Jump to Any Race
National Politics|By Charlie Cook, April 13, 2017
This story was originally published on nationaljournal.com on April 10, 2017

Wichita is 1,250 miles from Wash­ing­ton, D.C., but you can bet quite a few people in­side the Belt­way will be check­ing elec­tion re­turns from there Tues­day night.

Nor­mally, few out­side of Kan­sas would care much about the res­ults of a spe­cial elec­tion to fill a staunchly Re­pub­lic­an House seat like Kan­sas’s 4th Dis­trict, a va­cancy caused by Mike Pom­peo’s de­par­ture to be­come CIA dir­ect­or. After all, how com­pet­it­ive could a dis­trict be that voted for Don­ald Trump by a 27-point mar­gin (60 to 33 per­cent for Hil­lary Clin­ton) and four years earli­er for Mitt Rom­ney by 26 points (62 to 36 per­cent for Pres­id­ent Obama)? Old-timers will re­mem­ber this dis­trict as the one once held by Demo­crat Dan Glick­man, who won in 1976 as Jimmy Carter de­feated Pres­id­ent Ford and lost in the Re­pub­lic­an tid­al-wave elec­tion of 1994 to Todd Ti­ahrt. Ti­ahrt then served eight terms be­fore giv­ing the seat up to run un­suc­cess­fully for the U.S. Sen­ate—los­ing to fel­low GOP Rep. Jerry Mor­an in the Re­pub­lic­an primary—and has lost two sub­sequent ef­forts to re­claim it.

Cook Polit­ic­al Re­port House Ed­it­or Dav­id Wasser­man puts the chance that Demo­crat­ic nom­in­ee James Thompson will beat Re­pub­lic­an state Treas­urer Ron Estes in this race at 30 to 40 per­cent. But pros will be watch­ing the mar­gin closely to see wheth­er there is something to this idea that Demo­crats are build­ing a large and in­tense en­thu­si­asm gap over Re­pub­lic­ans, even in ruby-red con­ser­vat­ive dis­tricts. The 4th Dis­trict has a Cook Polit­ic­al Re­port Par­tis­an Vot­ing In­dex (PVI) rat­ing of R+15, mean­ing that over the last two pres­id­en­tial races, this dis­trict voted 15 points more Re­pub­lic­an than the coun­try as a whole, mak­ing it the 74th-most-Re­pub­lic­an vot­ing dis­trict.

Hav­ing been the state treas­urer for sev­en years dur­ing a peri­od of great fin­an­cial dif­fi­culty for the state and a school fund­ing crisis cer­tainly could un­der­cut Estes’s strength. If his vic­tory mar­gin over Thompson, a civil rights at­tor­ney, is much un­der a dozen points, one can as­sume the anxi­ety level among Re­pub­lic­ans will height­en go­ing in­to a far more com­pet­it­ive April 18 spe­cial elec­tion in Geor­gia’s 6th Dis­trict.

In the seat va­cated by Tom Price, who be­came sec­ret­ary of Health and Hu­man Ser­vices, Mitt Rom­ney’s mar­gin over Pres­id­ent Obama was 24 points in 2012 (61 to 37 per­cent) but Trump beat Clin­ton by about a point and a half (48 to 47 per­cent). The PVI for the 6th Dis­trict is R+8, mak­ing it the 165th-most-Re­pub­lic­an dis­trict in the coun­try. While this Geor­gia seat should also be re­li­ably Re­pub­lic­an, it isn’t nat­ur­al Don­ald Trump ter­rit­ory; it has a high pop­u­la­tion of up­scale, col­lege-edu­cated whites, many ori­gin­ally from out of state, more com­fort­able with a Rom­ney, John Mc­Cain, or George W. Bush than Trump. If there is a path to a 218-seat House ma­jor­ity for Demo­crats, it will more likely be through up­scale sub­urb­an dis­tricts, not the rur­al and small-town-ori­ented dis­tricts in which they used to be very com­pet­it­ive.

The Kan­sas and Geor­gia spe­cial elec­tions, along with the Vir­gin­ia gov­ernor race, are the ca­nar­ies in the coal mine that ana­lysts will be ex­amin­ing for hints of what’s to come in the 2018 midterm elec­tions. Pres­id­en­tial elec­tion years have be­come in­creas­ingly more par­lia­ment­ary, as every 2016 Sen­ate elec­tion and 400 out of 435 House races voted the same way as the pres­id­en­tial races in those con­stitu­en­cies. While our midterms are get­ting more par­lia­ment­ary rather than the “I vote for the per­son, not the party” ap­proach that voters used to claim to fol­low, there is also a very strong ref­er­en­dum ele­ment to non-pres­id­en­tial cycles. A polit­ic­al party hold­ing a ma­jor­ity in the House and hav­ing a sit­ting pres­id­ent with ap­prov­al rat­ings in the 30s or lower has the po­ten­tial for real prob­lems.

It is ab­so­lutely true that there are many few­er com­pet­it­ive con­gres­sion­al dis­tricts than there used to be. The num­ber of swing dis­tricts—seats that have a PVI of between D+5 and R+5, with­in 5 points of the coun­try as a whole—dropped from 164 in 1997, when the PVI was first cal­cu­lated, to just 72 now, a very steady pro­gres­sion down­ward caused by a com­bin­a­tion of pop­u­la­tion sort­ing, people choos­ing to live in en­virons with like-minded voters, and par­tis­an ger­ry­man­der­ing—draw­ing le­gis­lat­ive and con­gres­sion­al dis­trict lines with par­tis­an in­tent. But while this has had an ef­fect of di­min­ish­ing the volat­il­ity in the House, it does not in­ocu­late a party from the im­pact of an un­pop­u­lar pres­id­ent, a leth­ar­gic base, and an en­er­get­ic op­pos­i­tion party.

Lower voter turnout in midterm elec­tions means that who chooses to show up is vi­tally im­port­ant. If one party’s base is pas­sion­ate, or bet­ter yet, angry, and the oth­er party’s base is com­pla­cent or dis­ap­poin­ted, big and of­ten un­usu­al things can hap­pen. The im­pact of a wave elec­tion may be less than it used to be, but it doesn’t mean that it can’t be a wave. The kind of wave that once would have gen­er­ated pres­id­en­tial-party losses of 63 House seats (2010), 55 seats (1946), 54 seats (1994), 48 seats (1958, 1966, and 1974), or 30 seats (2006) might trans­late in­to losses of just two or three dozen with today’s con­gres­sion­al dis­trict bound­ar­ies and pop­u­la­tion-dis­tri­bu­tion pat­terns. But when just 24 seats will flip the House, it can cer­tainly hap­pen.

The late Demo­crat­ic House Speak­er Tip O’Neill fam­ously pro­nounced, “All polit­ics is loc­al.” In wave years, as op­posed to nor­mal “all polit­ics is loc­al” elec­tions, there is a cas­cad­ing ef­fect, with seats flip­ping that would nev­er flip in a nor­mal polit­ic­al en­vir­on­ment. That’s why my vari­ation of the O’Neill ad­age is, “All polit­ics is loc­al, ex­cept when it’s not.” Nor­mal years are loc­al­ized; wave years are not. The ques­tion is wheth­er any giv­en elec­tion will be­come na­tion­al­ized, will be a wave elec­tion—wheth­er it has the dy­nam­ics that can turn a nor­mal year in­to one that isn’t.