Big States vs. Red States
This post came out of comments on another thread. Everyone is familiar with Hillary's claims that her primary wins in "big states" demonstrate her electability vis-a-vis Obama's. The Obama campaign and supporters have been countering that Obama's wins in wavering "red states" demonstrate his electability. This argument is elegantly presented in a Daily Kos post. (Thanks to ColoradoJones for the link.) Kos breaks down the electoral college numbers to show that "Obama is the more competitive November candidate for the Democratic Party."
Both arguments, however, are specious. Kos himself writes, "If you were to make the moronic assumption that only the winner of the primary could win those states, that would add up to 74 electoral votes for Obama, 49 for Clinton." But if it's a moronic assumption (and it is a moronic assumption), then how can it lead him to confidently conclude that Obama is more electable than Clinton?
Why is it a moronic assumption? Winning a primary means that you are the favorite among Democratic voters of that state, not that the Democratic voters hate the other folks. After all, somebody has to win the primary. If Obama weren't in the race, Hillary would have won Virginia, and if Hillary weren't in the race, Obama would have won New York. (OK, I'm discounting Edwards, but you get the point.) Consider Massachusetts. There is probably no Democratic primary candidate, even Mike Gravel, who would not win MA in the general election. Therefore, losing the primary to another Democrat does not mean that you will lose the state to a Republican in the general election. And vice-versa, winning the primary against other Democrats does not mean that you will win the state in the general election. The primary winner may have a slightly higher probability of winning the state, but that's very difficult to quantify in a way that allows you to predict the outcome with any confidence.
But Kos has another argument which is not based on the "moronic assumption" He looks at the SurveyUSA matchup polls against McCain to predict who would win which state. But the assumption that whomever is currently outpolling McCain in a particular state will win that state is equally moronic. We've already seen the head-to-head polls fluctuate during the primary season, and the election is still eight months away. Folks, eight months ago, Giuliani was heavily favored in the Republicans polls, and McCain had been all but dismissed. Obama was a distant second to Clinton. The general election campaign hasn't even started. The Democratic-Republican skirmishing has barely begun. We've got hundreds of speeches, press conferences, talk shows, news reports, unpredictable world events, and debates (hopefully not hundreds of those) before the election. The head-to-head polls at this point are almost worthless as an election predictor. Around this time in 2004, polls, showed that voters preferred Kerry by as much as 5%.
So Kos's argument is flawed, but of course, so is Clinton's, since it also relies on the "moronic assumption". No one with an ounce of humility should confidently predict the probability of the general election based on primary or poll results. It may be somewhat more reasonable to look at voter demographic indicators like age, gender, and political affiliation as well as voter preferences on the issues, but even then, I caution against confident conclusions. How does Obama's strength among independents and young voters measure up against Clinton's strength among women and older voters? How does Obama's change message hold up against McCain's security message? You can speculate as much as you like, but in the end, there's no way to know, so it's inappropriate to treat your predictions as anything more than speculations. That goes for both of you, Hillary and Barack.
Both arguments, however, are specious. Kos himself writes, "If you were to make the moronic assumption that only the winner of the primary could win those states, that would add up to 74 electoral votes for Obama, 49 for Clinton." But if it's a moronic assumption (and it is a moronic assumption), then how can it lead him to confidently conclude that Obama is more electable than Clinton?
Why is it a moronic assumption? Winning a primary means that you are the favorite among Democratic voters of that state, not that the Democratic voters hate the other folks. After all, somebody has to win the primary. If Obama weren't in the race, Hillary would have won Virginia, and if Hillary weren't in the race, Obama would have won New York. (OK, I'm discounting Edwards, but you get the point.) Consider Massachusetts. There is probably no Democratic primary candidate, even Mike Gravel, who would not win MA in the general election. Therefore, losing the primary to another Democrat does not mean that you will lose the state to a Republican in the general election. And vice-versa, winning the primary against other Democrats does not mean that you will win the state in the general election. The primary winner may have a slightly higher probability of winning the state, but that's very difficult to quantify in a way that allows you to predict the outcome with any confidence.
But Kos has another argument which is not based on the "moronic assumption" He looks at the SurveyUSA matchup polls against McCain to predict who would win which state. But the assumption that whomever is currently outpolling McCain in a particular state will win that state is equally moronic. We've already seen the head-to-head polls fluctuate during the primary season, and the election is still eight months away. Folks, eight months ago, Giuliani was heavily favored in the Republicans polls, and McCain had been all but dismissed. Obama was a distant second to Clinton. The general election campaign hasn't even started. The Democratic-Republican skirmishing has barely begun. We've got hundreds of speeches, press conferences, talk shows, news reports, unpredictable world events, and debates (hopefully not hundreds of those) before the election. The head-to-head polls at this point are almost worthless as an election predictor. Around this time in 2004, polls, showed that voters preferred Kerry by as much as 5%.
So Kos's argument is flawed, but of course, so is Clinton's, since it also relies on the "moronic assumption". No one with an ounce of humility should confidently predict the probability of the general election based on primary or poll results. It may be somewhat more reasonable to look at voter demographic indicators like age, gender, and political affiliation as well as voter preferences on the issues, but even then, I caution against confident conclusions. How does Obama's strength among independents and young voters measure up against Clinton's strength among women and older voters? How does Obama's change message hold up against McCain's security message? You can speculate as much as you like, but in the end, there's no way to know, so it's inappropriate to treat your predictions as anything more than speculations. That goes for both of you, Hillary and Barack.
