The Sen­ate races are start­ing to heat up as the pri­mary elec­tion sea­son winds down.

Here is this month’s map:

Now for the details.

 

“Con­tin­u­ing” refers to the seats in Sen­ate Classes 2 and 3, which are not up for elec­tion this cycle.

The big story this time is Indi­ana, where Repub­li­can incum­bent Sen­a­tor Richard Lugar fell this week in the Repub­li­can pri­mary elec­tion to Richard Mour­dock, who will face off against Rep­re­sen­ta­tive Joe Don­nelly (D-​​Granger, IN), who cur­rently rep­re­sents Indiana’s Con­gres­sional Dis­trict 2. This takes the seat from a vir­tu­ally cer­tain hold to a tossup. But how do we arrive at this con­clu­sion? How does one hand­i­cap this race?

In my case, I’m start­ing by rec­og­niz­ing that this is a statewide office, which means we need to look at the statewide his­tor­i­cal vote break­down. This can be accom­plished by look­ing at the Cook Par­ti­san Vot­ing Index (PVI), which for Indi­ana is R+6. This is a mod­er­ately Repub­li­can num­ber (com­pare it to, say, Texas at R+10, Okla­homa at R+17, or Utah at R+20). One can, in fact, con­vert the PVI to a roughly equiv­a­lent DW-​​NOMINATE score to com­pare the vot­ing record of a mem­ber of Con­gress to the state’s vot­ers’ posi­tion on the polit­i­cal spec­trum. To do so, I treated PVI R+20 as equiv­a­lent to a DW-​​NOMINATE of +1.00, and a PVI of D+20 as equiv­a­lent to a DW-​​NOMINATE of –1.00, and scaled every­thing else lin­early between the two.

Using this con­ver­sion fac­tor, Indiana’s vot­ers come to a DW-​​NOMINATE of +0.30. How does this stack up against the two lead­ing can­di­dates for Sen­ate in November?

We can pretty directly gather Donnelly’s DW-​​NOMINATE, since he has served two com­plete terms in the House. His DW-​​NOMINATE is –0.05 (I’m round­ing to the near­est twen­ti­eth, since the PVI scale is no finer than that), plac­ing him 0.35 to the left of the state’s vot­ers. (Inci­den­tally, he’s 0.15 to the left of his district’s vot­ers, but redis­trict­ing has made the dis­trict more con­ser­v­a­tive, which is one fac­tor influ­enc­ing his deci­sion to run for the Sen­ate seat.)

Fig­ur­ing Mourdock’s DW-​​NOMINATE is more dif­fi­cult, since he has never served in Wash­ing­ton. He has, how­ever, given us sev­eral clues in his cam­paign­ing to date. The two most telling are his stance of zero com­pro­mise, and his stated intent to be most like Sen­a­tor Jim DeMint (R-​​SC). The lat­ter is par­tic­u­larly valu­able, since we do know DeMint’s DW-​​NOMINATE; it’s +0.85, higher than any Sen­a­tor in the 111th Con­gress other than Sen­a­tor Tom Coburn (R-​​OK). Assum­ing Mourdock’s com­paign­ing is accu­rate, that would put him 0.55 to the right of Indiana’s voters.

In other words, Mour­dock appears to be fur­ther to the right than Don­nelly is to the left of the state’s vot­ers. And not just a lit­tle fur­ther — we’re talk­ing about a 57 per­cent greater dis­tance. This sug­gests that if Mour­dock doesn’t mod­er­ate his tone in the next few months, he’s likely to find him­self los­ing to Don­nelly, though prob­a­bly by only about three or four points.

I’m wait­ing for polling to sup­port my analy­sis. In the mean­time, I’m putting Indi­ana in the “Tossup” cat­e­gory. And, if you recall “The Ide­ol­ogy Gam­ble”, that’s prob­a­bly just fine for many Repub­li­cans, par­tic­u­larly within the Tea Party faction.

Here are some other high­lights of the past month:

  • Nevada: Both Pub­lic Pol­icy Polling and Ras­mussen show incum­bent Repub­li­can Sen­a­tor Dean Heller with a lead over his likely (the pri­mary elec­tion is June 12) Demo­c­ra­tic chal­lenger, Rep­re­sen­ta­tive Shel­ley Berkley (D-​​Las Vegas, NV). Account­ing for Rasmussen’s house bias, it appears that Heller has about a five-​​point lead, mov­ing Nevada from “Tossup” just a lit­tle into the “Leans Repub­li­can” column.
  • Florida: The Sun­shine State shifted pretty far left this month, with both Pub­lic Pol­icy Polling and Ras­mussen show­ing double-​​digit leads for Demo­c­ra­tic incum­bent Sen­a­tor Bill Nel­son over his most likely Repub­li­can chal­lenger, Rep­re­sen­ta­tive Con­nie Mack, IV (R-​​Fort Myers, FL). This takes Florida from the “Tossup” col­umn and puts it instead solidly in “Leans Democratic”.
  • Wis­con­sin: A new Ras­mussen poll shows a 12-​​point lead for for­mer Gov­er­nor Tommy Thomp­son, who is con­sid­ered the most likely Repub­li­can nom­i­nee, over sole Demo­c­ra­tic can­di­date Tammy Bald­win. Dis­count­ing Rasmussen’s house bias still ties the Tammy-​​Tommy tiff to Thomp­son ter­ri­tory, but the state remains rather lightly polled, so I’d call it just barely in “Leans Republican”…for now.
  • Mis­souri: Ras­mussen had another recent poll here, too, show­ing Repub­li­can for­mer state Trea­surer Sarah Steelman’s lead over Demo­c­ra­tic incum­bent Sen­a­tor Claire McCaskill shrink­ing to seven points. With the house effect, this returns to “Tossup”.
  • Maine: for­mer Gov­er­nor Angus King, run­ning as an inde­pen­dent, showed a com­mand­ing lead in the one poll pub­lished by the Maine People’s Resource Cen­ter. It’s hard to imag­ine any­one else com­ing close.
  • Mass­a­chu­setts: Ras­mussen shows Repub­li­can incum­bent Sen­a­tor Scott Brown tied with pre­sump­tive Demo­c­ra­tic nom­i­nee Eliz­a­beth War­ren. Mass­INC gives War­ren a two-​​point edge. With those con­flict­ing data, it’s still a “Tossup”.
  • Ari­zona: Rep­re­sen­ta­tive Jeff Flake (R-​​Mesa, AZ) seems most likely to secure the Repub­li­can nom­i­na­tion. For­mer US Sur­geon Gen­eral Richard Car­mona is the only declared Demo­c­rat in the race. While Flake appeared to have a solid lead a month ago, more recent polling sug­gests that it may be nar­row­ing. As with Mour­dock in Indi­ana, Flake’s PVI-​​rounded DW-​​NOMINATE of +1.00 puts him far to the right of the state’s pro­jected value of +0.3, based on the PVI of R+6. It’s dis­tinctly pos­si­ble that, as Ari­zo­nans gain addi­tional under­stand­ing of the two can­di­dates, Carmona’s more mod­er­ate views will lead to a Demo­c­ra­tic pickup here. This is a race to con­tinue to watch.

Four tossups, evenly split on both sides. Repub­li­cans need three of them to cap­ture the Sen­ate. Intrade is giv­ing them a 56 per­cent chance of doing it. How cred­i­ble do you think that is?

Related arti­cles