers for their favorite government shutdown. My favorite shutdown is the original acts of partisan brinksmanship in 1995 and 1996. Why settle for cheap copies when you can have the original?
The ascent of Newt Gingrich was a turning point for the modern Republican Party. The slide toward the right on issue positions began in 1974. This accelerated, and the tone took a sharp turn too, when Newt and Company came in. I was working in the U.S. Senate as a committee staff fellow. People were still adapting to the shock (or rush, depending on one’s party) of the Contract With America, which seemed to be filled with unrealistic goals. As a scientist naive to the appropriations process, I attended a briefing by the GOP budget director. I asked why one would link the debt ceiling with day-to-day funding of the government. To me, it seemed like asking for trouble. She replied that it was “a strategic move.” What an understatement!
All these steps were mind-blowing and extreme at the time. It’s all with us today. Back then it was so fresh, so new.
It seems that the departure of John Boehner as Speaker dramatically decreases the likelihood of a shutdown this year (though we’ll see what happens in December, when the temporary spending bill runs out). Furthermore, if House Republicans have any sense of self-preservation, they’ll refrain from doing so in the run-up to the 2016 election. That’s a big if. Still, one wonders if the next shutdown will have to wait until the first term of a hypothetical President Hillary Clinton (there, I said it).
Tags: 2016 Election
September 25th, 2015, 5:23pm by Sam Wang
From Math with Bad Drawings comes this gem, depicting what probability means to a political journalist. “The threat is real.” Heh.
There’s a whole series of such drawings. Check them out!
Tags: 2016 Election
September 21st, 2015, 1:59pm by Sam Wang
In time for the autumn equinox on Wednesday, the Summer of Trump is waning. Surveys are now out with samples entirely postdating GOP debate #2. Taking a median of the last 4 pre-debate polls gives Trump 33.0±0.7% (estimated SEM). Then, in the last 4 pre-debate polls, Trump is now at 26.5±3.0%. The combined uncertainty (sigma) is 3.1%. He has dropped by 6.5%, which is greater than two times sigma and therefore statistically notable. Where does this end? Has Trump peaked and will he continue to head down? It’s hard to say .
Note that the HuffPollster graph above is shown just to give a visual impression, not for extracting the numbers I give above. As useful as HuffPollster is, their smoothing algorithms tend to boost the size of swings in the last few data points. “Less smoothing” gives the closest result to the simpler calculation of medians, which is why I show it.
The fraction of GOP voters willing to go with outsiders who have never held office (Trump, Carson, Fiorina) still adds up to over 50 percent, but they have redistributed a bit. Considering Fiorina’s problems with accuracy and her dismal performance as Hewlett-Packard’s CEO, it seems likely that her rise will also be transient. If 2012 is a guide, she will last one or two months then fade, just as Cain, Gingrich, Santorum, and Bachmann did. And now, maybe Donald Trump – though his unique combination of issue stands (maintained domestic programs and extreme nativist attitude toward immigrants) and his persona could keep him from dropping all the way. It seems likely that many Trump/Carson/Fiorina voters will eventually turn to establishment candidates…but who?
Walker’s fallen below 2% [P.S. and has dropped out] and Jeb! is stuck around 8%, for now anyway. Assuming those trends persist, the highest-finishing serious candidate is Marco Rubio. As I have said before (link to The New Republic), Rubio is a relatively likely consensus candidate. In the past, the GOP has usually settled on a strong general-election candidate. Rubio polls relatively well against likely Democratic nominee Hillary Clinton. If the past is a guide, then at some point party donors/actors and primary voters may fall into line.
Tags: 2016 Election
September 16th, 2015, 6:34pm by Sam Wang
Step 1: Turn on CNN.
Step 2: Refer to Matt Taibbi’s Debate Drinking Game.
Step 3: Comment!
Tags: 2016 Election
September 10th, 2015, 9:07am by Sam Wang
This is interesting, from CNN/ORC: those who list Trump as their first choice added to those who list him as their second choice now add up to 50%. In principle, this opens the possibility of a clean win if the whole narrowing-down process were to play out immediately, using today’s opinion among GOP voters.
However, an actual Trump win by next year’s convention is still highly unlikely because of opposition by Republican Party leaders – and primary voters too. Of GOP voters surveyed, 21% said they would be “upset” if he were the nominee, not far from the number who would be upset about Jeb Bush.
Who was least upsetting? Ben Carson at 5%, followed by Marco Rubio at 7%. If the primary process were allowed to play out in an orderly manner (a big if, with Trump acting as a wrecker, to use the old Soviet term), these two would seem to be in a better position to make it to the final stages of the primary season.
Tags: 2016 Election