Princeton Election Consortium

A first draft of electoral history. Since 2004

An open-source thank you

November 13th, 2012, 12:00pm by Sam Wang


Benchmarking of Election-Eve snapshots continues. Drew Linzer wrote me: “basically we all succeeded in different ways,” as seen at Applied Rationality and Margin Of Error. His update is here, and a recent interview with the two of us on KUSP-FM is here. A separate topic is genuine long-term predictions, which I will evaluate later (“A Presidential/downticket prediction challenge,” October 28).

Today I want to focus on what has been distinctive about the Princeton Election Consortium – and the role you played as readers.

Here are two indicators for how we did.

  1. This year PEC got over 4 million page views. That’s three times our traffic in 2008, and occurred despite several brief service outages in September and October, including one just before the election. It is consistent with an estimate (see survey) in which we had three new readers in 2012 for every reader who started following us in 2004-2010.
  2. Same as 2004-2010, I used simple statistical methods to identify knife-edge contests where readers could maximize their leverage. Because President Obama was ahead all season, this meant downticket races. The ActBlue thermometer at left shows one metric. This exceeded expectations by over 6-fold. The Crossroads GPS link does not indicate how much PEC readers gave Karl Rove. I imagine it was a small fraction of that organization’s $300 million in expenditures.

We did well on questions that affect next year’s legislation and messaging: Presidential electoral-vote and popular-vote outcome, specific Senate outcomes, and House seat and popular-vote counts. In all areas, the predictions were very close.

The ActBlue list did identify genuinely close Senate races, with the exception of Nevada (Berkley v. Heller, decided by a margin of 1.2%). Specific House races were harder to identify because of the absence of polling data – DCCC/RNCC was generally a better bet. Correctly identified knife-edge races should have been won 50-50 by each side. Although this was true, further refinement of this process is needed.

But these outcomes are just part of the picture. The real story is the mechanism of this project successes. Three principles were at work:

  • Elementary statistical methods were all we needed to reveal truths about the 2012 race. Medians and other outlier-rejecting tools were enough. Adjustments like house biases and chewing over polling internals were not necessary. Good statistics can cut through the noise – if you have the nerve to ignore the chatter. I deeply believe this to be true for many problems in public policy and science – not just elections.
  • Our data came entirely from public sources. Pollster.com was a rich source of unbiased information.
  • We used open-source methods. All of PEC’s code is posted. Every bit of the activity was open to you, the reader. This led to wonderful benefits. Two examples are the depth and analysis added by Rick in Miami and Froggy.

Using these simple principles, we matched benchmarks such as FiveThirtyEight, Votamatic, and Pollster.com – and gave the sharpest moving snapshot of the race that you could find anywhere on the Web.

My own writing was also open-source. I wrote about what you discussed in the comment threads. Readers were a rich source of discussion and ideas. The project succeeded in large part because of your thoughtful feedback. There are too many of you to mention. This year we hit a sweet spot by having lots of readers – but not too many. Andrew and I thank you.

One more thing. Our Congress has become increasingly paralyzed in the last 20 years by Senate supermajority rules. I encourage you, especially if you donated or are a constituent of a swayable Senator, to make contact. Encourage him/her to vote for a rule change. It’s important for our democracy, no matter who controls the chamber.

In coming weeks posting will continue, though less frequently. To follow us, use the RSS feed in the top right corner, which will ping you when there’s something new. Also, please suggest topics – on any topic, whether political, economic, scientific, medical, whatever. I am interested in where this platform can go.

I end by quoting one of our most frequent commenters, Wheeler’s Cat.

Rasmussen cheating, Gallup antique methodology, non-gaussian structure of reality, asymmetrical political behavior, all fell to Dr. Wang’s pure mathematical statistics.

you were right.

but it sure was fun talking about all that stuff.

it was an epic tale of meme-warriors and Poll Jedi battling the sinister forces of disinformation and spin and “gut-feelings”.

i felt a lot of esprit de corps with my fellow commenters– we were like a rapid deployment force of truth commandos slapping down republican eumemes and pundit spin every day.

Tags: 2012 Election

124 Comments so far ↓

  • E L

    Contact your Senator about filibuster reform. DO. IT. Now.

    • Craigo

      Every Democratic Senator plus Sanders and King voted in favor of the Udall reform package, with the exception of Kerry, Reed, Reid, Inouye, Pryor, Feinstein, and Baucus.

      Now Reid has obviously changed his mind. Assuming all previous Yeas stay firm, and the new Dems and King stay true, that puts us at 49 votes, one short of the 50 needed to bring in Biden’s tiebreaker. That’s the target list.

    • 538 Refugee

      Craigo. Does Biden favor it? Does Biden even get to use his discretion or would he vote on orders from Obama?

    • Craigo

      Constitutionally, the VP’s authority as President of the Senate is completely independent of the President himself.

      (For a sad illustration of that, read up on the Federal Elections Bill of 1890, or the so-called Lodge Force Bill. It would have federalized American elections for Congress and President and protected the voting rights of African-Americans, but died of a filibuster that VP Levi Morton refused to break, despite President Harrison’s enthusiastic support.)

      Biden has repeatedly denounced the casual filibuster, and voiced support for reform in 2010. Moreover, he may still want to run in 2016, and pissing off the entire Senate Democratic caucus would start him off on the wrong foot.

    • Craigo

      I should note that Kerry, Inouye, and Feinstein did not vote on the 2011 resolution at all, so they might be a bit more ripe.

    • xian

      I despair of Feinstein ever supporting a change that reduces the privileges of an elite she belongs to.

  • Jack Barry

    I notice that Florida is still white in your map above. Any reason for this?

    I’m so encouraged by your methods and openness – it is so counter to the fog that the Republicans seem to have been lost in.

  • Craigo

    Given how many votes are outstanding in AZ-SEN, can anyone calculate the odds that Carmona wins? They’re not good, but an estimate would be nice.

    • Olav Grinde

      I would be grateful if anyone could cast light on why there are so many outstanding votes to be counted in Arizona. Over 600,000!? That just seems totally disproportionate.

      How does 600,000+ compare to previous elections?
      Was there a massive purging of the voter rolls in Arizona?
      Was there a large number of provisional votes?
      If so, what types of voters had to vote provisionally?

    • MAT

      As far as I can tell, provisional ballots were used in Arizona for 2 main reasons:

      1) Someone showed up to vote without proper ID. A provisional ballot would be cast, and then the voter would have to follow up and present ID at a designated location by a designated date (which I think is tomorrow). This process has been a gigantic mess from everything I’ve read.

      2) If voter requested an absentee ballot, but claimed not to receive it, they would have to vote provisionally on election day. This apparently happened a lot also.

      From the numbers, one of these problems happened to about 1 out of every 6 voters in AZ.

      The way we do it in NC makes a lot more sense to me – even if you have requested an absentee ballot, if the Board of Elections hasn’t received it by the time you show up to vote in person, then you are allowed to vote. Should the absentee ballot show up later, then it would be dealt with appropriately to avoid a double vote.

    • Olav Grinde

      MAT, thanks for casting light on this.
      Question: Are the procedures you describe in any way different this year, or is this just business-as-usual in Arizona?

      1 in 6 sounds awfully high!

    • Paul K2

      I have been looking at the Arizona vote, and apparently a lot of people who registered to vote, were not on the voter rolls. Apparently, these voters cast a provisional ballot, but then must return after the registration has been checked, to show ID.

      This seems like a convoluted process. Why not get the ID checked when the provisional was filled out?

      As of last night, about 5.4% of ballots hadn’t been counted, with another 5.5% of provisional ballots still not included in the vote total. If all the outstanding + provisionals are counted, then the turnout will hit about 77% of registered voters.

  • mediaglyphic

    dr. wang, it looks like the PEC house forecasts were not as accurate as other forecasts. Since PEC was saying 210+/1 10 for democratic house. Was there in fact a statistical difference in these estimates (not sure what the margin of error was). Do we know what the sources of error might have been in this forecast?

    • Cherith Cutestory

      Dr. Wang initially estimated that the built-in advantage for the GOP through redistricting (gerrymandering) was roughly +2.5%, but this turned out to be low–it was more like +4-5%. I also believe that the Dems did actually end up with 200 seats, so still within the moe that Dr. Wang initially forecasted.

  • 538 Refugee

    I have never used RSS before but it wasn’t hard to set it up in Thunderbird but you do need to know the feed url. To save others a little groping it is : http://election.princeton.edu/feed/

    Sam, how do you want to receive suggestions?

  • jayackroyd

    As I said to Sam when we spoke last Thursday, this is one of the most impressive comment communities I’ve seen in many years living in the world of the commentariat.

    • Craigo

      I think I learned as much statistical theory from the people here as I did in school. (Which granted, wasn’t much – even in poli-sci.)

  • securecare

    “…non-gaussian structure of reality…”

    Benoit Mandelbrot had some things to say about that in “(Mis)behavior of Markets”. Worth reading for statistics folks.

    http://www.amazon.com/s/ref=nb_sb_ss_i_2_10?url=search-alias%3Dstripbooks&field-keywords=mandelbrot+misbehavior+of+markets&sprefix=Mandelbrot%2Cstripbooks%2C357

  • &y

    Suggested area for further application of the best tool for clearly showing which of two things is bigger: the climate. Talk to the good people at RealClimate. See if there’s any way you could help. Thanks for PEC. See you in the RSSphere.

  • don in fl

    pryor,baucus.isnt it a stretch to call these two dems? mayb e with the president winning a second term they may show some backbone.

  • Dave Goldberg

    Well done on “More or Less” this week. It was particularly cathartic to hear your take on the role of pollsters after the insufferable blowhard from Gallup got finished saying that despite the 8 point (!) error in their national polling (or _only_ 4 points, if you grant them the cheat afforded by Hurricane Sandy) that aggregators would be nothing without the actual pollsters.

    But with pollsters like Gallup, who needs pollsters?

  • Rick in Miami

    Thanks for the mention, Sam. The greatest strengths of PEC are the relative simplicity of the approach (state polls only) and the open source code.

    The philosophical approach is opposite from the big black box approach of, say, 538 – so when both are predicting nearly the same thing, it’s a STRONG statement. There will always be value in poll aggregators that keep things as simple as possible – PEC reigns supreme (and alone?) in this department.

    Because the code is open, that means that we can always tweak and create alternate versions. For example, I find myself wondering “what if house effects were removed”, “what if we built in Lindzen’s idea that each state margin time series is the sum of a national and a local time series”, etc. Given free time, I can explore those questions!

  • Hayford Peirce

    I made a comment (complaint) several days ago on an earlier topic in which I said that in spite of all my efforts on three different browsers I could NOT get ActBlue to process my donations for two different candidates in states other than my own.

    This comment was “waiting for moderation” for several days and I have given up trying to find it to see if it ever got published and, if so, there were any replies to it.

    I would *really* like to know why ActBlue did not work for me. I gave money to a number of other candidates but was frustrated in my attempts with these two.

    Also — why are MY comments subjected to several day’s wait for the moderators rather than being processed practically instantaneously like the others?

    Because I’m Harvard ’63 and not a Princetonian?

    • Hayford Peirce

      Hmm, this one *did* go through instantaneously!

      Maybe because they thought I was a Yalie?

    • 538 Refugee

      I haven’t yet figured out why some of my posts end up in moderation and others don’t. I’ve suggested that they change the comments section to a forum so that threads/thoughts are easier to follow. As the community grew this method has been out grown.

    • Some Body

      There is moderation on some posts and not others. I guess Sam didn’t put your post through thinking here’s not the right place to ask questions about ActBlue (which is, well, true).

    • Michael

      I think your first comment got shunted to moderation because it was your first comment. The fact that this one went right through suggests that your first one eventually passed. At least, that’s been my experience. If I change my name or email, the comment goes to moderation again.

    • Michael

      Forgot to add, I have no idea why you had trouble with ActBlue. I do recall having had some sporadic issues with their interface, but never hunted down root causes.

    • Craigo

      Sam was also getting annoyed with people who insisted on freaking out over every single poll, so if you mentioned specific firms in your post it sometimes got sent to moderation limbo.

    • mediaglyphic

      “heavens a yale man.”

      Thurston Howell the 4th

    • Hayford Peirce

      Thanks for the comments, people!

      To answer a couple of them — no, my comment about ActBlue was actually my second comment. The first one *did* go to Moderation, but then showed up a couple of hours later. That was a day or two before I made my ActBlue comment.

      And no, I was not mentioning any particular polling firm in my other comment — and I’m certainly not like those hysterical people at 538 who would, as you say, freak out about *any* poll that didn’t indicate a clear Obama or Romney victory, depending on their taste….

    • Matt McIrvin

      I got lots of my posts sent to moderation over the course of the year. I never did figure out a pattern to it, but didn’t sweat it.

  • MAT

    We should also thank Brian Cohen for the wonderful Google Chrome PEC extension, which he updated repeatedly during the outages to enable those of us addicted to PEC to stay on our steady drip of updates.

    https://chrome.google.com/webstore/detail/princeton-election-consor/mopbhdfigdcejolaebakppgnffbhlamd/reviews

    • Brian C.

      I appreciate it, MAT. It was fun! A couple things I’m thinking for the next election cycle(s):

      – A Firefox Add-on
      – Side by side data from PEC, 538, Votamatic, RCP, Pollster, and Electoral-Vote.

      What do you think?

    • MAT

      Funny you should mention that, because I was considering doing an iOS app that did much the same thing. So you would save me the trouble ;-)

  • Olav Grinde

    Dr Wang, I am immensely grateful to you for this excellent site. You have consistently provided a clear, fact-based analysis that has cut through the noise. The posts and comments here have been even better than in 2004 and 2008 — and that’s saying a lot!

    Thank you!

  • BNW

    Preofessor Wang – Thank you for the ActBlue list. I contributed to those candidates and other close races. Congratulations on surpassing your fundraising goal six times over. Next time around, you could consider going for a longer list, perhaps 10 candidates, or more. I think the contribution this year shows that contributors are ready to help win those elections.

    And thank you for your analysis of the House the other day. The structural bias created by gerrymandering is a very significant problem for democratic governance in our country.

    And for the same reasons, yes, filibuster reform – now!

    • Hayford Peirce

      And I do indeed want to thank you for this wonderful site — the effort you have put into it is stupendous!

      And please, if you do continue with the ActBlue list, get some input from people like me who couldn’t make it work — and in my case it wasn’t from want of trying!

  • Some Body

    Thanks a lot from this one reader!

  • BNW

    Here’s a list of websites and contact information for incumbent Senators:

    http://www.senate.gov/general/contact_information/senators_cfm.cfm?OrderBy=party&Sort=ASC

    For new Democratic Senators:

    Heinrich (NM): http://heinrich.house.gov/contact/

    Murphy (CT) (see office numbers at bottom of page): http://chrismurphy.house.gov/

    Baldwin (WI): http://tammybaldwin.house.gov/contact/write-or-phone-my-offices.shtml

    Donnelly (IN) (see contact info at bottom of page): http://donnelly.house.gov/index.php

    – The rest are the campaign websites –

    Heitkamp (ND): http://heidifornorthdakota.com/contact/

    Kaine (VA): http://www.kaineforva.com/contact

    Wareen (MA): http://my.elizabethwarren.com/page/s/contact

  • RLJ

    Have you considered using neural nets to help improve the polls. Gallup has defended its methodology, but clearly they used a WAG to predict the likely voting population in 2012. There are a (limited) number of time series showing the trend of voting in elections, and demographic trends that your neural net could use to make a projection about the likely voting population in 2014 and 2016. It might be crude now, but it should improve with time.
    Great job this cycle and for the last three election cycles.

    • mediaglyphic

      RLJ,
      State polls are a very good source of information. Why complicate matters with neural nets?

  • Steve16748

    Yes everyone, lobby you senators against the filibuster. I have emailed my senators, Cantwell and Murray of Washington State. Below is what I wrote Patty Murray.

    Senator Murray, thank you for your service and congratulations on your great success as leader of the Senate re-election campaign. My wife and I have been long time supporters of you because we believe you sincerely represent ordinary people and small business. We donated to your last fight against Rossi.

    The Senate is not working, the filibuster must go. If we let the filibuster stymie Democratic efforts, we cannot shine a bright light on the obstruction of the Republicans in the House. We must be able to advance legislation or eventually we will start to lose support and begin to lose elections as a party.

    End the filibuster. In short Patty, as I say to my friends, I am getting old; I would like to live in a modern western democracy before I die!

  • Amitabh Lath

    Sam, thanks for putting up one of the most elegant analysis sites about elections.

    It was fairly clear from August on that Obama was going to win, but the margin surprised me (I had assumed 300-310 EV). You said as much multiple times, and urged us to look downticket. I donated to a few and they won!

    The huge victory by the quants in this election is heartening. I hope it inspires a humanities major or two to pick up a book on probability and discover the beautiful world of mathematics and statistics.

  • Rick in Miami

    Is this the same Rasmussen that ran the polls? If so, little wonder that they were so skewed:

    http://www.theblaze.com/contributions/the-coming-landslide-and-the-inevitable-fallout/

    • Michael

      No, different Rasmussen. The pollster is Scott, not Don. But man, reading this stuff never gets old:

      “In fact, most polls show Republicans enjoy a double digit enthusiasm lead going in to Tuesday. Yet polling models continue to show narrow wins for the president in key battleground and an overall electoral advantage. My own analysis suggests that Democrats are being overrepresented by an average of just under 3 points, enough to swing CO, IA, NH, NV, OH, WI and Maine’s 2nd CD while locking down VA, FL, NC and Nebraska’s 2nd CD. It also puts him within spitting distance in PA, MI, MN, and OR.

      That would give Romney a historic rout with 355 electoral votes. Compare that to the Real Clear Politics average showing him currently leading in states worth just 235 electoral votes. This disparity will become the stuff of teeth-gnashing, consternation and soul-searching among pollsters and the media that relies on them when the ballots are counted. Hopefully it will lead to better methodology and less reliance on polls to drive news cycles in the future, but in the short run it will just be fun to watch Nate Silver trying to explain how he could have gotten it so very, very wrong.”

      How’s that landslide thing working out for you, Donnie-boy?

    • 538 Refugee

      Maybe Sam could let him do a follow up article. He probably isn’t welcome back to that site judging by the comments section. I’m sure most of us would just love to hear what he has to say now. :D

    • Hayford Peirce

      Just read Don’s article. Even by Karl Rove and Dick Morris standards, this one is hard to believe….

  • mediaglyphic

    http://www.salon.com/2012/11/13/obamas_grand_bargain_ing_leverage/

    Robert Reich has a good opening salvo on fiscal negotiations.

    Does anyone think 55% top marginal tax rate is too high?

    • Michael

      That depends on where the top bracket starts and what the deduction/loophole structure looks like. With the rest of the current code left intact, my guess would be that most wealthy people would continue to have little trouble avoiding the top marginal rate no matter what it might be raised to.

      I’d be happy to just get rid of preferential treatment of capital gains. After all, it was good enough for Ronald Reagan.

    • mediaglyphic

      Michael, agree wit you on brackets and deductions. the thing that intrigues me, is that no-one is saying that revenues alone can deliver 4 trillion.

  • Dave Kliman

    I looked at your histogram on election eve and declared to all my friends that it would be Obama: 332…

    It’s cool how math works.

  • Bill Moore

    Outstanding job! I really enjoyed following PEC this election season. Open-source, transparent solid statistical analysis is the way to go! Thanks.

  • Dean

    Thanks again for this great site. I feel great that science won over propaganda and lies. Thanks also for reminding us to contact our senators to change the supermajority rules.

  • Eric Walker

    This reflects my carelessness, but I’m wondering if Dr Wang could re-post here the list of Senatorial candidates who appeared at one time or another on his ActBlue page, as I never kept track but donated to all.

    I ask because I now want to send anti-filibuster emails to all whom I financially supported, but cannot recall for sure who they all were.

  • 538 Refugee

    Republicans have a new GOTV strategy to replace the failed ORCA.

    http://natmonitor.com/2012/11/13/wife-mows-down-husband-with-car-for-allowing-obama-to-win/

  • wheelers cat

    In the post-election evaluation of poll houses and methodology, it seems like online polls were the most accurate. I wonder how polling methodology will change for next cycle?
    Given OFA’s elegant exploitation of behavioral science and SNT, I wonder if there are ways to quantize social media indicators for forecasting?
    And what about RAND? I havent seen them included in the post election analysis.
    Some commenters even here insisted RAND wasnt a poll at all…but is RAND the pollform of the future?
    Foe me…..the interesting thing about RAND is that it likely educated the respondents over time, and made them interested– made them pay attention.

    • MAT

      An interesting question – other than for mass media consumption, does polling even play a role in campaigns anymore, at least on a presidential level? After reading ‘The Victory Lab’ by Sasha Issenberg, it doesn’t appear that the Obama campaign relied on polling, but rather depended on prior voter history, massive voter contact and demographic information to build a huge predictive model for every voter in the nation. It sounds a lot like the PECOTA model Nate Silver built for baseball, which predicts performance of players based on similarity with other player’s characteristics. According to Issenberg, the model updates daily, giving a near real time accessment of what conditions on the ground look like. The fact that Silver was called in by Chicago in 2008 for some consulting ( according to his book) sorta reinforces my idea that they are running a massive voter version of PECOTA.

      Having this model spit out, ‘hey, we need a little help with demographic ‘x’, sounds like when you call in the behavioral scientist to figure out how to rope ’em in.

      So, back to my first statement, how long will people keep paying for polls if campaigns don’t use them, particularly if voter predictive models start to trickle down?

    • wheelers cat

      MAT why wouldnt all polling just go internal to the campaigns? e.g., why consume Rasmussen if hes wrong? Elite polling technology could become proprietary and LV models patented.
      Internal polling could become super-precise, and be used to fine-tune moneyball style GOTV.
      Another fascinating thing to me was how the campaigns used twitter and Facebook and social media.
      The Obama campaign actually bought trends like #47percent.
      And the “Dream Team” of behavioral scientists that volunteered for Obama’s campaign developed techniques that been labeled as proprietary — one of those people is Susan Fiske of Princeton.
      Remember Dr. Wangs article on undecideds from 2008?
      That is what I mean about the Team Red needing geeks worse than hispanics. 94% of scientists are not republican– they get no volunteer dream teams.

    • wheelers cat

      so maybe…..this is the wave.
      http://www.politico.com/blogs/burns-haberman/2012/11/obama-pollster-gallup-public-polling-needs-a-retooling-149807.html
      Nate said once that internal polls add 6points….but what if that isnt true?
      what if Obamas internal polls just did what Dr. Wang did, and distilled the true shape of the race from the noise?

    • MAT

      @Wheelers,

      If you replace ‘likely voter models’ with ‘PECOTA style voter modeling’, and ‘polling ‘ with ‘massive voter contact updating the model daily’, everything you said above makes perfect sense. I think polling as we know it is going to be relegated to public media, not campaigns, because better tools already exist. And there will be a lot of pressure from State parties for these new tools and techniques to be used for down ticket races, so in time they will trickle down.

      There is the very real possibly that Gallup could go the way of Twinkies.

  • Christian

    I have a suggestion to which topic you could devote some of your energy, apart from predicting election results: teach the public that the US are not and cannot go broke. The US are the creator of the $, so they can always pay everything that is denominated in $ and that is quite a lot. That does not mean that it is wise to do so. The real capacity of the economy to absorb the spending has to be taken into account, to avoid inflation. But with that in mind, one can start arguing about what spending is actually meaningful and what is not, when the issue of affordability is out of the way.

    Warren Mosler wrote an easy to read book, called the Seven Deadly Innocent Frauds, that will introduce you to how our monetary system actually works. The foreword is written by James K. Galbraith: “All in all, this book is an engaging and highly instructive
    read – highly recommended.”

    http://moslereconomics.com/wp-content/powerpoints/7DIF.pdf

    • wheelers cat

      sheesh, the basic problem with our economy right is fanatical faith in the “freed” market.
      Its the religious fervor that drives all this makers and takers crapology.
      The free market is teleologically incapable of improving the human condition — it only improves the condition of the overclass.
      That worked great for Americans in the past when America was the overclass of the world.
      Not so much anymore.
      The other problem with the freemarket is that its an ecophagy– in times of resource starvation (post peak oil for example) it eats itself.
      Thus the middle class.

  • Amitabh Lath

    Sam, your approach really needs to be published.

    Your method of calculating the median is the simplest and most elegant one out there.

    You hit the mark without much in the way of priors or economic inputs.

    I do not know if you did any checks against data selection criteria (increase/decrease the time window for relevant polls, change the function used for prob. dist. etc…). I suspect your method is pretty darn robust against small tweaks like that.

    You have developed a simple yet highly sensitive technique that takes accurate national/state level temperature. Look at the sharp jumps at Ryan VP,
    47% remark, debate#1,#3 etc.

    Surely the 1st derivative at those points says something about the speed of dissemination of the media narrative (or in some cases, the fight between dueling media narratives).

    This has utility beyond entertainment for us political junkies.

    • wheelers cat

      Amitabh, did you see my link above to Wired? It is an article about Opensourcing Nate Silver that even mentions Dr. Wang.
      Every media brand is going to want their own court statistician now after 538’s success.
      I think Dr. Wang would be a natural fit for Wired.

  • Olav Grinde

    Dr Wang, today Mother Jones has an interesting article on gerrymandering, and what it meant for the election.

    http://www.motherjones.com/politics/2012/11/republicans-gerrymandering-house-representatives-election-chart

  • Olav Grinde

    “Most Americans voted for Democratic representation in the House. The votes are still being counted, but as of now it looks as if Democrats have a slight edge in the popular vote for House seats, 49%-48.2% … the 233-195 seat majority the GOP will likely end up with represents the GOP’s “second-biggest House majority in 60 years and their third-biggest since the Great Depression.”

    “So how did Republicans keep their House majority despite more Americans voting for the other party…? Because they drew the lines.”

  • Olav Grinde

    There’s an interesting graph for seven states comparing the popular vote with the number of House seats: Florida, Wisconsin, Pennsylvania, Ohio, Virginia, North Carolina and Michigan.

    Needless to say, the discrepancies are striking.

    For balance, the article mentions Maryland and Illinois, two states where gerrymandering apparently favors the Democratic Party.

    Nevertheless, the most striking graph is the one at the end of the article where it is shown, relatively speaking, how many votes it took to elect a Republican or Democratic congressman in the respective states.

    • Amitabh Lath

      Olav, that is a stunning set of bar graphs. Republicans controlled the state houses in 2010, and this is what they did.

      Is there any sensible anti-gerrymandering bill that has a ghost of a chance?

    • 538 Refugee

      Andrew had a good comment that explains why it isn’t all gerrymandering. I also saw an article about how the civil rights act gives cover to the people drawing the lines to keep urban areas in one district to help ensure minority representation. Now when you draw the rest of the lines for the rural districts they will favor the conservatives.

    • mediaglyphic

      @538, if one party wins the popular vote and the other side gets most of the seats, thats the definition of gerrymandering. There are no two ways about it.

      @olav, thank you for posting this from Ma Jones. Your post is one of the reasons i love this board.

    • Michael

      “if one party wins the popular vote and the other side gets most of the seats, thats the definition of gerrymandering”

      Not necessarily. It is at least theoretically possible that, even with district boundaries drawn with zero partisan input, one party could win 55% of the seats by very narrow margins while the other party wins 45% of the seats by very large margins, thus giving the majority party the minority of the seats. Of course, in reality a lot of this really is gerrymandering, but I could certainly believe that it isn’t all gerrymandering.

    • Michael

      Oops. Should have read all the way to the bottom before responding.

  • 538 Refugee

    “Also, please suggest topics – on any topic, whether political, economic, scientific, medical, whatever. I am interested in where this platform can go.”

    It took me a while to digest the headline on 538 today. Nate has moved on to baseball. If you start doing sports I’ll probably become PEC Refugee somewhere. ;)

  • weichi

    @mediaglyphic

    Not true. I posted this at Kevin Drums site:

    Imagine a state with four counties, each with the same population. The state is 50/50 D/R overall. One county is 90% D, 10% R. The other two counties are 36.7% D, 63.3% R. You end up with 3 R seats and 1 D. No real gerrymandering required here, just use the existing county barriers (you might note the similarity here to states like Pennsylvania).

    That is *not* to say that gerrymandering isn’t going on. Of course it is! It just says that you can’t look solely at things like the difference between vote margins and seat margins to determine the extent to which gerrymandering is happening.

    • mediaglyphic

      @weichi,
      i understand the substance of what you are saying. Thank you for reiterating them.

      Gerrymandering, of course is very loosely defined as a drawing of districts which results in them looking like salamanders.

      Its my belief that when the composition of seats does not reflect the popular vote, there is an issue. It may be overt or covert, but its still an issue.

      You cannot disprove this by looking at the composition of county population. Its a case of Godels’ incompleteness theorem, part 1. All truths cannot be reached deductively.

    • Michael

      Pennsylvania is an excellent case-study. Obviously some Republican shenanigans involved in the last redistricting, but equally obvious that there are going to be a couple of CD’s in Philadelphia that are overwhelmingly Democratic unless you gerrymander in the other direction.

  • 538 Refugee

    Republicans are threating their first filibuster of the new term.

    John McCain, center, Kelly Ayotte, left, and Lindsey Graham vowed to block the possible nomination of Susan Rice to the State Department. Photograph: J Scott Applewhite/AP

    Is this playing right into the hands of Democrats seeking the rule change?

  • Craigo

    Latest innumeracy from pundits and crazies: They’re shocked – SHOCKED – to find that 99%+ black precincts sometimes vote 99%+ for the Democratic candidate.

  • Stephanie

    As far as topics that could be covered in the future, it would be very interesting to see analysis on various “fiscal cliff” (or slope, or whatever) solutions and other economic issues. Also Obama Care impacts, etc. When the pundits talk about this stuff they cover a very narrow portion of the subject. I think it would be very helpful to see impacts of the policies proposed in the immediate and the long term 10-years that are always talked about and how that impacts the deficit/debt. Also, I think it would be interesting to see in more detail what is actually going on with the deficit/debt in a more digestible way than what the government and media provide thus far. I think that would make this site relevant 24/7 until the next election. You could even have a topics bar at the top: politics, economic policy, medicine, etc.

    Also, I was wondering if there was a way to cut through the political posturing to cut down on the noise, in the same way that you did for the polls? I felt like I had a pretty good handle on the election due to the consistency your site provided and the analysis that pointed out noise. Maybe you could start doing similar analysis based on the national attitudes of these issues as well (as supposedly politicians take that into consideration). Even more interesting, would it be more accurate still to take it to a state by state level to figure out what individual politicians would do given the region they represent? I know that is all probably too much to ask, but I think it would help focus the geeks like myself on where the important tipping points are, etc, and if we live somewhere where we could make a real difference then we could act instead of wonder.

    All in all, I LOVE the site and hope you do many different things with it- I really think the sky is the limit for you.

  • pechmerle

    Thinking about continuing topics for this site:

    As Sam, and several of us, noted during this election season, there was an unprecedented amount of money spent. A large slug of it was “independent” spending by Super-PAC’s as permitted under Citizens United.

    The results don’t look very good this cycle for the Super-PAC’s. And sometimes, it appears, an ad message chosen by a Super-PAC hurt rather than helped the favored candidate.

    What I’m getting at, though, is whether the quantitative methods favored on this site can be applied to the effects of spending in our elections? There are quite a few sites that collect the data on candidate spending and independent spending, so the raw materials should be available. What I don’t know (no clue, actually) is whether there is quantitative meta-analysis that draw some useful conclusions out of that data.

  • bks

    Thinking ahead to 2016 it would be nice if all the aggregators could put time zones in the date stamps for the comments section.

    –bks

  • Michael

    I discovered your site about a month before the election and have enjoyed taking in your insights. I’m glad you will be posting your assessment of the longer term predictive value of various election forecasting models. As you have pointed out, this is much more challenging than an election-eve prediction. Your approach seems to presuppose that the stochastic process characterizing poll movements through the summer and fall won’t experience any significant structural shifts prior to the election. I am not convinced that you can rely on that assumption. Are you? Also, how many election cycles would you need to incorporate into your evaluation before you could make any fair model comparisons?

  • pechmerle

    Another thought about future topics for this site (before 2014 and 2016 roll around that is):

    Sam and others have worked on the amount of gerrymandering of House districts that there has been, and it’s un-leveling of the playing field.

    An important topic is: what is a fair, defensible way of drawing House districts? We all tend to think that allowing the foxes to guard the chicken house — umm, allowing state legislators to draw the districts — leads to considerable gerrymandering.

    But, as others have noted, the Voting Rights Act introduces some constraints of its own. Baker v. Carr, and Reynolds v. Sims — the famous one person, one vote cases — also constrain the line-drawers.

    But it goes beyond those. I have seen academic comment (the cite for which I have lost) that appointed commissions (such as California now has, with representatives of both major parties and non-affiliated members) don’t do any fairer job than legislatures do.

    So the question is, what is a good, fair method. And can PEC-type quantitative analysis usefully address and clarify it?

    Are there foreign models that have merit? How do districts get drawn in other large democracies (India, Australia, Britain, France, . . . )? Do they have problems with gerrymandering, or other demerits to their systems, that need to be taken into account?

    There is a large political science literature on redistricting. I don’t know how quantitatively sophisticated much of it has been though.

    • xian

      I think the California commission did a great job. I think the trend of nonpartisan district drawing and open primaries may moderate some of the old “extremist in a safe seat who only has to win primaries of party fanatics” problem.

    • Brian C.

      I apologize in advance, this is an incomplete thought. I believe the key here would be to first define the problem, and right there you get into murky waters.

      Take a simple example, Idaho (where I live): there are two congressional districts, both of which have a Republican congressman. Republicans control the statehouse with a 78% majority. There are a handful of Democratic strongholds scattered around the state. Now, on the one hand, Republicans might want to slice the Democratic areas up (drawing lines straight down the biggest ones) to minimize the chance it could turn a whole district blue. On the other hand, maybe they actually WANT that; if they feel the state is getting bluer it might be advantageous to have us all in one district to “limit the damage”.

      Maybe the commissions’ goals have nothing to do with partisan voting blocks. It’s possible the goal is easier administration, or closer alignment between statewide and federal districts.

      Regardless, if the goals aren’t defined and universally agreed upon, I think the “gerrymandering” accusations will always fly anytime districts are re-drawn.

    • Andrew

      Its worth asking what the purpose of dividing an electorate into districts is.

      In my view, we need small districts to allow politicians to represent a small enough subset of the population in a given political body that retail level politics is possible. For example, in Pennsylvania, the state house has districts of just 62,500 people, small enough that any citizen can walk a district and compete and gain an entry into politics. Its also small enough that most communities of interest can be set aside into their own seats, giving political representation to various minority communities. Thus in Philadelphia, there are districts for the Puerto Rican community, districts for Italian neighborhoods, districts for urban Republican areas, districts for middle class black neighborhoods, etc. allowing every stripe of the rainbow to have a voice. It would certainly be easy to just draw lines across the map that ignore these communities in the city (or in the state) and divide them up, but what purpose would that serve in a democracy that is supposed to be representative of all of the people?

      When you look at the congressional districts in Pennsylvania, a similar sense of order emerges from underneath. There is a seat for Pittsburgh (14), one for York County (4), one for Lancaster County (16), one for Bucks County (8), one for Montgomery County (13), one for Erie (3), one for the Allentown-Bethlehem (15), one for Altoona (9), one for Johnstown (12), one for Scranton/Wilkes Barre (11), one for the black half of Philadelphia (2), and one for the white and hispanic half (1), etc. The one really goofy seat, the 7th District, looks the way it does because the congressman for the Lancaster County district refused to move out of Chester County, which had been joined to Lancaster in the past when Pennsylvania had more seats. In order to not force him into a primary with another Republican, the legislature drew the squiggly lines of the 7th district. The original intention was to not have squiggly lines and to make the 6th and 7th districts more compact, with the 6th having northern Chester County and Berks County and the 7th the southern half and Delaware County

    • Andrew

      Its worth asking what the purpose of dividing an electorate into districts is.

      In my view, we need small districts to allow politicians to represent a small enough subset of the population in a given political body that retail level politics is possible. For example, in Pennsylvania, the state house has districts of just 62,500 people, small enough that any citizen can walk a district and compete and gain an entry into politics. Its also small enough that most communities of interest can be set aside into their own seats, giving political representation to various minority communities. Thus in Philadelphia, there are districts for the Puerto Rican community, districts for Italian neighborhoods, districts for urban Republican areas, districts for middle class black neighborhoods, etc. allowing every stripe of the rainbow to have a voice. It would certainly be easy to just draw lines across the map that ignore these communities in the city (or in the state) and divide them up, but what purpose would that serve in a democracy that is supposed to be representative of all of the people?

      When you look at the congressional districts in Pennsylvania, a similar sense of order emerges from underneath. There is a seat for Pittsburgh (14), one for York County (4), one for Lancaster County (16), one for Bucks County (8), one for Montgomery County (13), one for Erie (3), one for the Allentown-Bethlehem (15), one for Altoona (9), one for Johnstown (12), one for Scranton/Wilkes Barre (11), one for the black half of Philadelphia (2), and one for the white and hispanic half (1), etc. The one really goofy seat, the 7th District, looks the way it does because the congressman for the Lancaster County district refused to move out of Chester County, which had been joined to Lancaster in the past when Pennsylvania had more seats. In order to not force him into a primary with another Republican, the legislature drew the squiggly lines of the 7th district. The original intention was to not have squiggly lines and to make the 6th and 7th districts more compact, with the 6th having northern Chester County and Berks County and the 7th the southern half and Delaware County. A similar quirk explains the odd lines in Philadelphia. Bob Brady, the white leader of the Democratic Party in Philadelphia lives in a mostly black neighborhood in West Philadelphia. In order to not have Brady represent the black community, a thin squiggly line connects his home neighborhood with the rest of the 1st district, which is a series of mostly working clsss white and Hispanic neighborhoods in Philadelphia and Delaware County, and leaves a relatively compact 2nd district which encompasses most of the black and Jewish neighborhoods of the city and nearby suburbs. The lines of the 6th and 7th and 1st and 2nd districts could certainly have been drawn more compactly, but it wouldn’t have changed their political composition – instead it would just reduce the seniority of Pennsylvania’s congressional delegation by forcing out two incumbents. Since Pennsylvania lost one seat in redistricting, the legislature combined two Democratic held seats in western Pennsylvania (Johnstown and Beaver Falls), where the population has been static or falling. This seat became the new 12th district which the Democrats then proceeded to lose in the general election because of the ongoing swing of western Pennsylvania to the Republican party, not from any “partisan gerrymandering” intended to hurt Democrats.

      Another example is the 6th district. When first drawn in 2002, the district was a mix of heavily Democratic towns including Lower Merion, Norristown, Pottstown, Reading, Phoenixville, Coatesville, and Downingtown, along with some moderately Republicans exurbs. The district actually leaned Democratic, but was won by a popular moderate Republican state senator, Jim Gerlach. The Democrats then had multiple shots at unseating him, but repeatedly came up short despite the district voting for Senator Casey and President Obama. Then the legislature redrew the district, but still left it as a district which voted in a majority for Obama in 2008, so hardly unduly partisan. And the Democrats still lost in 2012. Is the seat “Gerrymandered” just because the Republican won again?

      I hope this helps people understand some of what is behind squiggly districts, and why the squiggly lines are often nothing more malicious than trying to maintain seniority of the delegation, and thus power in Washington to help the residents of the state.

  • Olav Grinde

    I thought my good friends here on this forum might enjoy this beautiful story. Not all politicians are the same…

    http://www.bbc.co.uk/news/magazine-20243493

  • Olav Grinde

    A topic that I would like to see receive a stronger focus, is he role of campaign money from corporations and big donors.

    There has been excellent focus on Karl Rove and the embarrassingly low batting average of his SuperPACs, for instance in Dr Wang’s recent post. But the role of big money is far broader than just funneling money to the Presidential, Senate and House elections.

    Here is a modest example: Monsanto launched a campaign to stop California from becoming the first state in the union to pass a proposition requiring labeling of GMO-containing food products.

    Initially, California’s Proposition 37 enjoyed support of 65% of those polled. Monsanto and its allies “invested” $46,000,00 to defeat the initiative. They succeeded.

    http://www.mercurynews.com/elections/ci_21949842/proposition-37-defeated-food-movement-vows-fight

    http://truth-out.org/opinion/item/12741-how-monsanto-flipped-the-vote

    • Amitabh Lath

      Hi wcat, Olav beat Andrew Sullivan to it! See the link to the same gerrymandering plots a day earlier, up above.

      It’s horrible that local statehouse elections that maybe had 20% participation by the electorate end up determining the eventual makeup of Congress.

    • wheelers cat

      Unfairly, a vrai dire, Sullivan gets more eyes than Olav. Hopefully that will change in future.
      I’m encouraged that the anti-democratic nature of redistricting is getting eyeballs.

  • Olav Grinde

    Senator Chris Coons is embracing the call to fix elections at the state level. He has a very interesting set of proposals, whereby states would be rewarded for coming up with plans that include:

    • Providing flexible registration opportunities, including same-­day registration;
    • Providing early voting, at a minimum of 9 of the 10 calendar days preceding an election;
    • Providing absentee voting, including no-­excuse absentee voting;
    • Providing formal training of election officials, including State and county administrators and volunteers;

    http://www.huffingtonpost.com/2012/11/15/chris-coons-election-reform_n_2134224.html

    • MAT

      I gotta say, from your list, NC has it’s act together. Other than same day registration, we do all of the above, and you can register while early voting up to the Sat before the election.

    • Olav Grinde

      Yes, that absolutely seems to be the case. I have been positively surprised to repeatedly read articles underscoring that, with regard to enabling voting, North Carolina is one of the most progressive states in the union.

      If I remember correctly, North Carolina’s governor had also implemented an efficient program for quickly re-enfranchising ex-felons.

      That is quite a contrast to states that dis-enfranchise felons for life — even after they have paid their debt to society.

  • wheelers cat

    More evidence of how disenfranchised conservatives are from contemporary culture…
    obamagifts
    http://andrewsullivan.thedailybeast.com/2012/11/obamagifts.html

Leave a Comment