The Monkey Cage

Here’s a graph of the 50 states (actually, I think Alaska and Hawaii are missing), showing the average economic and social ideology of adults within each state. Each of these is scaled so that negative numbers are liberal and positive are conservative; thus, people in Massachusetts are the most liberal on economic issues and people in Idaho are the most conservative:

West Virginians are on the liberal side economically but are extremely socially conservative, whereas Vermont is about the same as West Virginian on the economic dimension but is the most socially liberal of all the states. Coloradans are economically conservative (on average) but socially moderate (or, perhaps, socially divided; these are averages only).

How do these rankings fit with our usual rankings of states? Here’s a plot showing average economic and social ideology for each state, plotted vs. George W. Bush’s vote share in 2000:

Democrats and Republicans separately

The next step is to break these voters down into Democrats and Republicans (based on self-reported party identification and following the usual practice among political scientists of throwing the “leaners” into the regular party categories). In the graph below, each state is shown twice: the avg social and economic ideologies of Democrats in the state are shown in blue, the avgs for Republicans in red.

We made these graphs during the primary election season, and one thing we noticed was that South Carolina (“SC”) is in the middle of the pack among Democrats and among Republicans, but it’s one of the most conservative states overall. My take on this: South Carolina is a strongly Republican state, and the moderates in South Carolina are likely to identify as Republican. This pulls the Republican average to the left (as they includes the moderates) and also pulls the Democratic average to the left (as they are not including so many moderates).

But the big thing we see from the graph immediately above is that Democrats are much more liberal than Republicans on the economic dimension: Democrats in the most conservative states are still much more liberal than Republicans in even the most liberal states. On social issues there is more overlap (although in any given state, the average Republican is more conservative than the average Democrat).

Details on data

David Park and I made these graphs from the Annenberg pre-election survey from 2000 (with its huge sample size), creating indexes based on issue opinions, giving each respondent an economic and social ideology score. We scaled these so that each had a national average of 0 and standard deviation of 0.5. (We used these scales in our Red State, Blue State book, but these particular graphs never made it into the book.)

P.S.

Yes, I know the graphs could be better. We made them a few months ago and haven’t organized them into any final form.

Posted by Andrew Gelman at 09:35 PM | Permalink | Comments (3)

Back in February, I posted on the left-handedness of recent presidents, here. That posting captured the fancy of the entire nation — well, maybe not, but it did spark an unusually high number of responses (some speculative, some research-based, some confessional, as in “I’m left-handed”).

Now that both the Democrats and the Republicans have done the right thing (er, the left thing) by naming portsiders Obama and McCain as their presidential candidates, the media are beginning to catch up to “The Monkey Cage,” as evidenced by this newspaper story.

The story quotes Daniel Geschwind, a professor of neurology and psychiatry at UCLA, to the effect that left-handers’ tendency toward bilateral brain function could enable them to visualize problems more broadly and with more complexity and could relate to the social and interactive skills needed to be successful in politics.

That’s pretty speculative, but the story goes on to note one well-established difference between left-handers and right-handers. According to Amar Klar, a scientist at the National Cancer Institute, “Handedness is related to the way the hair spins on the back of your head.” The whorl for right-handers curls clockwise in 92% of cases. In left-handers, the distribution is random, with half exhibiting a clockwise whorl and the other half spinning counterclockwise. (Sort of like toilets flushing counterclockwise in Australia, I guess.) The relevance of the whorl phenomenon to presidential politics is not immediately apparent, but perhaps one of the new generation of political scientists studying the physiological bases of political behavior can forge the link.

P.S. I know of no evidence that Bob Barr is left-handed, but it really doesn’t matter.

UPDATE: In response to popular demand (see the comments below), I launched an extensive research effort to find an answer to the vital question of “Is Ralph Nader left-handed?” Here’s what I found in my landmark three-second Google search:

[Hat tip to Erik Voeten]

Posted by Lee Sigelman at 10:03 PM | Permalink | Comments (2)

Political scientists frequently moan about the reluctance of political journalists to read their work and use it when making horse-race prognostications. So it’s nice to see Tom Edsall (previously at the Washington Post, now at HuffPo) do a long piece on the implications of recent work on political cycles for the upcoming election. As a non-Americanist, I’m not very well grounded in this literature (my broad impression is that it’s a bit skimpy on the causal mechanisms), but even so.

Posted by Henry at 11:30 AM | Permalink | Comments (0)

Boris forwarded to me this passage from The Audacity of Hope which was noted by Jim Geraghty:

Increasingly, I [Obama] found myself spending time with people of means - law firm partners and investment bankers, hedge fund managers and venture capitalists. As a rule, they were smart,interesting people, knowledgeable about public policy, liberal in their politics, expecting nothing more than a hearing of their opinions in exchange for checks. But they reflected, almost uniformly, the perspectives of their class; the top 1 percent or so of the income scale that can afford to write a $2,000 check to a political candidate. They believed in the free market and an educational meritocracy; they found it hard to imagine that there might be any social ill that could not be cured with a high SAT score. They had no patience with protectionism, found unions troublesome, and were not particularly sympathetic to those whose lives were upended by movements of global capital. Most were adamantly prochoice and were vaguely suspicious of deep religious sentiment… I know that as a consequence of my fund-raising I became more like the wealthy donors I met, in the very particular sense that I spent more and more of my time above the fray, outside the world of immediate hunger, disappointment, fear, irrationality, and frequent hardship of the other 99 percent of the population - that is, the people I’d entered public life to serve.

Continue reading "Obama on rich liberal political donors" »

Posted by Andrew Gelman at 12:54 AM | Permalink | Comments (0)

Eric Rauchway has an interesting post on the reasons why the Democrats lost the South.

ever since Nixon’s “southern strategy”—it’s been commonplace to assume that the Republicans picked up where the Democrats left off in courting bigoted whites, in the South and elsewhere. … Along come some political scientists to tell us this Republican racism is a bit of a side show, that the real story of the GOP’s new southern eminence has to do with the emergence, at long last, of a New South, ushered (ironically) into being by Democratic programs of New Deal and wartime mobilization. … One of the best-known works in this line was co-written by a friend and former colleague of mine, Byron E. Shafer, together with Richard Johnston, titled The End of Southern Exceptionalism. They argue that more, richer white southerners means more Republican white southerners.

Shafer and Johnston sometimes contribute to this confusion themselves, nowhere more than in the short section they devote to the influence of George Wallace on southern politics. Now, we think we know this story, too: Wallace helped loosen the loyalty of southern whites to the Democratic Party in 1964 and in 1968; the sort of person who voted for Wallace in 1968 was the sort of person who’d voted for Goldwater in 1964 and if he couldn’t have Wallace in 1968, he’d rather have had Nixon than Humphrey … But—say Shafer and Johnston—not so fast. … in the South the Democrats ultimately kept the Wallace voters, while the Republicans picked up the Johnson voters. It’s provocative, all right. Is it true?

Eric argues (using graphs of county level data and the aptly named ICPSR variable #666, George Wallace’s share of the presidential vote in 1968), that Shafer and Johnson may be missing part of the picture.

I don’t see what Shafer and Johnston see—I’d bet from looking at this that Nixon got the Wallace vote in ’72, when probably not so much had changed demographically since ’68. But in later elections, people have moved around a lot and just looking at race and income, the constituencies for the same counties look pretty different. Which isn’t to say Shafer and Johnston are wrong, per se—(1) I’m looking at counties, not districts; (2) I spent all of a couple of afternoons on this; (3) I may well be missing something incredibly obvious, such that it’s better to look at districts than counties, or something. Political scientists in the readership who want to look at this and tell me how I got it all wrong are welcome. It’s just to say, I don’t see what Shafer and Johnston see.

Any political scientists out there who’d like to comment?

Posted by Henry at 11:09 AM | Permalink | Comments (0)

The recent Washington Post poll I mentioned earlier raises some interesting questions about the role of racial attitudes in an Obama-McCain election. The Post’s description of the poll results suggests that for whites, racial attitudes have an independent impact on vote choice, even when controlling for party identification. John presented some recent findings suggesting that whites’ racial attitudes, at least in recent elections, aren’t particularly influential.

In order to shed a bit more light on this issue, I looked around for recent polling data that included questions on racial attitudes in addition to the usual horse race questions. Fortunately, the Roper Center has a Newsweek poll from April that includes some of these questions. In particular, it asks respondents the following:

If Barack Obama were to become president, do you think his administration’s policies would favor African-Americans and other minorities, would favor whites, or would NOT favor any group in particular?

This question gets at the most common and most critical stereotype facing black candidates, that they are defined by their race, that they are motivated to advance black interests over those of whites in a zero-sum racial spoils system, and that they are thus unable to represent the interests of whites. Such stereotypes have been common to white racist concerns about black political power since Reconstruction and are far more prevalent than stereotypes about black intellectual or moral inferiority. Overall, about 20 percent of respondents and 22 percent of whites thought that a President Obama would favor African-Americans. In contrast, only about 4 percent of African-Americans thought so.

Continue reading "More on Race and Voting" »

Posted by Phil Klinkner at 01:40 AM | Permalink | Comments (1)

Appropos of Phil’s post on this Washington Post piece, as well as this Newsweek piece cited by Brendan Nyhan, here are three studies — two published, one informal — that speak to this question.

First, see Zoltan Hajnal’s book on white attitudes toward black mayors. He concludes:

[U]nder most black mayors there is real, positive change in the white vote and in the racial attitudes of white residents. This change occurs because black incumbency provides concrete information that disproves the fears and expectations of many white residents. These findings not only highlight the importance of black representation; they also demonstrate the critical role that information can play in racial politics and point to the ability of at least some whites to change their minds about blacks and black leadership.

Second, as I’ve noted before, Ben Highton looked at a large sample of House races and found that white voters were no more or less likely to vote for black candidates. He writes:

To conclude, although African American “victories [in majority-white areas] attract attention precisely because of their exceptional nature” (Lublin, 1995, p. 112), one should not automatically assume that their rarity results from discrimination by white voters. Given the evidence presented here, the barrier presented by white voters in general elections does not appear especially daunting, especially in relation to the barrier it is often perceived to be.

His paper is here.

Finally, Seth Masket recently undertook a quick study of black candidates in Senate and gubernatorial races in 2002-2006. As he notes, the important caveat is that there have been few such candidates — 10 altogether, and 2 of those (Obama and Alan Keyes) actually ran against each other — and so that data are thin. Nevertheless, Seth also finds that race did not have a significant impact on the candidate’s vote share.

He then did a second analysis, focusing on black Democratic candidates. He computed the “expected” vote share of these candidates, based on several factors, such as the state’s partisan leaning and incumbency status. He then compared these candidates’ expected vote to their actual vote. He found, at best, a small discrepancy:

In each of the four contests, the African American Democrat ran behind the expected vote, although not by much…In only one of these cases (Ford) did the African American candidate lose where he “should” have won.

Incidentally, the story is much less consistent for the black Republican candidates. In two cases, they ran ahead of the expected vote, and in two they ran behind. None got within 10 points of winning, though.

Of course, there are ways in which this year’s presidential race may not be comparable to Senate, House, gubernatorial, and mayoral races. Perhaps presidential races are just different. Perhaps Obama is different. But these studies are worth taking into account going forward.

Posted by John Sides at 08:18 AM | Permalink | Comments (2)

Syracuse University political scientist Jeff Stonecash has been doing public opinion research for the last quarter-century, using university facilities and paying Syracuse students to serve as interviewers. He has conducted surveys for non-profit organizations and the local newspaper, among other sponsors, and for numerous candidates for public office — Republicans and Democrats alike. Now Stonecash is being told by Syracuse University authorities to cease and desist — this following complaints from a Democratic candidate about a survey that Stonecash recently conducted for his Republican opponent. The complainant alleges that Stonecash is a partisan Republican whose survey results are biased. (Actually, Stonecash is a registered Democrat.) When informed of these concerns, Stonecash indicated that he “would have done the same thing for [the Democratic candidate], but they never asked. I’m not a partisan pollster.”

This episode raises all sorts of questions about “applied” research in academic settings, especially in politically charged circumstances. For details, click here.

[Hat tip to Carol Sigelman]

Posted by Lee Sigelman at 04:32 PM | Permalink | Comments (0)

Tony Schwartz has died. He was the creator of the “Daisy” ad. The New York Times obituary is here. The Washington Post obituary is here. The Daisy ad is here. Schwartz’s own website is here.

Interesting tidbits from his life:

• Schwartz was agoraphobic and rarely left his home. Political clients traveled to him.

• Schwartz also made field recordings of folk music and ambient noise in New York City. See here.

• Schwartz said, “The best political commercials are Rorschach patterns. They do not tell the viewer anything. They surface his feelings and provide a context for him to express these feelings.” That statement echoes a scholarly literature demonstrating that political ads “prime” or make salient certain considerations in the viewer’s mind.

• Schwartz also said that the Daisy ad was “the most positive commercial ever made.” That, I suppose, is in the eye of the beholder.

Both the Times and Post obituaries speak of the Daisy ad’s effectiveness without citing any particular source or evidence — e.g., “was credited with contributing to Johnson’s landslide victory at the polls in November” in the Times. To my knowledge, there is no evidence that the Daisy ad — which aired but once, albeit with news coverage thereafter — had any effect on the 1964 election. In fact, I have open Jim Stimson’s book Tides of Consent: How Public Opinion Shapes American Politics. In his chapter on campaigns, he presents LBJ’s percent of the vote throughout 1964. The Daisy ad aired on September 7, but LBJ’s share of the vote did not change at all from essentially the beginning of August until just before Election Day.

It is difficult to show that individual ads affect candidate fortunes or election outcomes, and the conventional wisdom that certain ads mattered in particular elections is typically based on conjecture and lore. The Daisy ad likely constitutes such a case.

Addendum: Here is Stimson’s graph:

Posted by John Sides at 09:43 AM | Permalink | Comments (3)

Recent discussion about the “Reagan Democrats” and their role in the 2008 election begs the question of who these voters were. The standard story is that these voters were blue-collar whites, conservative on social issues, and disenchanted with the Democratic Party. Unfortunately, in many respects, this characterization is inaccurate.

The “New Republicans” were not drawn disproportionately from the middle to lower strata of the population; their conservatism was not more marked on social issues than on economic issues; they were neither more religiously oriented nor more alienated from government than other voters; finally, they bore little similarity to the constituency that provided the core support for Wallace in 1968.

This is from an analysis of the 1980 election by Jerome Himmelstein and James McRae, Jr., which was published in the Public Opinion Quarterly in 1984. “New Republicans” are simply those who said they voted for Carter in 1976 but Reagan in 1980. This may not be what everyone means by “Reagan Democrat,” but given the amorphousness of the term, it is a reasonable working definition.

Himmelstein and McRae find that these New Republicans supported Reagan for much more ordinary reasons: they disapproved of the performance of President Carter, especially with regard to the economy. This kind of behavior — retrospective voting based on the performance of the national economy — is well-documented and, well, a lot more pedestrian than prevailing theories either then and now.

A question raised by Ezra Klein in the post linked to above is: what happened to the Reagan Democrats? Did they permanently defect to the Republicans, or did they return to the fold in later elections? There is no good data on this, in part because we have few surveys that interview the same voters over multiple elections. But Himmelstein and McRae’s findings strongly suggest that Democratic defections to Reagan had more to do with current economic circumstances than some sort of deep-seated ideological discomfort with the Democratic Party. This likely explains why not very long after Reagan, Democratic loyalty to the Democratic presidential nominee was at or above its historical norm (see this previous post).

The Himmelstein and McRae paper is here (gated). If anyone can find a non-gated version, please put the link in the comments.

Posted by John Sides at 08:45 PM | Permalink | Comments (1)

This is how they do political science at Cambridge (U.K. version). They also seem to think that doing a poll is equivalent to “throwing darts at a board.” To give credit where due, though, they do take their darts pretty seriously over there . . .

Posted by David Park at 10:48 PM | Permalink | Comments (2)

Andy warns (ok, rants) about people overinterpreting early presidential polls. This time the culprit is an astrophysicist rather than a reporter. . . .

Posted by David Park at 03:21 PM | Permalink | Comments (0)

Awesome dynamic graphics here. See Andy for comments on how it could be even better but it’s pretty impressive as is.

Posted by David Park at 11:21 PM | Permalink | Comments (1)

Andy writes,

One thing I learned in econ class in 11th grade was that government policy should be counter-cyclical (spending more in recessions and cutting back in boom times), but that there’s a lot of pressure to be pro-cyclical, which will tend to exacerbate business cycles. (Except I suppose they didn’t say “exacerbate” in 11th grade.) At a personal level, too, it’s natural to spend more when we have more and cut back when we aren’t doing so well. Every now and then you hear about a “rainy day fund” but my general impression is that these are never big enough to counter the business cycle.

Political parties seem to apply a similar pro-cyclical behavior in their congressional election campaigns. Consider 2008. It’s expected to be a good year for the Democrats, and so now should be the time for them to make some investments in new, young candidates. They should encourage lots of their incumbents to retire, because in 2008, they can win a lot of these districts without needing the incumbency advantage (estimated to be about 10% of the vote, i.e., enough to take you from 50% to 60%). Conversely, this is the time for the Republican Party to hold on to what it has, and to keep all their incumbents in, trying to hold out until 2010 when the pendulum might swing back in their favor. But we don’t see that—actually, something like 30 Republican House members are retiring this year. Republicans retiring, Democrats sticking around—that’s a recipe for big Democratic gains this year. But then in 2010, or 2014, or whatever year it is when the Democrats get wiped out—then a bunch of their incumbents will probably retire, and boy will the Democrats wished they had put in younger incumbents back in 2008 when they had a chance!

One of the difficulties here is that I’m talking about the long-term goals of the parties, but “the parties” are, to a large extent, simply their officeholders. And congressmembers’ incentives can be much different from those of the party as a whole. In particular, it makes sense that an incumbent congressmember will want to quit in a year when he or she would be facing a tough reelection battle, and when the prize for winning is to remain in the minority. Conversely, why step down when you’re facing an easy reelection and the prospect of some juicy committee assignments? So the individual officeholders have an incentive for pro-cyclical behavior, even if it harms their party’s long-term interest.

Beyond the benefits or lack thereof to the individual parties, pro-cyclical behavior would seem to increase the size of political changes, making the swings in congressional representation larger than would be expected simply based on swings in public opinion. Actually, many political scientists would consider this a good thing (an increased “swing ratio”); my point here is that some of this swing is “endogenous” in the sense of arising from pro-cyclical decisions of individual congressmembers deciding whether to run for reelection. It would be interesting to see if this happens with state legislatures as well.

We also see this in the Senate. For example, 84-year-old Frank Lautenberg is running for reelection in New Jersey. This is a Democratic year when the Democrats might do well with just about anybody. (Or maybe not; I don’t really follow New Jersey politics and am just extrapolating from national polls.) In 6 years, they’re going to need to find someone new, and at that point they might wish they had an incumbent already in the slot.

Posted by David Park at 11:12 AM | Permalink | Comments (4)

Here’s what might happen to the House in 2008:

It’s from this paper by Kastellec, Gelman, and Chandler. See Andrew’s blog for a quick summary.

Posted by David Park at 09:05 PM | Permalink | Comments (0)

The Sunday Washington Post ran excerpts from five interviews conducted by the Presidential Oral History Program at the University of Virginia’s Miller Center of Public Affairs. The topic was how presidential nominees choose their running mates, and the interviewees were all individuals who had been intimately involved in the process in past campaigns: Richard V. Allen, on tutoring Sprio Agnew on international affairs in 1968; Martin Anderson, on the wheeling and dealing between Gerald Ford and Ronald Reagan in 1976; Richard Moe on Walter Mondale’s decision to join Jimmy Carter’s ticket in 1976; Lyn Nofziger on Reagan’s choice of George H.W. Bush in 1980; and Stuart Spencer on Bush the Elder’s selection of Dan Quayle in 1988.

The excerpts are fascinating.

Here, to get you started, is the excerpt from Richard V. Allen:

I had the joy of providing [Agnew] with his first foreign policy briefing, which was a hoot … We sat in beach chairs out on the beach, with a map of the world on the ground and four stones holding the map down, and I with a pointer and Agnew in the chair wearing shorts, very casual.

Before I started the briefing, he said, ‘You know, I want to tell you something Dick. I’ve never been out of the country before, except to go to Greece, and I came straight back.’

And I said, ‘Well, that’s going to complicate things a little bit, so I’m going to take you through the world, a tour d’horizon, and I’m going to tell you what our policy is in each area. … I got down to South Africa, I had my pointer, I said, ‘Okay, now we come down here.’ He said, ‘Don’t tell me. I think I know this one. … That’s a black government, right?’ Well, it surely wasn’t a black government in 1968. And so we had to walk back from that one.

To read the four other excerpts, click here, gated. Or if you want to proceed directly to the full texts of the interviews for the Carter and Reagan presidencies (which cover much more ground than vice presidential selection and involve many more interviewees than those named above), click here. Those who can stomach watching political sausage being made will find some wonderful morsels here.

Posted by Lee Sigelman at 07:53 PM | Permalink | Comments (1)

John Heilemann of the New York Magazine, makes an interesting case for Mayor Bloomberg for Veep - for either McCain or Obama. Heilemann notes that Bloomberg does nothing in the electoral math deparment for either candidate. NY is going Obama. However, if you’re willing to “stretch your mind a bit” and believe that the economy will be the central issue in the election, then Bloomberg has something to offer both candidates - his entrepreneurial background and “financial stewardship” of Gotham. Specifically, he could help McCain in FL, NJ and PA, and Obama in FL. There’s a host of reasons why McCain and Obama shouldn’t pick Bloomberg, and Heilemann goes through as well, but it is an interesting idea.

PS Of course the real reason for pushing Bloomberg as VEEP is that New Yorkers feel like they should have someone on the national stage since, for a fleeting moment, they thought they were going to have an all NY contest for the presidency - Giuliani vs Clinton.

Posted by David Park at 07:15 AM | Permalink | Comments (0)

In his influential Atlantic magazine article, “One Nation, Slightly Divisible.” published after the 2000 election, David Brooks compared Montgomery County, Maryland, the liberal, upper-middle-class suburb where he lives, to rural, conservative Franklin County, Pennsylvania, a short drive away but distant in attitudes and values, with “no Starbucks, no Pottery Barn, no Borders or Barnes & Noble,” plenty of churches but not so many Thai restaurants, “a lot fewer sun-dried-tomato concoctions on restaurant menus and a lot more meatloaf platters.”

In Brooks’s home state of Maryland, there is no clear pattern of county income and Republican vote, and it was not difficult for him to go from Montgomery County, the prototypical wealthy slice of blue America, to a poorer, more Republican-supporting county nearby. Here are the data from 2000:

Brooks lives in a liberal, well-off part of the country. It is characteristic of the East and West Coasts that the richer areas tend to be more liberal, but in other parts of the country, notably the South, richer areas tend to be more conservative. A comparable journey in Texas would go from Collin County, a wealthy suburb of Dallas where George W. Bush received 71% of the vote, to rural Zavala County in the southwest, where Bush received only 25% of the vote:

When we showed our graph of Texas counties to another political scientist, he asked about the state capital, noted for its liberal attitudes, vibrant alternative rock scene, and the University of Texas: “What about Austin? It must be rich and liberal.” We looked it up. Austin is in Travis County and makes up almost all of its population. Travis County has a median household income of $45,000 and gave George W. Bush 53% of the vote in 2000, about midway between Collin and Zavala counties. (Austin has its own red-blue divide, with a highly Democratic university area and urban center, and strongly Republican suburbs.)

This is not to dismiss Brooks’s insights but rather to place them in the national context of income mattering more in poor states than in rich states, which is the subject of our forthcoming Red State, Blue State book (from which the above graphs were taken).

Posted by David Park at 12:19 PM | Permalink | Comments (7)

Apparently, someone at the Los Angeles Times reads The Monkey Cage, because they asked me to write a short piece apropos of these two posts about the “divided” Democrats and the reinforcement effect in campaigns.

The op-ed is here. The gist of the argument is:

Despite ugly battles and policy differences that sometimes seem intractable, the reality is that presidential campaigns tend to unify each party behind its nominee. Political scientists call this phenomenon the “reinforcement effect.”

Some references for works and data cited in the piece:

• The Lazarsfeld, Berelson, and Gaudet study.

• The National Annenberg Election Study.

• The Pew center poll.

• The piece by Ben Highton. Thanks to Ben for doing some additional data analysis for me.

• The National Election Study data showing that 90% of the public identifies with or leans toward one of the major parties.

• Gary Jacobson’s study of public opinion during the Bush administration.

Posted by John Sides at 09:01 AM | Permalink | Comments (3)

That’s what Jodi Kantor says in the New York Times:

But in recent presidential elections, Jews have drifted somewhat to the right.

Really? I took the state-level exit polls for each presidential election from 1988-2004 and combined them into one big file for each election year. This produces relatively large samples of Jewish voters (~1,000 each year). Here is the percentage of Jewish voters choosing the Democratic candidate:

Looks like a leftward drift. In 1988, 71% of Jewish voters supported Dukakis. In 2004, 79% supported Kerry. And you don’t even have to crunch the data. I typed “jewish vote” into Google and the very first link told me the same thing.¹

Kantor is not wrong about her central thesis, although she fails to cite any systematic data to prove it: Obama’s current support among Jewish Americans is lower than that of these past Democratic candidates. Since Kantor doesn’t present any polling data about an Obama-McCain match-up, I went and found some. Just like before, it was hard: I typed “poll jew obama” into Google. Here was the first hit, a poll cited in the May 9th edition of the Jerusalem Post:

A new Gallup survey found that 61% of Jewish voters prefer Obama to McCain, who got 32% of the Jewish support.

That’s how you write a story about what voters think. Don’t make vague claims about recent history, such as the “rightward shift,” without backing it up. And then, if you really want to make the case, don’t only talk to a handful of people in Boynton Beach and Boca Raton. Find some hard evidence. Google will even do it for you.

ADDENDUM: See Andrew Gelman’s post with data from the National Election Study.

[1] This webpage’s data doesn’t precisely match mine, perhaps because they are using the (smaller) national exit poll file rather than the merged state-level files that I use. In fact, they find that the percent of Jews supporting Dukakis in 1988 is even lower (64%); the difference may arise because there wasn’t a separate exit poll in every state that year and so my combined state file is missing some data. In any case, the “rightward shift” is not plausible. In fact, if anything, the leftward shift is larger, especially if you compare 1988-2004 to earlier elections in the 1970 and 1980s.

Posted by John Sides at 09:19 PM | Permalink | Comments (12)

Being of the early-to-bed, early-to-rise persuasion, I don’t stay up late enough to watch “Saturday Night Live.” Thus, over at Daniel Drezner’s blog, I happened upon what for me, but not for millions of other Americans, was brand-new material: a biting lampoon of Hillary Clinton explaining why she’s staying in the race and will be a stronger candidate than Barack Obama. Clinton supporters will probably hate this, but I assume that in short order SNL will come up with something equally biting about Barack Obama, and when that happens I’ll try to feature it here, too. Ditto for John McCain.

[Hat tip to Daniel Drezner]

Posted by Lee Sigelman at 06:13 PM | Permalink | Comments (2)

A journalist queried me about whether anyone had studied efforts at voter suppression, such as “the distribution of flyers or use of robocalls in black neighborhoods with false information about election dates, polling locations,
eligibility requirements, and so on.”

Readers, is there any credible study that has actually sought to measure the effects of such tactics? Please leave any tips in the comments.

Posted by John Sides at 08:44 PM | Permalink | Comments (5)

Since I was recently complaining about the lack of political science in journalism, let me also give credit where credit is due. This piece, by John Judis in The New Republic, does a nice job surveying relevant literature about race from political science and psychology, including work by Tali Mendelberg, David Sears, Donald Kinder, Nicholas Valentino, Vincent Hutchings, and Ismail White. Judis’ piece is distinctive for treating social science as if it has some nuance — in particular, discussing existing debates rather than treating the debates as settled. There are, of course, further nuances and debates that he does not discuss, but, given the limits of magazine journalism, he still deserves kudos.

[Hat tip to Clyde Wilcox]

Posted by John Sides at 04:29 PM | Permalink | Comments (1)

In the spirit of John Sides’s selfless offer of himself and David Park as regression-runners and findings-interpreters for New York Times columnist David Brooks (HERE), I hereby volunteer John and David’s services to ABC News political analyst George Stephanopoulos as well. (I myself am much too busy blogging about cats to become involved in such an undertaking.)

This offer is prompted by the following exchange between anchorman Charles Gibson and Stephanopoulos on May 6, 2008, the evening of the Indiana and North Carolina primaries:

GIBSON: And joining me now is our chief Washington correspondent, George Stephanopoulos, who is here in New York tonight. …

STEPHANOPOULOS: … Let’s look at those numbers. … We did ask a question I know in the exit polls about Reverend Wright, Barack Obama’s former pastor and whether that was influencing voters. What did we find? Right down the middle. About half said it’s important, about half said it was unimportant. Of those who said it was important, look at this in Indiana, 70% went for Senator Clinton. Of those who thought it was unimportant, again right down the middle, 65% for Barack Obama. So what you thought about the importance of Reverend Wright basically determined your vote.

Now, George Stephanopoulos, who graduated summa cum laude from an Ivy League school (Columbia), is no dummy. The concluding sentence of his summary of the survey results is so fundamentally flawed, though, that one can only wonder what they were teaching in all those political science courses that he took (he was a political science major) or whether the time he spent in the Clinton White House so addled him that it’s all gotten jumbled in his head.

The most basic lesson one is supposed to learn in one’s very first research methods course is — repeat after me, everybody — “Correlation is not the same as causation.” Now, with that thought in mind, go back and re-read the offending sentence: “So what you thought about the importance of Reverend Wright basically determined your vote.” Do you begin to see the problem?

The causal spin that Stephanopoulos put on these survey results assumes, in effect, that heading into the Indiana and North Carolina primaries most voters weren’t committed to one candidate or the other. Then the Reverend Wright issue emerged, and — presto! — it split the Democratic electorate into two groups, Clinton voters and Obama voters.

Instead, let’s begin with a decidedly different, and altogether more plausible assumption. Let’s say that heading into these two primaries, which after all occurred very late in the primary season, most voters were already pretty well committed to one or the other of the two candidates. Then the Reverend Wright issue emerged. For many of those who were already intending to vote for Clinton, that issue reinforced their existing inclination; certainly it wouldn’t have transformed them from Clinton supporters into Obama supporters. For many (presumably most) of those who were already intending to vote for Obama, though, the Wright issue wasn’t sufficient to persuade them of the error of their ways and move them into the Clinton column.

Then on election day along came an exit poll interviewer asking them how important the issue had been to them. Naturally, many of the Clintonites cited it as a reason for voting for Clinton rather than Obama — because it had been consistent with that intention and also because it had also emerged late enough to still be retrievable from memory. But how many Obama supporters would cite it as a reason for having voted for Obama? As the issue played out, that simply wouldn’t make sense; it was a one-way or “valence” issue, and its valence was anti-Obama. So Clinton supporters were more likely to cite the Wright issue as underlying their vote choice than were Obama supporters — the same statistical finding that Stephanopoulos reported but with a very different interpretation.

The problem, stated more generally, is that when pollsters ask voters which issues were most important to them, what the voters tell them can’t be taken at anything approaching face value. People just aren’t reliable reporters of their own mental processes. (The classic statement of this point is Nisbett and Wilson’s “Telling More Than We Can Know: Verbal Reports on Mental Processes,” Psychological Review 84 (May 1977), pp. 231-259.) In the context under consideration here, this unreliability typically takes the form of ex post facto rationalizations: Those who were predisposed to vote for Candidate X cite the issues that X had raised as the decisive ones, and those who were predisposed to vote for Candidate Y cite Y’s issues as decisive. Look back over several decades of ANES post-election surveys and you’ll see this pattern time and time again. Or save yourself the trouble and hunt down a copy of Wendy Rahn, Jon Krosnick, and Marijke Breuning’s “Rationalization and Derivation Processes in Survey Studies of Political Candidate Evaluation” (American Journal of Political Science 38 (August 1994), pp. 582-600), the bottom line of which is that “Voters’ reports of the reasons for their preferences were principally rationalizations.”

My point isn’t that the Reverend Wright matter was wholly inconsequential. It may well have figured into the decision calculus of some late-deciders in Indiana and North Carolina. But that’s a world away from concluding based on the data that Stephanopoulos had in hand that “What you thought about the importance of Reverend Wright basically determined your vote.”

Columbia has an excellent political science department. Perhaps as a public service they should bring back their erstwhile star student for some refresher courses in research methods and electoral behavior. Or maybe Stephanopoulos will decide to take me up on my sincere offer of John and David’s assistance.

Posted by Lee Sigelman at 07:22 PM | Permalink | Comments (9)

Below is the percentage of Democrats voting for the Democratic nominee for President, drawing on National Election Studies data from 1952-2004.

This graph should be the starting point for any discussion of a “deeply divided” or “fractured” Democratic Party. The party loyalty of Democrats has been increasing over time and has essentially hovered at 90% since 1992. (And Republicans are similarly loyal to the Republican nominee.)

A central finding of political science research into campaign effects is this: during the general election, presidential campaigns tend to reinforce one’s preexisting partisan leaning. In other words, they tend to unify the party.

The tendency now, as in the NY Times and Washington Post pieces linked above, is to ponder, ominously, “divides” within the Democratic Party and, elsewhere, to quote meaningless exit poll data about how many Clinton voters say they will stay home or vote for McCain. Instead, journalists and commentators should be grappling with the reinforcement effect and party loyalty.

Early data from this election also portends a high degree of party loyalty. See this, from an April Pew poll. Already, large percentages of Democrats (77%) say they will vote for Obama against McCain. (Party loyalty among Democrats in the unlikely Clinton-McCain race is 81% — which, given sampling error, is indistinguishable from 77%.) Republicans are, right now, more unified (85% say they will vote for McCain), but this isn’t surprising, since the Republican race has been decided for much longer.¹

Thus, it’s not clear why Pew entitles the table “Obama’s Struggle within the Democratic Base.” Obama does just as well as Clinton among Democrats. Moreover, and most importantly, he will likely do better as soon as Clinton drops out and the general election campaign begins its inevitable reinforcement effect.

*****

Addendum: Below is a picture of what reinforcement looks like. The graph derives from National Anneberg Election Study data. It presents the percent of Democrats who said they would vote for Gore over Bush. The data derive from samples taken daily from December 14, 1999, until November 6, 2000. Early on, about 78% of Democrats said they would vote for Gore. That increased to 87% over the course of the campaign.

[1] Independents, by the way, favor Obama over McCain, 52-41. That difference should be close to statistical significance.

Posted by John Sides at 09:05 AM | Permalink | Comments (2)

“Not really” is not what you’d expect to hear, if you are reading about:

• the “problem” that Wright poses (according to this New York Times piece)
• the “threat to Obama’s ability to show that he could unify the Democratic Party” (again, the New York Times)
• the “damage to his presidential candidacy” (the Washington Post)
• the “troubling trends” (this post at the Washington Post’s Behind the Numbers blog)
• the “diminished aura of inevitability” inferred from this CBS/New York Times poll

Let’s leave aside the easiest rebuttals — namely that Behind the Numbers is cherry-picking recent polls and relying on the always-dubious “does X make you more or less favorable” question (they, of all people, should know better), and that the CBS/NYT poll’s real main finding is that Obama’s lead over Hillary has actually grown since the last CBS/NYT poll.

Instead, let’s take a more systematic look at the available data. First is this bravura graph from Seth Masket over at Enik Rising, comparing Obama’s performance in the OH and PA primaries among various demographic groups:

Even among those groups who should perhaps be most bitter over being called “bitter” — weekly church-attenders, union members, white Catholics — Obama loses only a few percentage points in PA relative to OH. Among every other group, he does better.

More importantly, look at the two graphs below: the national Obama-Clinton split and the Obama-McCain trial heat (both courtesy of pollster.com).

Obama’s lead over Clinton has actually gotten slightly larger in the last several weeks. Obama has also improved viz. McCain.

I assume a few other voices are shouting in the wilderness on this issue (Seth writes that Bill Schneider of CNN noted the lack of movement in the polls). But clearly the Gang of 500 hasn’t gotten the message. Perhaps we’re on the cusp of a real drop in the polls for Obama; if so, all the above quotes from our leading journalistic lights would be justified. But until then, the “problem” is in their eyes only.

Posted by John Sides at 08:56 PM | Permalink | Comments (5)

The other day I noted the contributions that the Brennan Center at NYU and the Political Advertising Project at the University of Wisconsin have made to our understanding of how often campaign ads are shown, at what cost, to how many people, and with what effects (click HERE). Now, though, comes word that one component of these data — the cost estimates — may be less reliable than had previously been supposed.

Let’s begin with a brief review of where these cost estimates come from in the first place. (The source of the following is Michael G. Hagen and Robin Kolodny’s article, “Finding the Cost of Campaign Advertising,” in the latest issue of The Forum, published by the Berkeley Electronic Press. For an abstract and directions about downloading the article itself, click HERE.)

For the last decade, the Campaign Media Analysis Group (CMAG), which is a division of TNS Media Intelligence, has been monitoring political, public affairs, and issue advocacy advertising on seven TV broadcast networks, 44 national cable networks, two Spanish-language networks, more than 200 syndicated programs, and broadcast stations in the 100 largest TV markets in the country. CMAG catalogues an array of information about each airing (date, time, length, sponsor, and so on). It also estimates, for each airing of an ad, the cost of the airtime. As Hagen and Kolodny outline the cost estimation process:

The estimates are based initially on a monthly survey of television station executives, advertising agents, and time buyers, who report the average price in their market of a thirty-second spot to be aired during the month ahead at a particular time of day. CMAG uses this “daypart” information about advertising rates in general to approximate the cost of particuilar spots on particular stations purchased by particular political sponsors. The process for constructing these specific estimates is proprietary, but it is grounded in ‘specialized research on advertising trends for political and issue advocacy advertising,” and includes adjustments for the fact that “rates for political advertising often can deviate from the standard rates used for non-political advertising.”

That’s obviously an extremely comprehensive data gathering process. Even so, CMAG misses some expenditures because it monitors “only” the major stations in the major markets and because it doesn’t monitor local cable networks. Moreover, campaigns often have to pay premium rates to secure particular non-preemptible time slots, and TV stations often raise their prices as the end of the campaign nears. These considerations imply that, if anything, the CMAG expenditure figures should underestimate the total that is spent on campaign ads — a claim that is routinely made when these expenditure figures are presented.

This is where things get interesting.

Continue reading "Estimating the Cost of Campaign Advertising: A Cautionary Note" »

Posted by Lee Sigelman at 07:41 AM | Permalink | Comments (0)

Televised campaign ads live two lives. First, there are the paid-for showings of the ads themselves. An ad may get shown just once in a single media market or dozens of times in hundreds of markets. In recent years, thanks to data made available by the Brennan Center at NYU and the Political Advertising Project at the University of Wisconsin, we’ve learned a great deal about how often these ads are shown, at what cost (though I’ll have more on that in an upcoming post), to how many people, and with what effects.

Second and, increasingly, there is free media coverage of the ads. Campaign ads, in a weird sort of “pseudo-events create new political reality” way, have transmogrified into one of the major media stories of modern gubernatorial, congressional, and presidential campaigns. To some extent, such media coverage may blunt the force of a particular ad by exposing misstatements or half-truths therein. Offsetting this blunting effect or perhaps, in many instances, overwhelming it, though, are the time-honored ideas that “There is no such thing as bad publicity” and “If you tell the Big Lie often enough, people will start to believe it.” That is, the charges and countercharges in a campaign ad may, thanks to the attendant publicity, take on a life of their own. And the ultimate test for a candidate may not prove to be whether those charges and countercharges are “right” or “wrong” (and few these days seem to be all of one or all of the other), but rather how the candidates react to them. If they do so slowly and haltingly (e.g., Kerry in 2004) or virtually not at all (e.g. Dukakis in 1988), then the attacks, having been aired by the parties and their allies and then given further prominence by media coverage, may sound a death knell for their candidacy. Letting oneself get put on the defensive seems increasingly to be taken as bona fide evidence that one is unqualified to serve in high political office.

In a forthcoming article in the Political Research Quarterly (“Free Advertising: How the Media Amplify Campaign Messages”), Travis Ridout and Glen R. Smith of Washington State University turn the spotlight on the second life of campaign ads by intensively exploring media coverage of commercials in ten of the 2004 Senate campaigns. Ridout and Smith report two basic results:

• “Ad amplification” via media coverage of paid commercials is indeed widespread. I think it’s fair to say that we already “knew” this, but having our impressions and intuitions validated via careful analysis is valuable nonetheless.

• The most media-amplified ads are the comparative and negative ones. Positive ads capture less media attention.

It’s the second result, of course, that’s really interesting. Media definitions of what is “news” encompass dimensions like conflict, drama, corruption, and malfeasance. By these and related standards, an ad proclaiming that Senator X has been voting against the public interest because of financial contributions from malevolent lobbying groups — well, that’s likely to bear looking into as a good news story. Meanwhile, an ad proclaiming that Governor Y has done an effective job of introducing some new programs and administering some existing ones — that’s no story at all. (Until, that is, Governor Y’s opponent calls those claims into question. Now, that’s news. And so it goes.)

Even if media coverage of a particular ad is critical, identifying distortions or counterpoints, it keeps the ad alive by exposing those who missed it when it aired on TV to what the candidates are saying about one another – adding fuel to the fire sparked by the ad itself and, given the media’s disproportionate focus on negative and comparative ads, thereby adding to the viciousness of this particular vicious cycle.

Posted by Lee Sigelman at 04:45 PM | Permalink | Comments (1)

Brian Knight and Nathan Shiff have an interesting new NBER paper on momentum and voter choice.

This paper provides an investigation of the role of momentum and social learning in sequential voting systems. In the econometric model, voters are uncertain over candidate quality, and voters in late states attempt to infer the information held by those in early states from voting returns. Candidates experience momentum effects when their performance in early states exceeds expectations. The empirical application focuses on the responses of daily polling data to the release of voting returns in the 2004 presidential primary. We find that Kerry benefited from surprising wins in early states and took votes away from Dean, who held a strong lead prior to the beginning of the primary season. The voting weights implied by the estimated model demonstrate that early voters have up to 20 times the influence of late voters in the selection of candidates, demonstrating a significant departure from the ideal of “one person, one vote.” We then address several alternative, non-learning explanations for our results. Finally, we run simulations under different electoral structures and find that a simultaneous election would have been more competitive due to the absence of herding and that alternative sequential structures would have yielded different outcomes.

This brings the US primary system, which I’d never really thought about in this way before, under the rubric of things that I find deeply weird as a foreigner about the US political system. I won’t pretend to comment on the modeling strategies or econometrics in the Knight and Shiff piece - but if the estimate that they provide (or even the estimate under an alternative specification that Iowa voters were six times as influential as Super Tuesday voters) is at all ballpark, isn’t there something problematic about a system in which the disparities in influence are as gross as this? And yes: the same point does apply to Senate voting too - but that’s impossible to change without changing the constitution, while a simultaneous election would only require calling Iowa’s and New Hampshire’s bluff (this is one situation where the one-shot game prediction that all states would converge on the earliest possible date would be superior to the current arrangement).

Posted by Henry at 12:10 PM | Permalink | Comments (0)