The general belief seems to be that it is important to "clear the field" in primaries to get a winning hand in the fall general election. The belief is founded on a number of factors. Many interest groups will not back a candidate with a primary election opponent. Primary election campaigns can be costly and challengers generally have less money to spend than incumbents. Anecdotal evidence points to a number of campaigns easch cycle where a strong primary is followed by disappointing results in the fall.
I decided to test this thesis by looking at election results for all 31 Democratic pickups in the House during the 2006 cycle (including Peter Welch as a pickup in Vermont) and comparing the results to close losses. The close losses were not systenmatic but I looked at 20 races that fit the bill.
Overall, 14 of the 31 pickups (45%) were preceded by primary elections, a higher than expected number. Although some of these were blowouts, a surprising number were close and in many cases surprise winners emerged despite less money. As a comparison, among the 20 close but losing elections only seven (35%) were preceded by primaries and only one of those was close: the Tammy Duckworth-Christine Cegelis- Scott duel in IL-6.
