Barron's BS

Brendan Nyhan points out the flaws in the prediction by Barron's magazine that the Republicans will keep control of the House. According to Barron's:

We studied every single race -- all 435 House seats and 33 in the Senate -- and based our predictions about the outcome in almost every race on which candidate had the largest campaign war chest, a sign of superior grass-roots support. . . .

Is our method reliable? It certainly has been in the past. Using it in the 2002 and 2004 congressional races, we bucked conventional wisdom and correctly predicted GOP gains both years. Look at House races back to 1972 and you'll find the candidate with the most money has won about 93% of the time. And that's closer to 98% in more recent years, according to the Center for Responsive Politics. Polls can be far less reliable. Remember, they all but declared John Kerry president on Election Day 2004.

That 93 percent number is right, but only because so many House races lack decently funded challengers. When you look at races where both candidates have adequate resources, the numbers are much different. In his book, The Politics of Congressional Elections, Gary Jacobson looks at the relationship between campaign spending and election results.

	Challenger Spending
Incumbent Spending	<$200K	$2-$400K	$4-$600K	$6-$800K	>$800K	Row Average
<$200K	0	0	0	0	0	0
$2-$400K	0	0	0	0	0	0
$4-$600K	0	6.2	8.7	0	0	0.9
$6-$800K	0	3.1	11.6	35.7	18.2	2.7
>$800K	0.4	7.1	10.5	17	30.3	9.6
Column Average	0.1	5.4	10.1	18	28.5	4.6

From 1986 to 2002, where the challenger spent less than $200,000 (in 2002 dollars) they won only 0.4 percent. Where challengers spent between $200,000 and $400,000, only 5.4 percent won. Challengers spending between $400,000 and $600,000 won 10.1 percent of the time. Between $600,000 and $800,000, challengers won 18 percent of races. And challengers spending more than $800,000 won 28.5 percent of their races.

And this relationship holds, even when challengers are outspent by incumbents. Challengers who spent between $600,000 and $800,000 won 17 percent of the time when up against incumbents who spent over $800,000.

Finally, this analysis looks at all years, regardless of the partisan tides running in a particular election. Jacobson points out that challengers running in "good years" for their party (1974, 1982, and 1996 for the Democrats; 1980, 1984, and 1994 for the Republicans) did even better. In those years, challengers spending more than $700,000 won more than 50 percent of their races.

The Republicans may manage to retain control of the House, but contrary to Barron's prediction, it won't be because of their spending advantage.