How McCain Won Michigan

Going into Tuesday's primary election in Michigan, pollsters were calling it a dead heat between George W. Bush and John McCain.

Yet by 8:30 p.m., only half an hour after the polls closed, ABC News called the election in McCain's favor. Their analysis of exit polls showed the following:

Yes, Michigan has an open primary. That allows Democrats to vote in a Republican primary, and vice versa. It also allows those of use without a fixed party affiliation to vote. (I've campaigned for both Democrats and Republicans in the past. In 1980, I voted for John Anderson.)

Some might say open primaries skew the system, since people can vote for a weaker candidate of the opposing party. But in Michigan that's not a big deal - we've always had a strong group of moderates who refuse to embrace either party label. I think an open primary is more fair to those who have no party affiliation or are disillusioned with the party they usually support. I also believe it is more representative of potential outcome in November, especially when the field has essentially narrowed to two candidates.

(No, I haven't forgotten Alan Keyes. However, at this point in the process, he is no longer a viable candidate. Maybe he'll get the vice presidential nomination.)

Michigan is strongly Democrat in the Detroit, Flint, and Ann Arbor side of the state, where unions and University dominate. Republicans dominate most of the rest of the state. For as long as I can remember, we've sent one Democrat to the Senate and one Republican. It's the independents who make the difference in close Michigan elections - and it's the independents who helped McCain defeat George W.

Despite (or perhaps because of) strong support by the Michigan Republican Party and Governor Engler, Bush didn't carry the day. Two-thirds of traditional Republicans supported him, but the independent and Democrat participation in the primary showed the Texas governor doesn't have broad support outside his party. Come the November election, that will count for more than Republican support.

Why? Because over 90% of Republicans are going to vote the party ticket, whether that's Keyes, Bush, or McCain. That pretty much assures that the Republican candidate will take 40-45% of the vote in November.

Likewise, almost all self-identified Democrats will vote the party line for Gore or Bradley. That's a safe 40-45% for the Democratic presidential nominee.

Here in Michigan, it will be the independents who make the difference in November. Given the choice between the conservative Bush and the moderate McCain, two-thirds chose McCain on Tuesday. Given the choice between a conservative Republican and a liberal Democrat in November, it could go either way.

But given the choice between a moderate Republican and a liberal Democrat, Michigan is likely to swing Republican in November.

There are two issues at stake: party support and popular support. People like Barry Goldwater, George McGovern, and Michael Dukakis have had the backing of their parties, but their extreme positions drove the independent voters (and sometimes members of their own parties) into the other camp.

I see the American electorate as distributed along a bell curve. There are a few arch-conservatives and a few flaming liberals. There are more "regular" conservatives and liberals. And there are a lot of moderate, middle-of-the-road people who vote the candidate, not the party.

What happens in each election is that each candidate finds a point on the political spectrum. For the more conservative candidate, pretty much everyone to the right of his views will get his vote. For the more liberal candidate, pretty much those to her left will get her vote.

The question is how the voters in between will vote. For the sake of simplicity, let's assume you can draw a line midway between the more conservative candidate and his more liberal opponent. Voters on the right side of that line will almost invariable vote for the more conservative candidate; those on the left, for the more liberal one.

That means the best strategy for winning in November is running a candidate as close to the center as possible, since he is then likely to carry the broad center of the political spectrum against a strongly conservative or liberal opponent. Unless a strong third option becomes available, that should be enough to predict the election.

Let's say 0 represents the arch-conservative and 10 the flaming liberal. George W. is maybe a 2, McCain a 4. Gore and Bradley are probably in the 7-8 range (I really haven't followed the Democrats this year). If the November election is between Bush and a 7.5 Democrat, the midpoint between their views is 4.75, so the middle will swing just enough to elect the Democrat.

But if McCain runs against even a 7.0 Democrat, the midpoint moves toward the left at 5.5, moving enough of the middle to McCain to assure a Republican victory.

Of course, this model oversimplifies to an extreme. I assume a bell curve distribution of voters centered on 5.0, which may not be the case. If the curve is skewed toward the right, Bush could win. If the curve is weighted toward the left, Gore or Bradley could pull off a victory over McCain.

In the end, broad popular support means more for the Republican agenda (getting their man in the White House) than strong party support. In fact, it could easily be argued that the stronger the party support, the more likely the candidate holds more extreme views than the general public, and the more difficult it will be to gain support from the broad center.

Of the major candidates offered by the major parties, not only do I like McCain and his politics, but I believe he is the most electable candidate because of his moderate position on most issues (he usually falls between Bush and the Democrats).

It seems a lot of Michigan voters, especially the important independents, agree with me.

Further Reading

• Mighty Unusual GOP Primary, ABC News, 2/23/2000
• McCain Back in Race after Michigan Victory, Yahoo/Reuters, 2/23/2000