Jumping To Conclusions

Everyone links to this column, by political analyst John Ellis, who has begun handicapping the new, post-Gore Democratic presidential field for 2004. He suggests that South Carolina will be pivotal, winnowing the field of candidates down to two, and that those two are likely to be Dick Gephardt and John Kerry, with Joe Lieberman probably dropping out at that point. (He doesn't even mention John Edwards, which is interesting in itself.) And he's completely dismissive of Howard Dean and Wesley Clark.

But what he fails to address -- what so many analysts fail to address -- is the history of presidential elections. I fully recognize that John Ellis knows many more insiders than I do. If he tells us what these people are saying and thinking and planning, I'll buy it. But I can read history as well as anybody, and what I read is this: members of Congress don't win. Oh, sometimes they get nominated, though even that's pretty rare. But they don't get elected. Isn't that an important piece of information?

Since 1900, there have been 26 presidential elections. Unless I've missed someone, exactly two of those were won by someone coming directly from Congress. The most recent one was John Kennedy in 1960, forty years ago. (The other was Warren Harding, in 1920.) Indeed, if you expand the field to look at the losers of these elections, you only add three more: Barry Goldwater in 1964, George McGovern in 1972, and Bob Dole in 1996. (I only consider the major party nominees, for simplicity's sake.) So out of 52 election slots, we see just 5 sitting senators, of whom only two won.

Now, there's nothing deterministic about these statistics; there's no scientific law which prevents congressmen from becoming president. But doesn't the fact that nobody has been elected to the presidency from the House of Representatives since James Garfield in 1880 suggest something about Dick Gephardt's chances? (Indeed, barring an error on my part, I believe that was the last time a congressman even won his party's nomination.) Doesn't the fact that four decades have passed since a senator got elected tell us that John Kerry, Joe Lieberman, and John Edwards might not be in as strong a position as people think?

Yes, I concede that it seems pretty absurd right now to suggest that Howard Dean could actually win the nomination, much less the presidency. But on the other hand, who would have guessed in 1990 that Bill Clinton would have been nominated? Did Mike Dukakis really look like the strongest Democratic candidate in 1986? Of course not. But governors are far more likely to be nominated, and win election, than senators are. So why is this factor always ignored? If I had to guess, I'd say that most pundits live in Washington and get their information from Washington sources. And who do Washington sources know? Washington politicians. I doubt they spend much time in Little Rock, or Montpelier, or Albany. So they're not really in a position to assess the governors' strengths and weaknesses, and the governors don't have people dropping their names every day around journalists and pundits. Maybe that's too simplistic -- but regardless, I'm not going to bet the mortgage money against Howard Dean.