What Value are Endorsements?

by Brian on February 22, 2008

After the Super Tuesday primaries in the U.S., there was a lot of discussion that various big-name endorsements seem to have not made much difference. Most notably, despite being endorsed by Governor Patrick and Senators Kennedy and Kerry, Barack Obama got beaten heavily in Massachusetts. But what struck me at the time, and what seems to have been confirmed by subsequent contests, is that (at least in Democratic primaries) mayoral endorsements seem to make an enormous difference in the campaign. Not only does the candidate with the most endorsements seem to routinely win, they seem to outperform their poll numbers.

Chris Bowers has posted a table comparing how each state’s primary compared to the pre-election poll average. I’m ignoring caucus states, where there was typically little polling and for whatever reason Obama has dominated so heavily that it is hard to draw any conclusions about comparative factors. Apart from that, Clinton’s best performances, both absolutely and relative to polling, have largely come in states where she has had major local endorsements. These include

  • Massachusetts (Boston)
  • New Mexico (Albuquerque)
  • California (Los Angeles, San Francisco, Oakland, Sacramento)
  • New Jersey (Trenton, Elizabeth, Bayonne, Patterson)

Note that New Mexico is the only state other than New Hampshire where Obama has lost despite leading in pre-election polling. It’s true that in New Jersey Obama was supported by some prominent local officials, including the Mayor of Newark. But my impression from talking to some people who know New Jersey well is that the vast majority of local officials (including a large number in Newark itself) were supporting Clinton.

On the other hand, Obama’s best performances, especially his best performances relative to polls, have come in states where he has some crucial endorsements. These include

  • Georgia (Atlanta)
  • Connecticut (New Haven)
  • Virginia (Richmond, Norfolk, Alexandria, Roanoke, Charlottesville)
  • D.C. (Washington)
  • Maryland (Baltimore)
  • Wisconsin (Milwaukee, Madison)

The only state that really doesn’t fit this pattern is Missouri, where Clinton lost badly (and crucially) in St Louis despite being endorsed by the mayor. Still, on the whole it seems like endorsements from major mayors is worth several points compared to pre-election polling.

When I first started thinking about this, I thought it would be an indicator that Clinton would start to comeback on March 4th, because she had some crucial endorsements in upcoming states, including the mayors of Philadelphia and Providence. But two things happened in the last two days that make that judgment less clear.

First, Obama was endorsed by the mayor of Cleveland. Now both Clinton (Akron, Canton, Parma, Toledo) and Obama (Cleveland, Columbus, Youngstown) have solid local support. If Obama can put up a huge win in delegate rich Cleveland, that will make it hard for Clinton to have the massive win she really needs in Ohio to get back in the game.

Second, Clinton now seems to have alienated the mayor of Providence. I assume mayoral endorsements matter not because of their persuasive powers, but because mayor’s have a GOTV machine that they can put to work on their candidate’s behalf. If Mayor Cicilline is unhappy enough to simply not work hard for Clinton, it will make Rhode Island a much closer state. Though it should be noted that Cicilline’s political opponents in Providence are, as far as I can tell, politically similar to the groups who supported Clinton heavily in Massachusetts. So maybe she’ll have enough residual support to win Rhode Island anyway, but without the game-changing margin that a huge win could have produced.

Still, combined with Obama’s endorsement by the mayor of Austin, it’s getting harder to see where Clinton is going to stage her comeback.

Lists of endorsements are largely taken from Wikipedia for Clinton and Obama, with extra help from Raising Kaine and 13th Floor.

{ 2 trackbacks }

AMERICAN NONSENSE » Midday Open Thread
02.22.08 at 8:55 pm
News from Rob » Midday Open Thread
02.23.08 at 1:24 am

{ 27 comments }

1

David in Nashville 02.22.08 at 4:22 pm

One could also throw in South Carolina. While there are scarcely any current statewide Democratic officeholders, Obama had the enormously popular and universally respected Mayor of Charleston, Joe Riley, in his corner–as well as important statewide figures such as former state Superintendent of Education Inez Tenenbaum.

2

otto 02.22.08 at 4:47 pm

I heard it argued that the main effect of the Kennedy endorsement was to abruptly end the ‘ghettoization’ of Obama that was HRC’s tactic at the time.

3

Witt 02.22.08 at 4:54 pm

I’m not convinced that the Philadelphia endorsement will do that much for Clinton. It’s significant, especially in terms of the unions and other Democratic institutions, but there are a lot of other factors.

4

Paul 02.22.08 at 4:56 pm

One potential hiccup here is that while Clinton got the Menino endorsement here in Boston, Obama actually won Boston, and that while Obama got the two major state-wide endorsements in Kerry and Kennedy, Clinton won the primary by taking the rest of Mass big.

5

Brian 02.22.08 at 5:24 pm

Good point Paul. That does undermine the story somewhat. Although Obama winning 53-45, as I now see that he did, isn’t as big as he’s won other urban centres.

I was sort of guessing that Clinton’s wins in the Boston suburbs, which was where she ran up some of her biggest margins, were in part due to Menino (or Menino-like) machine backing. (By biggest margins, I mean things like the 35 point win in Quincy and the 38 point win in Braintree. It wasn’t just western Mass where she did well.) But it would be useful to have more actual evidence that there was a disparity in support from local officials, and it was causally relevant, to support the claim.

6

rootlesscosmo 02.22.08 at 5:32 pm

Another qualifying data point: Obama carried San Francisco by about 7 points, despite Clinton’s having been endorsed by the popular mayor, Gavin Newsom.

7

Brian 02.22.08 at 5:56 pm

True, but there I really think the endorsement helped Clinton. Compared to how Obama has done elsewhere in urban areas, holding him to a 7 point lead in SF is a great victory for her.

8

Paul Ding 02.22.08 at 6:18 pm

I suspect that endorsements do about as much good for politicians as they do for commercial products.

Chris Matthews says Democrats fall in love, Republicans fall in line, so the chief advantage of the Caroline/Teddy Kennedy endorsement has been to get people to take a *serious* look at Barack Obama.

Endorsements haven’t worked as well for McCain, because it hasn’t been long since he was in the doghouse with the theocrats for that “agents of intolerance” line, and they haven’t worked as well for Huck because, well, he isn’t getting many endorsements, possibly because he doesn’t seem to make intolerance a central theme to his candidacy.

The endorsement from the Teamsters might help Barack Obama. It says “we think we can work with this guy.” But at this point, Barack pretty well has the nomination tied up, so it doesn’t matter. I suspect that in their secret meeting, Barack asked Edwards to sit tight until the convention, because if he waits, Edwards would have more influence over a deadlocked convention than if he were viewed as being an Obama partisan.

9

R 02.22.08 at 6:38 pm

One reason for the power of mayoral endorsements is that mayors tend to be well-aligned with the groups that tend to lead activist-GOTV drives. So it may not be so much that people give great weight to what the mayor thinks, but that the mayors endorsement reflects the thinking of the city’s political community.

10

Brian 02.22.08 at 7:07 pm

R’s thesis might well true about be what’s driving this. (Correlation is once again not sufficient for causation.)

What I think is almost certainly true that there’s it’s the GOTV drive that’s crucial, not the popularity of the endorser. Teddy Kennedy is really really popular in South Boston, and his guy got blown out of the water there.

That is, I don’t think there’s a lot of evidence that endorsements change people’s minds. What I do think is that there is evidence that having people who know how to supporters into voters onside is important.

One related thing that I should have stressed in the post. I think one really fascinating question out of Super Tuesday was why Obama did so much better in Connecticut than he did in Massachusetts. (Nick Beaudrot has a very nice map of where the candidates did well in CT.)

I suspect having local networks, like the mayoral office in New Haven, onside was part of the solution. Maybe the networks that had been created to support Lamont were part of the solution. But whatever the answer, this seems like a question that will be worthy of political research when the history of this campaign is written.

11

jacob 02.22.08 at 7:47 pm

One thing to remember about New Haven is that John Destefano is not only the mayor but was the Democratic candidate for governor last time around, after beating the mayor of Stamford (I think) in a somewhat bitter primary. I suspect Destefano’s networks, including his GOTV network, is thus broader than just New Haven.

12

Ted 02.22.08 at 8:05 pm

One explanation on St. Louis:

Mayor Slay is wildly unpopular here after his firing of the African-American head of the Fire Dept. He held a shaky coalition before that, but his influence is done. The machines of Reps. Carnahan and Clay (both of whose districts cover STL) and Sen. McCaskill trumped whatever Clinton-support Slay could muster.

13

Davis X. Machina 02.22.08 at 8:26 pm

I was sort of guessing that Clinton’s wins in the Boston suburbs, which was where she ran up some of her biggest margins, were in part due to Menino (or Menino-like) machine backing.

Try white flight. Weymouth (‘Southie, with trees”), Braintree, Saugus, etc. the older, otherwise reliably democratic, not-rich suburbs is where Patrick under-performed

Suburban Boston, in towns without Volvo dealerships, is like Faulnker’s South, where the past isn’t dead, it isn’t even past. And it’s always 1975.

14

John 02.22.08 at 9:30 pm

In terms of Philly, I don’t think Nutter’s endorsement will mean that much. Nutter got elected more or less in opposition to the Democratic machine, and his relationship to it is tenuous and questionable.

The key local political figure in Philadelphia is Congressman Bob Brady, chairman of the city Democratic Party. Brady has not yet endorsed. He’s the one to look out for. Given that he’s a white guy in Philly who hasn’t yet endorsed Clinton, I’m going to guess that she’ll need substantial victories in Ohio and Texas to get him to back her.

15

thompsaj 02.22.08 at 10:21 pm

@1, yeah the peninsulares love Riley, but ask anyone from James Island!

16

Scott Hughes 02.22.08 at 10:32 pm

This is interesting. Maybe we don’t notice some of the value that endorsements provide. Without them, maybe the politicians would have done even worse in those states.

17

Brian 02.22.08 at 11:02 pm

Davis’s explanation (@13) is probably right. I may have been reaching a little to explain the voting patterns around Boston when it was all somewhat simpler than I made it out to be.

And that’s interesting to know about Philly, John. I wonder if it ends up being like North Jersey (with the sides reversed). One candidate had some high profile leaders, the other had the majority of the people who ran the political machines. And the second candidate won. If Brady supports Obama and brings a lot of people with him, Clinton is in trouble. A big win for Obama in Philly and it’s hard to see him losing the state by enough to make any serious difference to the delegate count. And then, it’s all over.

18

ikl 02.22.08 at 11:26 pm

Clinton didn’t do especially well in Western Mass, except for greater Springfield (where she had local backing from Congressman Neal). But she lost Franklin and Hampshire counties and Berkshire county was pretty close. That was not where she won MA, anyway.

19

nick s 02.23.08 at 1:16 am

There’s been an argument that long-standing city machines (of which Chicago, of course, is the classic example) can deliver votes, but in machine-politics areas like north NJ there was reporting that suggested otherwise.

Those machines, however, do have racial bases, and the anecdotal experience recorded by Josh Marshall’s correspondents — that many nth-gen ‘white ethnic’ voters simply won’t vote for a black man — doesn’t necessarily need a machine.

(On the MA endorsements: Obama came a lot closer than the polls were suggesting before those endorsements came in. RI’s going to be different still, given the size of the state in relation to Providence.)

20

Lee Sigelman 02.23.08 at 1:18 am

One of my co-bloggers at “The Monkey Cage,” John Sides, posted on this topic back in November, pointing to a new analysis that suggests that endorsements do matter. Click <a href=here for John’s post.

21

Lee Sigelman 02.23.08 at 1:20 am

Sorry. That link showed up nicely in preview and I can’t seem to crack the code here. It’s
http://www.themonkeycage.org/2007/11/why_endorsements_matter_in_pre.html

22

Answer Guy 02.23.08 at 4:35 pm

Re #19…. Voters in general, and Democratic primary voters in particular, in many smaller towns in Western Mass tend to be pretty liberal. Voting patterns there resemble Vermont more than they do the rest of Mass. The “it’s always 1975″ factor (an often forgotten but thankfully diminishing factor in election throughout the NE Corridor, as there are actually a lot of places like this between Boston and DC) referred to in #13 doesn’t exist out there in the same way.

The reason some of the bigger towns west of I-495 (Boston’s outer beltway) went for Hillary were the presence of more working-class voters and a bigger Puerto Rican population. For some reason, Clinton just cleaned Obama’s clock in that community; she won every town with a big PR population – Holyoke, Lawrence, and Worcester to name three.

None of this bodes especially well for Obama in Rhode Island, which is a lot like Massachusetts except without any “Baja Vermont.” Or for “beer track” Democratic voters in Ohio and Pennsylvania.

Except that Maryland was a very different story, which I’ll explain in another post since this is getting longwinded.

23

Answer Guy 02.23.08 at 5:17 pm

So I went at looked at Clinton vs. Obama in Maryland, specifically Baltimore County and Anne Arundel County, the two counties that contain the most white working-class Democratic voters in the state. They don’t believe in incorporated towns much in Maryland, so the area analysis would suck – except that Maryland thankfully breaks down by Congressional district within counties. In other words…precision gerrymandered Congressional districts to the rescue!

Baltimore County has pieces of 5 of the state’s 8 U.S. House districts. (The other 3 are entirely in the DC orbit.)
The 1st and 6th are the designated GOP zones. With regards to Balt Co, the area in the 6th is rural while the 1st is mostly heavily Republican outer suburbia. Neither have many of the voters I’m focusing on. Collectively, Clinton beat Obama narrowly in this territory, a pattern repeated intereestingly enough throughout “red” counties statewide.
The 7th is a black majority district based in the city, and the portions of Balt Co in the 7th (West Side – Woodlawn, Lochearn) are majority black as well. Obama won there by more than 3:1.
The 2nd and 3rd are the real points of focus.
The 2nd contains most of the East Side of the county (Dundalk, Essex, Rosedale, White Marsh) and then has an arm that includes parts of Towson that extends out to cover some territory in the northwest. Obama beat Clinton here 29.2K-23.5K, a pretty substantial win not far from his performance statewide. The east side isn’t all white of course, but white blue-collar is the dominant ethic in this district nonetheless. The Balt Co portion of the 3rd district consists of most of the North Side of the county (from Towson over to Owings Mills and up to Reistersown) just above the city, with an arm that includes Parkville and Overlea (usually considered East Side) and another part, joined by Baltimore City part of the 3rd, down on the South Side (Arbutus/Halethorpe.) Obama didn’t do quite as well here (fewer blacks than the 2nd, and I get the feeling Hillary did very well among the big Jewish community on the NW Side) but still carried the territory overall, 16.3K-15.4K.

The Anne Arundel County numbers betray a similar pattern, somewhat reversed. Most of the AA portion that’s in the 2nd (Brooklyn, Ferndale, pt of Glen Burnie) is quinessentially Reagan Democrat, and Obama won it 7331-5689. Obama did even better in the AA section of the 3rd (18.3K-13.8K), though that portion includes Annapolis, which is pretty diverse, and has an arm to the west that takes in a fair number of voters who are far more likely to be DC focused. There was a dead heat in the AA parts of the 1st (an area specifically designed to include as many Republicans as possible)and the 5th (another Republican area, except the 5th is dominated by the Democrats of Prince George’s and Charles Counties, so it’s blue overall.)

24

John 02.23.08 at 7:22 pm

Wow, Answer Guy, interesting stuff. The basic upshot, then, is that Obama did pretty well in the white working class areas?

Also – Jesus Christ, Maryland congressional districts sure have been gerrymandered, huh?

25

Answer Guy 02.23.08 at 8:09 pm

Yeah, far as I can tell, Obama did pretty well in working-class white areas, such as they exist, in Maryland. Which suggests that something changed between Super Duper Tuesday and Potomac Tuesday, or that Obama just had lots more boots on the ground in Baltimore.

26

sunflower7 02.23.08 at 11:52 pm

Endorsements don’t mean much in this society. People don’t question things. They just follow. It’s not as if the people are scrutiny these endorsements and making and informed decision.

27

Hogan 02.25.08 at 2:41 pm

I’d say there is no key figure in Philadelphia; the Democratic party here is a network of neighborhood machines and union GOTV operations who don’t much like each other, and Brady is more a broker of short-term deals and truces than a leader. (Witness his 15% vote in the mayoral primary last year.) There may end up being a consensus choice, but it will be the result of independent decisions by the warlords, not of people falling in line behind Nutter or Brady.

Comments on this entry are closed.