John Allen Paulos has a useful piece in today’s Guardian on the meaning of the US election and the tendency people have to draw sweeping conclusions about the US electorate from the numbers:
Excuse my mathematician’s obsession with coin flips, but consider this. There is a large bloc of people who will vote for the Republican candidate no matter what, and a similarly reliable Democratic bloc of roughly the same size. There is also a smaller group of voters who either do not have fixed opinions or are otherwise open to changing their vote.
To an extent, these latter people’s votes (and thus elections themselves) are determined by chance (external events, campaign gaffes, etc).
So what conclusion would we draw about a coin that landed heads two or three times out of four flips (or about a sequence of two or three Democratic victories in the last four elections)? The answer, of course, is that we would draw no conclusions at all.
Nice to know that dsquared and CT are ahead of the curve. Not that I agree, really.
External events an campaign gaffes may appear like coin flips to you but behind the scenes there are thousands of people trying to figure out what gaffes to claim the other guy made and what external events to play up, etc.
The guy/gal who figured out the “Kerry flip-flops” line is a genius (ok an evil genius, but really just a guy/gal who is very good at their job). People understood immediately what it meant and that it was bad and it stuck like glue and people were muttering it on their way to vote like a mantra.
To claim that the World Trade Centre attack was an inside job might be viewed as far-fetched, although I know people (over here in Europe) who believe that it was and draw parrallels with the burning of the Reichstag. It is however irrefutably true that President Bush ignored several calls to tighten up airport security on domestic flights before the attacks. And even if he didn’t order or at least allow the attacks, he certainly chose to talk about them repeatedly at every opportunity.
The good news from the US election is that it was pretty close. With good news management and the selection of the right candidate, there is some hope for change next time round. To suppose however that its just luck and that analysing it is pointless is to heavily miss the point and to risk loosing even more heavily next time.
It’s an interesting thesis, but it doesn’t hold up when you look beyond the presidential numbers. There has been a slow shift towards the Republicans in the Senate and the House as well. Statistically, that makes quite a difference.
I don’t think this piece is useful at all. I think it’s pernicious, anti-democratic, and full of faulty analogies.
First, the author undermines the idea that a 51% electoral victory can be taken as a mandate for action. Introducing some sort of popular-vote super-majority as a mandate for legitimacy is fundamentally contradictory to the democratic process.
Second, viewing voter behavior as analogous to random variables undermines the value of voter choice. This disrespects both the democratic process and demotivates voters.
I didn’t vote for Bush, but I think it’s clear he won fair and square this time (in the limited sense that any political campaign is every fair and square). Time spent besmirching his legitimacy may be emotionally therapeutic, but otherwise, it does more harm than good.
The second half of the editorial, regarding the regression analysis, is much better. In this view, Bush underperformed, thus possibly indicating a fundamental change in the electoral towards the Democrats.
…although, I would add… I have a lot of time for Paulos and have read his book (and do have a maths degree, from Oxford). There are a lot of spurious statistics out there being used to pollute our minds with rubbish ideas. But I don’t take his views on the election. Insofar as apparently “random” events decide the US election, valuable lessons can be learnt from seeing what those events were and making sure they happen to you next time.
Also ignoring the effect of Perot (not that I’m entirely sure we can accurately figure out exactly what the effect was) tends to make the thesis unreliable.
I don’t know if I agree with Paulos’s thesis, although it’s very consoling. But The gains made by the Republicans in the House and Senate cannot be used as evidence to the contrary. In the Senate the problem is simple: little, almost uninhabited states like North Dakota get the same two Senators as New York or California. And most of these little empty states vote Republican. This is a structural problem that the Democrats have to solve, but it says nothing about the relative popularity of the parties. The Democratic minority in the Senate probably represent at least half the population. As to the House, weren’t the small Republican gains all the result of blatant, and very possibly illegal, gerrymandering in Texas?
The fact that more than 10% of the electorate voted for the Bushies clearly indicates that the Democratic party is totally bankrupt, dead. Am I the only one who can see it? What am I missing?
I don’t know if I agree with Paulos’s thesis, although it’s very consoling. But The gains made by the Republicans in the House and Senate cannot be used as evidence to the contrary. In the Senate the problem is simple: little, almost uninhabited states like North Dakota get the same two Senators as New York or California. And most of these little empty states vote Republican. This is a structural problem that the Democrats have to solve, but it says nothing about the relative popularity of the parties. The Democratic minority in the Senate probably represent at least half the population. As to the House, weren’t the small Republican gains all the result of blatant, and very possibly illegal, gerrymandering in Texas?
I don’t know if I agree with Paulos’s thesis, although it’s very consoling, but the gains made by the Republicans in the House and Senate cannot be used as evidence against it. In the Senate the problem is simple: little, almost uninhabited states like North Dakota get the same two Senators as New York or California. And most of these little empty states vote Republican. This is a structural problem that the Democrats have to solve, but it says nothing about the relative popularity of the parties. The Democratic minority in the Senate probably represent at least half the population. As to the House, weren’t the small Republican gains all the result of blatant, and very possibly illegal, gerrymandering in Texas?
I take the author’s caution against drawing conclusions from these close elections, but…
Since the ‘64 Civil Rights Act, white flight to the suburbs and the Dixiecrats leaving the Democratic party, Republican presidents have won 7 elections, while the Dems have only won three. And the Democratic presidents were both southern governers.
This small dynamic might be enough to turn these otherwise close elections. But there are surely other small factors which, when isolated, could be argued as decisive.
My picks, in descending order of impact:
1. Chance events (e.g. weather in Cleveland)
2. Ballot characteristics (e.g. ordering of the candidates on the ballot)
3. Karl Rove
4. Pre- and post-poll manipulation
5. Demographic and ideological factors
Paulos is writing at a time when Democrats are wondering whether we need to radical reform the party or compromise our principles to remain electable.
Paulos is asking evidential questions. Should we conclude that the Democratic party is in terrible trouble because we narrowly lost this election? Do these results prove that we need to do something dramatically different next time? Not necessarily.
People are talking about this election like it was an ignominious defeat. It wasn’t. We nearly beat a wartime incumbent in a non-depression economy. We must have done a lot of things right.
Obviously, mistakes were made. But we only lost by about 3 million votes. We’ve got to keep things in perspective. We don’t need a scorched earth policy to pick up another 3-4 million votes in 2008.
I want radical reform in the Democratic party, but not as a reaction to a relatively minor electoral setback. If Bush had won in a landslide, I’d be wondering about the future of the party. Losing by a 3% margin is hardly a resounding rejection the Democratic party and its values, despite what the mean spirited pundits would have us believe.
The election was not lost by 3 million votes.
The election was lost by less than 150,000 votes in Ohio.
Possibly much less- I will let you know if we ever get a count out of Katherine Blackwell.
If the 3 million number meant anything, we would have been saying President Gore for the last four years.
Interesting idea, but I think its premises…
There is a large bloc of people who will vote for the Republican candidate no matter what, and a similarly reliable Democratic bloc of roughly the same size. There is also a smaller group of voters who either do not have fixed opinions or are otherwise open to changing their vote.
…are becoming less true, if they were ever true. Both parties are coalitions of interest groups. The constant aim of politics (at least in the American winner-take-all system) is to broaden the coalition without creating or exacerbating internal inconsistencies, which give the other party opportunities to split off disaffected groups. Thus the system tends toward 50/50 stability.
Case in point: the GOP used to be the party of fiscal discipline. But partly in an effort to win over some traditional Democratic interest groups (like seniors and African-Americans), they’ve spent huge amounts of Federal dollars things like on Medicare and education. So in response, the Democrats appeal to Republican voters disgusted by the soaring Federal deficit.
I think the blocs that will vote Democratic or Republican every time, regardless of candidate or platform, are getting smaller, while the bloc of undecided voters is getting larger. But I also think “undecided” misrepresents a large chunk of this bloc: many voters are more conflicted than undecided. For instance, some voters (including me) were turned off by Bush’s social agenda, but supported his foreign agenda. These voters had to decide which issue was more important to them.
Lastly, Paulos’s article ignores a rule that I learned in college poli sci class and have found quite instructive ever since: most people who develop an attachment to one party or the other do so because of some major life event. Millions of Americans became lifelong Democrats because of the Depression and FDR’s New Deal; millions of others left the Democrats for good because of the civil rights struggles. I believe 9/11 was just such a momentous event, and may have altered the political balance in this country for generations.
One other thing: I believe the closeness of the election can be chalked up to the weakness of Bush as a candidate. If it had been, say, John McCain in the White House these past four years, and events had unfolded exactly the same, the Republicans would have won by 10 or 15 points, rather than 3 and a half. Thus I’d say the only real strategy for the Democrats is to sit back and wait for the GOP to overreach, which they almost certainly will do. If they pass the FMA, or overturn Roe, they will push a lot of “national security” Republicans into the Democratic camp, and the pendulum will swing right back the other way. That’s about all the Democrats can do, anyway.
Paulos article omits important additional information: you can deduce something about an election if you have a lot of other information as well - the Cognressional shift to the GOP since 1994, the rise of GOP governors in otherwise democratic states (like New York and Mass), etc.
Chance the Gardener: your argument is undermined by the fact that the results were closer (both in absolute numbers and in percentages) in other states, and not always in Bush’s favor. Take Wisconsin: Kerry won the 10 electoral votes there by about 14,000 votes — less than half a percent of the total. To say that Ohio’s 130,000 is the only margin that matters is silly.
Paul said that ‘In the Senate the problem is simple: little, almost uninhabited states like North Dakota get the same two Senators as New York or California. And most of these little empty states vote Republican… The Democratic minority in the Senate probably represent at least half the population.’
Well, if you assume that on balance each senator represents half the population of his or her state, you find that there are 55 Republicans for 143 million Americans, and only 44 Democrats for almost 147 million Americans.
In other words, if you live in the backarse of beyond your vote counts for a lot more than if you’ve got more than a handful of neighbours.
As I’ve said on my site, it only takes 2,601,589 people to elect a Republican to the Senate, but you need a whole 3,337,420 people to get a Democrat there.
Paul Gottlieb - Just what is the Democratic Party expected to about Senate representation? It’s constitutionally guaranteed. Removing Senate representation from a state requires the express consent of the State. Even if you had permission from an individual state, other states of similar population makeup are required to pass any constitutional amendment.
Then there is the political component. Forget party viability for any political party even seriously suggesting this. Who in their right mind would vote for the party that advocates removing elected representation?
Bunk. He ignores a fundamental point: This election wasn’t a contest to see who could win the undecideds, the coin flippers. It was a contest to see who could get more of their decideds to actually show up at the polls. Which is why we had such a high turnout.
Just thought I’d chime in since I have a rare chance to say that I’m in complete agreement with Brett Bellmore: this election was a get-out-the-vote contest, not a convince-the-ditherers contest. But I think this supports, not debunks, JAP’s conclusion: if the difference in this election is that Democrats weren’t able to match the Republican 72-hour GOTV storm, then Paulos is perfectly right to say there’s little evidence for a “seismic shift” promising an era of Republican dominance. The question for Democrats now is then “What’s the best way to knock on doors and make sure people vote?” and maybe “How can we block redistricting and other procedural changes designed to keep Democrats out of office?” not “How can we overhaul our party so as to appeal more strongly to religious voters/ Hispanics / religious security soccer Hispanic moms?”
I think Paulos’s own reason for his conclusion is weaker — I’m just not confident there are enough data points for him to place any weight on the discrepancy between the regression analysis and the actual result. But I haven’t thought much about that analysis, so maybe it’s sturdier than I imagine.
No one’s mentioned what I thought was the best thing in the article — JAP’s observation that we’re biased against believing that really important things happen by chance, even when this is true. I’ll bet ogmb upthread is right to say that the weather in Cleveland was a bigger factor in the election than almost anything people have been writing op-eds about.
Some responses here seem to take Paulos’ comments to mean he thinks the result was random. I thought he was warning against drawing strong conclusions, given that the undecideds could decide in a different direction in some future contest. Saying that the election was “about” the get-out-the-vote effort is to take it from the perspective of what the campaigns did. While the campaigns had an impact on the outcome, the number of undecided voters in the weeks prior to the election was sufficient to determine the outcome of the election. That doesn’t mean either part can let down in its GOTV effort. However, calls for making big changes to the complexion of the party in a post-loss bout of self-flagelation are not supported by the numbers.
“if the difference in this election is that Democrats weren’t able to match the Republican 72-hour GOTV storm, then Paulos is perfectly right to say there’s little evidence for a “seismic shift” promising an era of Republican dominance.”
If Democrats couldn’t match Republican get out the vote efforts this year, with one of the more unpopular Republican presidents in recent history, with a war going on that is highly unpopular with Democratic voters, with the largest youth vote in the history of the US, with more money coming into the Democratic coffers than ever before, how can you fail at least suspect that something is going on?
Mark Ames’ take on what the Democrats need to do now: The eXile Solution to Middle America
Lol. He is so right.
Look, Bush is a lousy, maladroit President who’s pushing lots of policies that lots of people disagree with. Yet the Dems still managed to lose. That’s pretty sad.
My own hunch: it had less to do with slightly more than half the people voting for Bush than it did with slightly more than half the voters not liking what the Dems were selling.
You really should do better than that.
So, yeah, sure: you do deserve to have some doubts, a crisis of confidence, and a deep self-reevaluation. Can’t hurt. Lots of people out there just don’t like you. It may be worth wondering why.
And can I humbly suggest that deciding that the reason they dislike you is that they’re stupid is a wrong, or at least unproductive, tack to take? (And that your tendency to call them stupid may be part of the reason they dislike you?)
I didn’t vote for Bush myself, so don’t look at me that way.
Sorry: just needed to get that off my chest.
It’s true that the election was very close, but then the candidates’ positions were also very close—forced towards the center by the need to appeal to independent, moderate and undecided voters. The interesting question is, “to what degree is each party reconciled to the centrist positions that both major candidates were obliged to embrace?”
More here.
À Gauche
Jeremy Alder
Amaravati
Anggarrgoon
Audhumlan Conspiracy
H.E. Baber
Philip Blosser
Paul Broderick
Matt Brown
Diana Buccafurni
Brandon Butler
Keith Burgess-Jackson
Certain Doubts
David Chalmers
Noam Chomsky
The Conservative Philosopher
Desert Landscapes
Denis Dutton
David Efird
Karl Elliott
David Estlund
Experimental Philosophy
Fake Barn County
Kai von Fintel
Russell Arben Fox
Garden of Forking Paths
Roger Gathman
Michael Green
Scott Hagaman
Helen Habermann
David Hildebrand
John Holbo
Christopher Grau
Jonathan Ichikawa
Tom Irish
Michelle Jenkins
Adam Kotsko
Barry Lam
Language Hat
Language Log
Christian Lee
Brian Leiter
Stephen Lenhart
Clayton Littlejohn
Roderick T. Long
Joshua Macy
Mad Grad
Jonathan Martin
Matthew McGrattan
Marc Moffett
Geoffrey Nunberg
Orange Philosophy
Philosophy Carnival
Philosophy, et cetera
Philosophy of Art
Douglas Portmore
Philosophy from the 617 (moribund)
Jeremy Pierce
Punishment Theory
Geoff Pynn
Timothy Quigley (moribund?)
Conor Roddy
Sappho's Breathing
Anders Schoubye
Wolfgang Schwartz
Scribo
Michael Sevel
Tom Stoneham (moribund)
Adam Swenson
Peter Suber
Eddie Thomas
Joe Ulatowski
Bruce Umbaugh
What is the name ...
Matt Weiner
Will Wilkinson
Jessica Wilson
Young Hegelian
Richard Zach
Psychology
Donyell Coleman
Deborah Frisch
Milt Rosenberg
Tom Stafford
Law
Ann Althouse
Stephen Bainbridge
Jack Balkin
Douglass A. Berman
Francesca Bignami
BlunkettWatch
Jack Bogdanski
Paul L. Caron
Conglomerate
Jeff Cooper
Disability Law
Displacement of Concepts
Wayne Eastman
Eric Fink
Victor Fleischer (on hiatus)
Peter Friedman
Michael Froomkin
Bernard Hibbitts
Walter Hutchens
InstaPundit
Andis Kaulins
Lawmeme
Edward Lee
Karl-Friedrich Lenz
Larry Lessig
Mirror of Justice
Eric Muller
Nathan Oman
Opinio Juris
John Palfrey
Ken Parish
Punishment Theory
Larry Ribstein
The Right Coast
D. Gordon Smith
Lawrence Solum
Peter Tillers
Transatlantic Assembly
Lawrence Velvel
David Wagner
Kim Weatherall
Yale Constitution Society
Tun Yin
History
Blogenspiel
Timothy Burke
Rebunk
Naomi Chana
Chapati Mystery
Cliopatria
Juan Cole
Cranky Professor
Greg Daly
James Davila
Sherman Dorn
Michael Drout
Frog in a Well
Frogs and Ravens
Early Modern Notes
Evan Garcia
George Mason History bloggers
Ghost in the Machine
Rebecca Goetz
Invisible Adjunct (inactive)
Jason Kuznicki
Konrad Mitchell Lawson
Danny Loss
Liberty and Power
Danny Loss
Ether MacAllum Stewart
Pam Mack
Heather Mathews
James Meadway
Medieval Studies
H.D. Miller
Caleb McDaniel
Marc Mulholland
Received Ideas
Renaissance Weblog
Nathaniel Robinson
Jacob Remes (moribund?)
Christopher Sheil
Red Ted
Time Travelling Is Easy
Brian Ulrich
Shana Worthen
Computers/media/communication
Lauren Andreacchi (moribund)
Eric Behrens
Joseph Bosco
Danah Boyd
David Brake
Collin Brooke
Maximilian Dornseif (moribund)
Jeff Erickson
Ed Felten
Lance Fortnow
Louise Ferguson
Anne Galloway
Jason Gallo
Josh Greenberg
Alex Halavais
Sariel Har-Peled
Tracy Kennedy
Tim Lambert
Liz Lawley
Michael O'Foghlu
Jose Luis Orihuela (moribund)
Alex Pang
Sebastian Paquet
Fernando Pereira
Pink Bunny of Battle
Ranting Professors
Jay Rosen
Ken Rufo
Douglas Rushkoff
Vika Safrin
Rob Schaap (Blogorrhoea)
Frank Schaap
Robert A. Stewart
Suresh Venkatasubramanian
Ray Trygstad
Jill Walker
Phil Windley
Siva Vaidahyanathan
Anthropology
Kerim Friedman
Alex Golub
Martijn de Koning
Nicholas Packwood
Geography
Stentor Danielson
Benjamin Heumann
Scott Whitlock
Education
Edward Bilodeau
Jenny D.
Richard Kahn
Progressive Teachers
Kelvin Thompson (defunct?)
Mark Byron
Business administration
Michael Watkins (moribund)
Literature, language, culture
Mike Arnzen
Brandon Barr
Michael Berube
The Blogora
Colin Brayton
John Bruce
Miriam Burstein
Chris Cagle
Jean Chu
Hans Coppens
Tyler Curtain
Cultural Revolution
Terry Dean
Joseph Duemer
Flaschenpost
Kathleen Fitzpatrick
Jonathan Goodwin
Rachael Groner
Alison Hale
Household Opera
Dennis Jerz
Jason Jones
Miriam Jones
Matthew Kirschenbaum
Steven Krause
Lilliputian Lilith
Catherine Liu
John Lovas
Gerald Lucas
Making Contact
Barry Mauer
Erin O'Connor
Print Culture
Clancy Ratcliff
Matthias Rip
A.G. Rud
Amardeep Singh
Steve Shaviro
Thanks ... Zombie
Vera Tobin
Chuck Tryon
University Diaries
Classics
Michael Hendry
David Meadows
Religion
AKM Adam
Ryan Overbey
Telford Work (moribund)
Library Science
Norma Bruce
Music
Kyle Gann
ionarts
Tim Rutherford-Johnson
Greg Sandow
Scott Spiegelberg
Biology/Medicine
Pradeep Atluri
Bloviator
Anthony Cox
Susan Ferrari (moribund)
Amy Greenwood
La Di Da
John M. Lynch
Charles Murtaugh (moribund)
Paul Z. Myers
Respectful of Otters
Josh Rosenau
Universal Acid
Amity Wilczek (moribund)
Theodore Wong (moribund)
Physics/Applied Physics
Trish Amuntrud
Sean Carroll
Jacques Distler
Stephen Hsu
Irascible Professor
Andrew Jaffe
Michael Nielsen
Chad Orzel
String Coffee Table
Math/Statistics
Dead Parrots
Andrew Gelman
Christopher Genovese
Moment, Linger on
Jason Rosenhouse
Vlorbik
Peter Woit
Complex Systems
Petter Holme
Luis Rocha
Cosma Shalizi
Bill Tozier
Chemistry
"Keneth Miles"
Engineering
Zack Amjal
Chris Hall
University Administration
Frank Admissions (moribund?)
Architecture/Urban development
City Comforts (urban planning)
Unfolio
Panchromatica
Earth Sciences
Our Take
Who Knows?
Bitch Ph.D.
Just Tenured
Playing School
Professor Goose
This Academic Life
Other sources of information
Arts and Letters Daily
Boston Review
Imprints
Political Theory Daily Review
Science and Technology Daily Review