Because he wrote this today:
bq. The outcome of Iowa’s first-in-the-nation voting is skewed by two big factors. The turnout is ridiculously small, barely 20 percent of the eligible voters. And those who choose to caucus are hardly representative of the population as a whole
Broder concludes with this statement:
bq. New Hampshire is a more reliable, less distorted lens through which to view the presidential landscape than Iowa.
Is this true? Let’s define “more reliable” as “better predictor of the nominee.” Here are the winning candidates of the Iowa caucus and NH primaries for 1976-2004 (in order by year; a dash indicates no contested election or no candidate preference vote held that year).
IA Dem: Carter, Carter, Mondale, Gephardt, Harkin, – , Gore, Kerry
NH Dem: Carter, Carter, Hart, Dukakis, Tsongas, – , Gore, Kerry
IA Rep: Ford, Bush, – , Dole, – , Dole, Bush, –
NH Rep: Ford, Reagan, – , Bush, Bush, Buchanan, McCain, –
Which state is the best predictor of the eventual Democratic nominee? It’s a tie. IA and NH each were “reliable” 5 out of 7 times.
What about the Republican nominee? IA got it right 3 of 5 times, or 60%. NH got it right 4 of 6 times, or 66.7%.
So, if you want to hang your hat on the difference between 60% and 66.7%, then David Broder is right. I do not want to hang my hat on such a small difference, so I will say that David Broder is wrong. Neither IA nor NH is more reliable.
Related posts:
Would higher turnout produce different election outcomes?
Debunking Myths about the Iowa Caucus
Are Primary Electorates Representative?