Bar­ron’s thinks Repub­li­cans will retain their major­i­ty in the House:

Our analy­sis — based on a race-by-race exam­i­na­tion of cam­paign-finance data — sug­gests that the GOP will hang on to both cham­bers, at least nominally.

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. We ignore the polls.

This method­ol­o­gy is good only for fer­til­iz­er. The assump­tion that the size of a warch­est cor­re­lates with grass­roots sup­port is invalid.

Here’s a sim­ple exam­ple. Nan­cy Boy­da has raised $428,761 accord­ing to the Cen­ter for Respon­sive Pol­i­tics. Jim Ryun has raised $883,156. Her polling (includ­ing, I’m told, one poll not yet released) shows the race sta­tis­ti­cal­ly tied, as does a pol­ly by the state party. 

Clear­ly, the grass­roots sup­port for Ryun can’t be twice the size of Nan­cy’s. And we can fig­ure out why by look­ing at who con­tributes. Jim Ryun raised $409,792 from indi­vid­ual donors, while Nan­cy Boy­da raised $348,040 from indi­vid­ual donors. They are polling near­ly iden­ti­cal­ly because their grass­roots sup­port is near­ly tied. And that’s also why they are get­ting almost the same amount of mon­ey from indi­vid­ual donors. Jim Ryun is just rak­ing in the cash from the Repub­li­can mon­ey machine.

That illus­trates why Bar­ron’s found that “the can­di­date with the most mon­ey has won about 93% of the time.” Incum­bents win about 93% of the time, and incum­bents can tap into sources of finan­cial sup­port that insur­gent chal­lengers can’t. But that mon­ey does not trans­late to any­thing about the grassroots. 

Indeed some­one polling well while spend­ing less mon­ey indi­cates some­one whose mes­sage does­n’t need a lot of mon­ey to win peo­ple over. And in a cli­mate where peo­ple con­sis­tent­ly report dis­sat­is­fac­tion with Con­gress, with the coun­try’s direc­tion, and with their own rep­re­sen­ta­tives, it will cost a lot more to sell vot­ers on their cor­rupt incumbents.

Bar­ron’s, to their cred­it, acknowl­edges those crit­i­cisms, but responds to them in a way that’s even weak­er. They agree that 1994 and 1974 both showed that angry vot­ers can upset their finan­cial mod­el of elec­tions, but:

We just don’t agree that the out­rage has reached the lev­el of those ear­li­er times. The rea­son is that the econ­o­my in 2006 is health­i­er. And the econ­o­my is the only oth­er fac­tor that fig­ures in our analysis.

For most of what Bar­ron’s does, that atti­tude makes sense. But it does­n’t cut it today, espe­cial­ly in an econ­o­my where a ris­ing tide has scut­tled a lot of boats.