Flipping coins

If you think there have only been two flips of this particular coin–as you’ve been arguing in the CT threads–you’ve basically been paying no attention to the political history of American society since 1972.

Not to mention that the coin flip metaphor–which you don’t even seem to think is particularly metaphorical–is so insanely inappropriate in the context of ANY election, let alone this one, that it boggles my mind that your thinking runs along these tracks. If too many others are running along the same tracks, I think you can pretty much guarantee that this isn’t the last loss of this kind.

Well, if you are talking just a coin, of course I would only spend a couple weeks rather than an entire month. However, we are talking about every branch of our government, an increasingly compromised mainstream media, and the likelihood of greater assault on the institutions of progressivism. A little bit of strategery might be in order to figure out how to brace ourselves for the coming onslaught.

If you only were allowed to flip the coin every four years, and the fate of the free world depended on the outcome, wouldn’t you check to see if you could, say, blow on it to get a better result next time?

But each election is ~100,000,000 coin flips, no?

But each election is ~100,000,000 coin flips, no?

i repeat myself when under stress
i repeat myself when under stress
i repeat myself when under stress

It’s easy-peasy for you to be blithe about it. IIRC, your government complained when its citizens were labeled “enemy combatants” and then jailed indefinitely and without access to counsel; ours asserted that such was its right, as part of its war-making powers.

C) more than two, less than 100 million.

Look, the 1980s Labour party in the UK had a real problem and needed to change. They were bringing in about 30% of the vote and lost four elections on the spin. The democrats have a losing streak of two, and polled 49%. Surely there has to be some proportionality between problem and solution?

But each election is ~100,000,000 coin flips, no?

But each election is ~100,000,000 coin flips, no?

Nothing is wrong with the coin except that its a metaphor for a vastly more complex system, the human voter.

They lost to a chimpanzee

Labour lost to John Major.

Without any real gain in analytical sophistication, you can model the vote share as a normal variate. It’s still the case that nobody can tell the difference between 49% and 51% in two tries.

It’s true that there’s not a big difference (except in election results) between an electorate which is 51 percent for Bush and one that is 49 percent. I was making the case for remaining pessimistic with exactly this reasoning in that distant time so many hours ago when it seemed like Kerry would most likely win. A country where roughly half the voters favor a lying weasel who is bad on nearly every issue is a country with some sort of serious problem.

I think it’s more complex than the coin flipping model captures. A British General Election can be captured by coin-flipping: we all vote for MPs, aware that the party leader with the most MPs becomes PM (of course it’s meant to be more complicated than that, and it never is), and a few of us in any one year also get to vote for local councillors. The British system really isn’t set up to allow voters to consciously vote for one set of guys in the executive, another set in the legislature and let the two sets hammer out compromises on the composition of the judiciary and on other matters.

The American system is. There is plenty of evidence that the American people vote- often consciously by ‘splitting the ticket’- to achieve a divide in the governing parties in the executive and legislature. (The average British voter doesn’t know what the hell ‘splitting the ticket’ means, with good reason.)

But right now, anyone who voted for Bush almost certainly realised that this would lead to a rightwing Republican President; a rightwing Republican Senate majority; a rightwing Republican Congress majority, and one unprecedentedly disciplined; and that the President and Senate majority together could and would combine to appoint a rightwing Republican Federal Judiciary.

And they went ahead and voted for it. This isn’t normal: I’m too knackered to type out the details tonight, but FDR didn’t enjoy this lack of political challenge, and neither has a single of his successors, with the arguable exception of LBJ in the two years after 1964.

Hmmm, a coin toss? That’s an interesting analysis of a political process that has never once exuded the random walk hypothesis but I think I get the jist of you point…a tyranny of the 52% majority? Or is it that you think the coin (voting process) is “unfair?”
First, remind me, what percentage of the pop vote did Clinton receive in 92 when the coin first came up heads?
Second, land masses, Bush won 30 states to Kerry’s 20, so maybe the coin lands 3:2 in favor of tails? Or the urban vs. rural analysis is much more outspoken, I do not however have the correct odds for that coin, but my projection would be in around 3:1 in favor of tails in rural areas…maybe we should no longer allow sufferage for people who live in a population of LESS than, say, 1 million in a concentrated area. That way dems are guaranteed “fairness of the coin.”
In closing, maybe if you quit blaming you own failures on “a coin,” the election process, disenfranchisement, corporate scandals, the evil wealthy class, etc. and put some belief in the triumph of the human spirit in times of failure, then you may find your coin showing Heads more often…but as it is right now, the party of heads does not hold much belief in that spirit. To them IT IS a metaphorical coin toss, heads you win, tails you lose, and in one of the most prosperous and socially scalable nations in the world, you limit our existence and our potential to a coin toss. Sad. I’m really happy Frederick Douglass never believed his life was limited to this.

Land masses? Do you honestly think voting representation should be based on acreage rather than population (I realize it is somewhat)? And with regard to counting states, if California were subdivided into a dozen smaller states but each voter’s vote was unchanged, would that drastically change your perception of the election results? Maybe what you advocate should be called something other than democracy.

Andrew, your comment could be interpreted as meaning that you think it’s a realistic goal for either side to get 60% of the popular vote; this can’t be right, can it?

With all respet, Daniel, you’re slipping around now. If we’re talking about the electorate, then we’ve got, as many have pointed out, far more than two trials here — we’ve got over 100 million. A 51-49 outcome on 100 million trials does suggest a somewhat weighted coin.

In your initial post, where there are two trials in the past 5 years, rather than 200 million, seemed to be talking about election results as the reponse variable. In that case, I do think winning 60% of elections ought be possible. The Liberal party in my native Canada does it far more than that. There are plenty of other examples in functioning democracies of parties winning far more than 60% of elections.

When someone as bright as Daniel prefers to model human volition as a random coin toss, we have to question his critcal faculties.

This election was not a coin toss, just as the one I recently voted in was not, and just as Australia’s 1975 hammering of Whitlam was not a coin toss.

An election outcome is not a cooperative solution because there are losers. Nevertheless there are a million things in front of you to do to make the world a better place. Your influence on people’s volitional choices is your own free choice. The outcomes can be as good as your work for them.

Go for it.

They lost to a chimpanzee

that chimp was supported by a huge and very effective propaganda machine. it wasn’t W out there by himself making speeches on street corners.

If you flip a coin four times and it comes up heads, heads, tails, tails …

This example is only as good as the starting year you pick, and it’s hard not to notice you conveniently choose to begin just after the Democrats’ twelve years out of office. This problem really exists for any year you pick, but arguably the best starting point would be the moment when modern electoral politics emerged, as Nixon’s 1968 Southern Strategy won over the South to the Republicans, dramatically changing the US landscape and creating the realities we know today as Kerry lost even New Mexico on Tuesday. Since then the picture has been Tails, Tails, Heads, Tails, Tails, Tails, Heads, Heads, Tails, Tails. Save for a Carter and a couple of Clintons, it’s a forty year dearth of Democrats. Why you think the Democratic Party should be complacent about this I don’t know.

A 51-49 outcome on 100 million trials does suggest a somewhat weighted coin

Now you’re assuming independence between votes, which does seem like a stretch.

When someone as bright as Daniel prefers to model human volition as a random coin toss, we have to question his critcal faculties

I hear that they’re modelling mountains as pieces of paper down at the cartogrpahy department.

Cartography and modelling mountains: well yes, thats actually closely related to my work. And from a dispassionate distance the coin toss is an interesting model for a 51-49 outcome, and your point well worth discussing further.

ie: IS remodelling the Left actually necessary?

Wel, as a student I believed my leftie friends were mostly very caring people, while many of the Right were sub-adolscent jerks.

Thirty years of experience have convinced me that having the right motives do not mean your models of human behaviour are right.

The coin toss is independent to a collossal degree. Election outcomes are non-independent. The 51-49 numbers LOOK like coins but 45% rusted-on Labor/Left plus 45% brain-dead reactionaries leaves 10% to decide for themselves and they are not independent; they are diverse, spread-out herds. Each herd is subject to influence and influences others.

And I know you know this.

You point is good: no need to throw the left out or even remake it, because 49% will still vote for them even if they become apologists for mass murder.

Just shutting up for four years would guarantee an election because all anyone will remember is the screw-ups of the administration. But can the far Left STFU? Let’s see who wins next.

In a country with a uniform federal voting system, yes. In a country with a patchwork system that’s heavily balanced for rural areas, no.

Larger point: do you think a left coalition that can’t win the white working class is viable?

I think in 2000 the coin landed on its edge but some cheating b*stard nudged the table.

Anyhow only one way to find out – keep tossing that coin.

‘A 1-2% swing in either Ohio or Florida and it would have been the Republicans having this conversation.’

A 1-2% swing in either Ohio or Florida and it would have been the Republicans in control of the Congress, the Senate, with a majority on the Supreme Court and with an effective Senate veto on any liberal, or possibly even centrist, judicial nominees. Not to mention Republican control of the majority of Governors’ mansions. Not to mention the fact that there would have been an utterly shameless campaign against an ‘illegitimate’ President who hadn’t gained the popular vote.

President Kerry would not have been allowed to pass anything, except possibly water.

How about because (by the way, this is one of a whole range of arguments that could be deployed in favour of a sensible stochastic model of voter behaviour, rather than a silly deterministic one):

1) It was, in fact, very close
2) In close elections, the result is usually dependent on turnout
3) Turnout is dependent on a number of factors
4) Unless one believes that the policies of the opposition party can affect demographics and the weather, most of the factors affecting turnout ought to be modelled as random if one is creating a model of voter preferences.

I don’t ever use the word “Dude!”, so your attempt to parody my argument might work better if you didn’t use it either. Your argument would also work better if it didn’t depend on denying the plain fact of the matter that this was a very close election. That doesn’t leave you with much to work with other than “is it because I speak so authoritatively?”, but that’s actually quite a good gag, so I have high hopes for your rewrite.

A 700,000 swing in seven states would have given Kerry a 325-213 win (numbers very approximate). That’s a best-case scenario, of course.

Besides the series of Presidential reaces, there’s also been a steady 20-year-plus decline in Democratic power in Congress.

Part of Rove’s genius is eking out close victories by a frighteningly efficient ground game and the strategic placement of resources and dirty tricks exactly where needed. So the ideal case above would be much more likely to happen for Republicans than Democrats.

This kind of technical election mastery makes it possible for an extreme candidate to stay in office. Jesse Helms never won a blowout election. I believe that the Rove philosophy is that if you win by too much, it means that some of your political capital has been lying idle. Bush could have had an easy victory if he had played to the center, but relying on his base meant that he could be as radical as he wanted to be, which for Rovians is the reason for being in office.

The fact that Bush is not a real conservative is irrelevant. He has a program. He’s suckering everyone.

And I’d like to make clear that I’m not saying that the probability of one side winning an election “like this” is necessarily not “close to 50%”. I’m disputing the assertion that it is. I have no idea what the probability is. I can think of some ways that one might try to quantify such a thing, but your argument above (steps 1 through 4) doesn’t cut it. If we had a little complete world simulation and could play the election over and over again, I have no idea how much the results would vary. Neither do you.

Why the US is different (overlapping reasons):

1. Peculiarities in the pattern of early immigration (Scotch-Irish and Puritans), later only partly ameliorated by other immigrant groups (Germans, Scandinavians, Jews, Italians, Irish, etc.)

2. The heritage of slavery, together with the heritage of revolutionary and frontier violence.

3. 63 years of continuous military mobilization since 1941. I lot of Americans got most of their political education from the military.

4. No experience of occupation or war on home ground, making wars seem more fun than they really are.

By now the problem is in large part in the electorate, more than with the Democrats’ approach. This doesn’t mean that things are hopeless, but we need to moderate our hopes.

I believe that 1-3 above all distinguish the US from Canada and Australia, which is so similiar in many ways.

why do left-wing American parties do so much worse than their counterparts world-wide?

The French socialists haven’t done all that well, and the Italian Socialists are still in the doldrums after the Craxi affair ten years ago. I also take issue with your implied claim that conditions in the USA were ideal for unseating an incumbent; a khaki election is usually favourable, and whatever we think about the economy, Ray Fair’s model did not suggest that the economy would be a minus for Bush.

With respect, Keith, that’s exactly why I propose to model them as a random variable.

The French socialists haven’t done all that well, and the Italian Socialists are still in the doldrums after the Craxi affair ten years ago.

Yes, but we haven’t had a viable party that’s dared to use the word ‘socialist’ as anything other than an epithet in my lifetime.

This isn’t a new problem relating to this election, of course, but it does make it easy to leap to the conclusion that the Democrats have some massive structural problem: why are the social democratic ideals that are appealing to voters worldwide so comparatively ineffective as vote-getters to Americans?

My reasons why The Left doesn’t do very well in American elections cover the whole last 60 years. Since about 1941, the left has never done well in the US. In Europe, sometimes it does, sometimes not.
(Second try after error message).

Keith, you’re addressing quite deep issues here, with which I have a certain degree of sympathy, since I use similar arguments myself in economic contexts. I’ll try and deal with this at greater length in another post.

My reasons why The Left doesn’t do very well in American elections covers the whole last 60 years. Since about 1941, the left has never done well in the US. In Europe, sometimes it does, sometimes not.

Yes, and I think your reasons are pretty convincing. Unfortunately, what they aren’t is either changeable or suggestive of any course of action.

There may not be any fundamental problem with the Democratic party that is any more fixable than the reasons you list, but I think the hope that there might be, if we could just put our fingers on it, is what tempts people to look for major structural changes. It feels as if, if only we could just find the magic bullet, suddenly Middle America would start voting its economic interests.

(BTW, as a very occasional commenter on a lot of the blogs you comment on, I’d like to say how useful and informative I generally find your posts.)

Okay. Well, again, I’m not saying you’re necessarily wrong. My gut instinct is that US presidential elections are too “different” from each other to provide enough useful data to pick out the “noise” in the “signal”. But, truly, I have no idea. I can imagine some sophisticated and exhaustive methods that could pin it down to a range, anyway. Nevertheless, I’m arguing against your point on the basis on which you’ve defended it. And, so far as I can tell, that’s merely on the recent history of US presidential elections, the margin of this election, the assumption that the last election was very “like” this one…and that’s about it. My impression has been that you’re arguing on a gut basis that the small margin in this election is within the noise range. And it may be. It also may well not be.

If we could replay this election in the sense that you’re hypothesising—that is, leave everything the same except those supposedly random exogenous factors—frankly I wouldn’t be surprised to see a variation as small as within one-tenth of one-percent or a variation as large as ten percent. I just don’t know. And my point is that that 2% margin doesn’t mean anything by itself.

It’s pretty sad if the left – the left! – has to hope for EC technical wins in swing states to overcome a 3.5 million vote Bush margin.

If we were talking about elections in Sweden or the Netherlands, I would say yes, pretty sad. Given that we’re talking about America, the most capitalist, most religious and least leftwing developed economy in the world, I’d say it’s pretty good going.

That would be my contribution to the “Democrats and religion” debate – if people really want to see a world in which America has a structural social-democratic majority, as opposed to occasional progressive victories against a background culture of conservative capitalism, then yes, they probably had better start praying.

(Jason; I mean this with all respect and not at all sarcastically that you will probably have a happier life if you relocate from the USA. I have three or four Yank friends who have done this, and as far as I can tell, none of them have regretted it yet).

I think it was Peter Cook that explained to a British audience that there are two political parties in America – “The Republicans, who are the equivalent of our Conservative Party, and the Democrats, who are the equivalent of our Conservative Party”.

My Canadian brother is recruiting family members like crazy. At least one niece seems ready to go, and maybe my son. Me too, if they’ll have me.

I’ve really come to believe that some of the things I don’t like about the US are structural. So I feel a duty to a degree to keep things from getting even worse, but not much hope that they’ll get much better. It looks as if we’ll be spending the next year figuring out how to save Social Security, for example.