Turning the Tables

The pundits are at it again. They’re impressed that John Edwards took John Kerry to the wire in Wisconsin. They’re surprised that Edwards defied polls suggesting Kerry would blow him out. They’re intrigued that Edwards beat Kerry among independents. But a win is a win, they say, and Kerry has won nearly every contest.

Among self-identified Democrats, that’s true. But among independents and Republicans who have chosen to vote in Democratic primaries, the record is very different. In 10 of the states that have voted so far, the media have conducted systematic exit or entrance polls that clarified how independents voted. In seven of those states, exit polls have also measured how self-identified Republicans voted. What percentages of these voters have Kerry and Edwards won, respectively? Let’s look at the numbers:

Table 1	Independents		Crossover Republicans
.	Edwards	Kerry	Edwards	Kerry
Iowa	25	34	N/A	N/A
N. Hampshire	13	37	9	29
Arizona	6	33	N/A	N/A
Delaware	12	42	N/A	N/A
Missouri	28	37	37	30
Oklahoma	34	18	30	7
S. Carolina	48	22	54	16
Tennessee	34	32	28	22
Virginia	31	41	45	13
Wisconsin	40	28	44	18

Among independents in the exit-polled states, Kerry has beaten Edwards in six contests; Edwards has beaten Kerry in four. This month, the candidates are tied with four wins apiece. Since Feb. 10, Edwards has won two primaries to Kerry’s one.

The pattern among crossover Republicans is more lopsided. Kerry has won one contest; Edwards has won six. This month, Edwards has beaten Kerry among Republicans in all six states in which Republican votes were measured.

Remember, Democrats are as likely to vote for Edwards against President Bush as they are to vote for Kerry against Bush. It’s far more likely that independents and crossover Republicans will determine the outcome. In states where the choices of these groups have been measured, Edwards is matching Kerry among independents and beating him among crossover Republicans.

Fine, you say. But independents and crossover Republicans don’t control Democratic primaries. Democrats do, and they’re voting for Kerry.

That’s true. But the exit polls show that, by and large, Democrats aren’t voting for Kerry because they prefer him on the issues. They’re voting for him because they think he’s the Democrat most likely to beat Bush. What happens if they find out he isn’t? What happens if they realize that Edwards is doing as well as Kerry among independents and is doing better than Kerry among crossover Republicans?

Again, the numbers tell the story. In nine states, voters were asked in exit polls to clarify whether they chose their candidate primarily because he “agrees with you” on the issues or because he “can defeat Bush.” Here’s how Edwards and Kerry performed, respectively, among the “can defeat Bush” voters and the “agrees with you” voters.

Table 2	“Can defeat Bush” voters		“Agrees with you” voters
.	Edwards	Kerry	Edwards	Kerry
N. Hampshire	11	56	14	29
Arizona	6	59	7	32
Delaware	10	71	12	35
Missouri	21	68	29	39
Oklahoma	24	43	32	19
S. Carolina	42	42	48	22
Tennessee	24	58	34	31
Virginia	18	69	28	42
Wisconsin	28	59	39	29

Among “agrees with you” voters, Kerry has beaten Edwards in five contests; Edwards has beaten Kerry in four. This month, they’re even at four apiece. Since Feb. 10, Edwards is ahead, two contests to one. Consequently, if “agrees with you” voters continue their February pattern, and if “can defeat Bush” voters begin to split evenly between Kerry and Edwards, Edwards will beat Kerry in half the remaining states. But if “can defeat Bush” voters discover that Edwards is more electable and begin to vote for him instead of Kerry, Edwards will beat Kerry in most states, and the delegate count will turn.

From a horse-race standpoint, I’d rather be Kerry than Edwards tonight. But if I were, I’d be praying that the folks in Table 2 don’t find out about the folks in Table 1.