Handicapping the Whip Count – Odds Still Against Passage, But Margins Are Thin
March 18, 2010
Handicapping the Whip Count – Odds Still Against Passage, But Margins Are Thin
- We analyze the health reform whip count on the premise that members’ votes are more likely to reflect the politics of their districts than the pressures from leadership; and, that this is particularly likely to be true if opposing members can stand as a group with other members opposing on similar grounds (abortion, immigration, etc.). Specifically, we analyze the odds of ‘un-decideds’ becoming ‘no’s’, and of no-to-yes switches
- Of the 56 un-decided House Democrats, 11 are in toss-up races, and 10 of these are in districts that Obama either lost (5), or won by single-digits (5). Five of the members in toss-up races also are members of an opposing bloc; 2 are members of the Stupak bloc, 4 are Blue Dogs (one member is in both blocs). For each of these un-decided members a yes-vote is potentially career-ending; conservatively we expect 5 or more to vote no, even under the air cover of the Slaughter rule
- Two more of the 56 un-decided are not in competitive races, but are in seats that are relatively secure because of their tendency as Democratic members (Rahall, Altmire) to reflect the attitudes and interests of their conservative (McCain won both by > 10 points) districts. We expect at least one of them to vote no
- Democrats who lean no typically are in close races in more conservative districts. Exceptions include ‘safe’ conservative Democrats in otherwise Republican districts, who tend to be safe because of their conservative voting records – including no votes on HR3962. Gutierrez is a stark exception; Obama won his district by 72 points. Gutierrez opposes because of immigration issues and stands with the CHC, though we believe he’s a likely no-to-yes switch. On net, we expect Gutierrez plus one or more of 3 members with thin margins in more liberal districts to switch
- On net, we expect at least 6 ‘un-decided’s’ to become ‘no’s’, and at most 4 ‘no’s’ to switch to ‘yes’. This would add 2 ‘no’ votes to the current 37 total, yielding 39 no votes, where 38 are needed to defeat the measure.
- We believe our estimate of failure by two votes is reasonably conservative – first, it is the net of the low estimate of ‘un-decided to no’ switches and the high estimate of ‘no-to-yes’ switches. Second, we only assume no votes on grounds of opposing abortion language for members with right-to-life voting records spanning 10 or more years. Certainly more junior members have similarly strong convictions
It’s Called Whipping for a Reason
Narrow-margin whip counts create two layers of uncertainty. At the macro level, the leadership must create a sense of momentum and inevitability, whether momentum exists or not. At the level of individual members, the leadership exerts tremendous pressure, which can shift the votes of members who we have every objective reason to believe would not have shifted. The carrots and sticks are well known – but they’re bigger on this vote than on most. The carrot of Presidential attention – keeping in mind that the targets here are in the President’s own party — is amplified both by this President’s relative popularity (48%, v. 17% for Congress as a whole); and, by the great relevance of winning or losing to this President’s future. The stick – political isolation – is bigger than usual as well. To be labeled as the single Democratic member – or one of a very few – who stood in the way of reform passing is to live in political Siberia – at least within the Beltway.
All Politics is Local, and Strength in Numbers
We begin by assuming that no one who has committed to a yes vote will switch to a no, which creates a more conservative estimate. Rather, we focus on the no’s (firm and ‘apparent’) and un-decideds. We use TheHill.com’s whip count to determine who is in either of these categories; numerous counts are available, but in our view TheHill’s count is best for two reasons – members often will speak directly with TheHill; and, TheHill tracks members’ actions and comments fairly closely, and incorporates these secondary sources into their count.
As of this writing, the ‘no’s’ numbered 36; 2 more no’s (38 total) would kill the bill. Thus the analytical question becomes one of how many of the remaining 50-odd ‘un-decideds’ switch to no; and, how many of the current ‘no’s’ convert to yes.
As has been our approach throughout this debate, and as is consistent with decades of empirical evidence, we place the greatest emphasis on the politics of the member’s district, and what these politics suggest about how a member is likely to vote. Secondarily, we look for strength in numbers, i.e. participation of that member in a bloc that offers the company of other members who stand in opposition to the measure (e.g. Stupak’s bloc, the Congressional Hispanic Caucus (CHC), Blue Dogs, etc.). Beginning with the ‘un-decideds’ – we sort the members first according to the riskiest of the Congressional Quarterly (CQ) or Cook scores for their pending race (from riskiest to safest), and second by the number of points by which the President lost (negative values) or won that member’s district in the 2008 race (from greatest margin of loss to greatest margin of victory). The result is in Exhibit 1. We count 7 undecided members in toss-up races in districts that the President lost, 5 members in toss-up races in a district that the President won by fewer than 10 points, and 1 member in a toss-up race in a district that the President won by more than 10 points. 6 of these 13 members are also part of an opposing bloc, in this case either the Stupak bloc (2) or the Blue Dogs (5, one member is in both blocs). If we conservatively assume the retiring members vote with the leadership – as is largely traditional, despite all 3 having voted against HR 3962 – then we count 11 un-decided members for whom a yes vote on health reform is potentially career-ending. We would expect 5 or more of these members to become firm ‘no’s.’ As we proceed down the list the races become increasingly secure; however we note that both Rahall’s and Altmire’s relative security owes much to their habit of reflecting their conservative constituents’ interests – McCain won Rahall’s district by 14 points, Altmire’s by 11. Rahall is a member of the Stupak ‘bloc’ and Altmire is a Blue Dog. We expect at least one of Rahall or Altmire to become a firm no. Moving even further down the list, the races are securely Democratic and the President is extremely popular – so both local politics and Presidential pressure weigh heavily on these members. We count 6 members of the CHC, 6 Blue Dogs, and 1 Stupak ‘bloc’ member – eliminating overlaps, 9 of the members in safe races in districts that Obama won handily are members of an opposing bloc. We do not rely on any of these undecided members becoming no votes. Thus all in, our rough – and we believe conservative — estimate is for six or more un-decided members to vote no.
Turning to the ‘no’s’, we apply the same logic – local politics first, blocs second – but we sort the members in the opposite order as the un-decideds. Here, we’re looking for members whose local politics pull them away from their no vote, i.e. members in secure races in districts where the President is popular. We sort first by Cook and CQ scores (from safest to riskiest), and then by 2008 presidential margin (from greatest margin of victory to greatest margin of loss). The results are in Exhibit 2. Ten of these races are considered safe by both Cook and CQ – but all of them have a story. Gutierrez is opposing because of restrictions on un-registered immigrants’ rights to purchase care on the exchanges with their own money – i.e., he’s standing with the CHC. Davis is running for governor of Alabama, and the state’s conservative politics effectively require him to vote no if he hopes to be competitive – in fact Davis plans to interrupt his campaign and travel to DC for the express purpose of casting his no vote. Lipinski, Costello, and Stupak all oppose the Senate’s abortion language. Recall that Stupak’s ‘bloc’ consists of ten members who have voted in unison on all 15 major right-to-life priorities over the past decade, with the exception of a single member on a single bill. (We note that Congressman Kildee, who is a member of this bloc, has decided to vote for the bill, though Stupak re-asserted his (and the remainder of his bloc’s) decision to vote no after Kildee’s announcement.) The next 5 ‘solid’ Democrats – Peterson, McIntyre, Matheson, Boren, and Taylor – are ‘solid’ Democrats in otherwise Republican (McCain won all of these) districts for a reason – they vote conservatively, in line with their constituents’ attitudes and interests. And, as Blue Dogs, they have the added benefit of standing with other members who are likely to oppose the measure – 22 Blue Dogs are presently counted as ‘no’s’.
Moving further down the list of ‘no’s’, logic suggests the Democrats’ best candidates for switching are those members in races that are at least tipping in the member’s favor, and where Obama either won, or lost by a reasonably small margin. Six members fit this profile: Herseth-Sandlin, Kissell, Adler, Arcuri, McMahon and Carney. Herseth-Sandlin is rated as ‘likely’ to win her (Alabama) district, but at least part of this is because she voted against HR3962. And, neither the President nor the party can help her a great deal; the President lost her district by 8 points in 2008. The remaining five are in more competitive races (‘leans’) and the President’s 2008 margin was less than 10 percent in all of these districts (he lost two of these districts). Only Arcuri and Carney voted yes on HR3962; arguably the status of the other members’ races owes something to their no votes on ‘3962; i.e., if they switch to yes, they very likely jeopardize their chances for re-election. The balance of the ‘no’ list is made up of members in tight races where the President has limited popularity, and/or where McCain won outright. The potential exception to these pressures is Berry, who is retiring, but who has been increasingly critical of the party since announcing his retirement.
On net, our best estimate is that Gutierrez folds to Democratic pressure and votes yes, as do one or more of Arcuri, McMahon, or Boucher. Adding our two estimates – 6 or more un-decideds vote no, and 1 to 4 no’s switch – we conservatively expect 2 additional no votes, taking the current number of 36 no’s to 38 — the minimum required for the bill to fail. Note that our summation takes our low estimate of new ‘no’ votes and our high estimate of no-to-yes switches. And, note that we do not credit members as being part of the Stupak ‘bloc’ unless they have voted consistently with the bloc for at least a decade. We know that this leaves out some members that are pro-life, and who may be more likely to either become or remain ‘no’ votes than we have concluded. Thus all in, we believe our estimate that the measure fails – though we net to a margin of a single vote – errs on the side of being conservative.
- We’re only focused on Democratic ‘no’s’ here – we’re assuming there are no Republicans who will support the bill
- We prioritize race scores in election years, and particularly in election years after filing deadlines have passed – as those making the scores have the benefit of knowing who the challenger is, and how both the challenger and the incumbent are fairing. In non-election years, we tend to place greater weight on prior years’ election results.
- Where the Cook and CQ scores differ, we use the lower (greater risk of loss) score. We do this because the immediate trend has been for the scores to fall rather than rise, and the two services update at different times.
- We recognize the tactical support that will be given to these members in exchange for voting yes – which includes considerable campaign funding and support from the party, as well as the personal support of the President – though we stress this has to be viewed against the backdrop of electoral results in Va, NJ, and Mass.. More to the point, neither the party nor the President are that popular in the districts in question.
- Our definition of the Stupak bloc requires the member to have been in Congress for at least a decade – there are members who likely carry equally strong pro-life convictions that we don’t capture as members of the bloc.