October 2008
Sun Mon Tue Wed Thu Fri Sat
      1 2 3 4
5 6 7 8 9 10 11
12 13 14 15 16 17 18
19 20 21 22 23 24 25
26 27 28 29 30 31  


PPBW Sponsors


Recent Entries

Recent Comments


« September 2005 | Main | November 2005 »

27 October 2005

Sex and Corruption

In an undergraduate seminar last year my students and I discovered a curious relationship between the representation of women in the national legislature and corruption. Specifically, the percentage of parliamentary seats held by men and the perceived level of political corruption are strongly correlated, at least in OECD nations.


I can imagine several explanations for this relationship. Causation could run either direction, and lots of spurious relationships come to mind. Is there any literature to consult on this?

Posted by Barry Burden at 7:31 PM | Comments (2)

25 October 2005

Republican Hearts and Democratic Minds?

When asked which party they feel they are, women are more likely to identify as Republicans than when they are asked which party they think they are. This is an effect that is virtually nonexistent among male respondents. The effect is largest among high information women. These are the provocative findings that Barry Burden presented from his paper entitled The Social Roots of the Partisan Gender Gap at the American politics research workshop on Monday. Here is the abstract:

I suggest that the gender gap in party identification is partly an artifact of question wording and asymmetric stereotypes about men and women’s partisan preferences. The Michigan model holds that party identification is fundamentally affective, yet surveys have always asked respondents to “think? when answering. The literatures on gender, identity, and stereotypes suggest that this slippage could be quite consequential. A new survey experiment reanalyzes the gender gap by comparing the standard partisan battery to an alternative version that emphases feelings rather than thoughts. Bringing question wording into closer alignment with theory causes the gender gap essentially to disappear. This happens because the “feel? questions find women to be less Democratic than did the “think? questions. Moreover, the disappearance of the gender gap occurs mostly among highly sophisticated women not those usually susceptible to question wording effects. Contrary to popular wisdom, men and women appear to be more, not less, alike politically when feelings are primed. Taken together, the findings raise new questions about why the gender gap emerged at all.

In a randomized experiment, one group of respondents is asked: “Generally speaking, do you usually feel that you are a Republican, a Democrat, an Independent, or what?? The other group is asked the standard NES question wording: “Generally speaking, do you usually think that you are a Republican, a Democrat, an Independent, or what?? Comparing the two, we find that the gender gap shrinks dramatically among the feel group (see Table 2 in the paper).

Aside from being empirically very interesting, these findings have important ramifications for survey research. As Barry points out: "Though the original conceptualization [of party identification by the Michigan school} stressed affect, surveys questions focus respondents on their thoughts" (p. 3).

Workshop participants brought up some interesting points:

  • The survey was conducted shortly after 9/11 ("late 2001", p.19 ). Even though it is a randomized experiment, the effect of the question wording could be conditional on both gender and 9/11. The psychology literature that Barry references could suggest that the the question framing activated post-9/11 fears of terror and approval of the President but only among women -- and this is what led to the bump in Republican identifiers.
  • The partisan gap exists in terms of party ID as well as voting. What are the implications of the findings if the effect can't be achieved in an NES survey or in the polling both? Barry brought up the interesting possibility research that gender differences might be observed in a vote-by-mail election.

This is a very interesting piece of research, and I look forward to see where Barry goes with it (maybe he will make a post telling us).

Posted by Andrew Reeves at 7:58 PM | Comments (1)

24 October 2005

Frisby on Term Limits

Do term limits alter the decisions made by challengers? This week in PPBW we hear about Tammy Frisby's paper on term limits and strategic candidate entry in state legislative elections. Here's the abstract:

In comparison to our considerable knowledge of strategic challenger entry behavior in U.S. House and Senate races, we know far less about how major party candidates make entry decisions in U.S. state legislative races. I use the last eligible reelection bids created by term limits on state legislators to test for strategic entry behavior in state House elections when prospective challengers have information about impending incumbent retirement. Analysis of races in four term limit states; California, Ohio, Florida, and Colorado; from 1996 to 2004 indicates that once long-time incumbents are turned out of office, prospective challengers do not adhere to our traditional expectations of strategic behavior. Instead of waiting to run for the open seat, prospective challengers are just as likely to run against last term legislators as other incumbents. I also fail to find evidence that would indicate last termers face weaker challengers; when they are contested, last termers do not enjoy larger margins of victory. The paperconcludes by considering how the lack of a last term effect across all districts could be hiding strategic entry behavior on the part of candidates and parties. While handled fully in the subsequent portion of the larger project, here I preview of a theory of strategic challengers and public information about incumbent retirement in state House races.

Posted by Barry Burden at 10:03 AM

22 October 2005

Abramowitz Paper

Yesterday, the Political Psychology and Behavior Workshop returned to the topic of polarization in the electorate with Alan Abramowitz’s paper, “Is Polarization a Myth?? This paper, unlike the one presented by Ansolabehere, Rodden, and Snyder, finds that the American public is sharply divided in its opinions on a wide range of issues. Moreover, these divisions map more closely onto differences in religious attendance and fundamentalism. Abramowitz challenged the notion of the “Culture War? as a myth, finding that voters in red states and blue states really do differ in important ways.

The audience reacted favorably to the paper. The biggest critique of the paper was theoretical; the audience wanted more of a discussion of why polarization is an important concept for political scientists to study. What are the normative implications of polarization? Moreover, what are the implications for our theories of political behavior? Other comments emphasized the need to be clearer about the specifics of the analysis (including sample sizes), the potential benefits of using factor analysis for the ideology measure, and the need for multivariate analyses in order to provide stronger support for the hypotheses. After some discussion of the role of religious beliefs and practices in political behavior, the presentation ended with speculations on why Abramowitz and Ansolabehere, et al. come to such different conclusions.

Posted by Traci Burch at 12:10 PM | Comments (1)

20 October 2005

The Iraqi Election: Democratic Voting or Fraud?

The early results of the Iraqi election have been called into question, as some areas have reported extraordinarily high turnout. Among the concerns regarding the legitimacy of the outcome is the almost universal approval of the constitution in some areas, specifically those provinces that are inhabited predominantly by the Shiite and Kurdish people. One explanation for the abnormally high approval rate, offered by the planning minister Mr. Salih, is that the Iraqi people voted based on instructions given to them by their political and religious leaders. Americans have criticized the voters involved for failing to cast “individual? votes, and the leaders for circumventing the new democratic process. Yet the practice of a leader instructing members of his organization how to vote is not uncommon to American democracy. Following the 2004 general election, allegations of fraud were brought against a church in the state of Oregon, which holds all mail ballot elections. It is claimed that some 1100 members of the church met, having brought their mail ballots with them, and filled their ballots out together, casting votes as their pastor instructed them. This activity was alleged as fraud and likened to ballot-stuffing by some observers, but in actuality is perfectly legal. How a voter chooses to cast their ballot, including what information sources they use, is entirely their choice, and access to those sources is protected by law. As long as those who participated were all legally registered voters, casting only their own ballots, and they attended the church meeting of their own will, free from coercion, no fraud was committed. The behavior of Iraqi voters, if it occurred in the same fashion, is likewise the right of those voters.

Posted by Sarah Sled at 10:07 PM

18 October 2005

The 2006 midterms

Every day seems to bring a new lowest low to President Bush’s approval ratings, and at the same time large majorities of the public think that the country is heading in the wrong direction and won’t get any better in the near future. Common sense would seem to suggest that the Democrats might be looking forward to some success in the 2006 midterms.

Political scientists might also have cause to agree. There’s an extensive literature on partisan “surge and decline? in Congressional elections. When a President is on the ballot, he generates a "coattails effect" that helps members of his party win House seats (the surge) - but this effect vanishes at the next midterm when he’s not on the ballot (the decline). But the 2002 midterms didn't exactly fit this pattern - Bush seemed to engineer a new coattails effect by campaigning heavily for several Republican candidates who went on to win close contests.

National disapproval of Bush and the GOP agenda might not translate into higher approval for Democrats anyway. While the public generally dislikes Congress as an institution, they tend to approve of their individual Representative at higher rates: voters claim to choose “the candidate, not the party?. So as political scientists who claim to have assembled an array of generalizations about political behavior, do we have any way of gauging what’s likely to happen next fall?

Gary Jacobson’s 1989 APSR article argues that “strategic politicians? selectively enter races when national conditions favor their party over the governing party. These quality challengers (measured by whether they have held elective public office before) have more of the political skills necessary to win elections - so even when voters cast their ballot on the basis of local candidates, aggregate election results reflect national conditions.

So how are the Democrats doing in terms of their candidate recruitment for 2006? Larry Sabato’s page of the thirty House races that are likely to be close in 2006 reckons that nine of those seats have Democratic incumbents, fifteen have Republican incumbents and eight are open seats. Looking at races with incumbents, four of the Democratic incumbents are already facing opposition from a Republican candidate who has held elective office before, while nine of the Republican incumbents face similar opponents (several races have no declared candidates, and who the eventual nominee is obviously depends on next year’s primary races).

This would suggest that we’re not seeing a surge in the number of quality Democratic challengers that national conditions might lead us to expect. are not necessarily recruiting enough quality candidates to take back the House. But two questions remain. First, Jacobson’s original article noted that

…the crucial period is not when individuals decide to run for Congress but when they can no longer gracefully change their minds and decide not to run for Congress. Initial decisions to test the political waters may indeed be made right after the last election – if not earlier; final decisions to run or withdraw can be taken right up to the primary.

The numbers above represent declared candidacies, not potential interest in a race. Perhaps high-quality Democratic potential challengers are merely waiting to see whether the GOP’s string of bad news continues before jumping into campaign season?

Second, are we measuring the quality of candidates correctly? The justification for using previous elective office is that these candidates likely have the political skills and ability to campaign effectively, as well as a base and connections that allow them to raise money and start the election with some name recognition. But suppose that the “national issues? affecting the midterms are not economic (as Jacobson assumes they will be) or competency issues like the federal response to Hurricane Katrina. What if the national conditions instead become political reform? Should we then continue to measure quality in terms of previously holding elective office? Or do we want to start understanding quality in the traits of those who come from outside the realm of “politics as usual??

Posted by Phil Jones at 10:54 AM

16 October 2005

The Mother of Presidents?

By my count there are no fewer than 10 U.S. Senators who are seriously considering a run for president in 2008. Among the Dems are Senators Bayh, Biden, Clinton, Feingold and Kerry. Among the Reps are Senators Allen, Brownback, Frist, Hagel, and McCain. Without writing off any of these candidates, we should remember that only 2 presidents (Kennedy and Harding) have come directly from the Senate. It's far more likely that a governor such as Bill Richardson or Mitt Romney will finish first. The Senate produces a lot of presidential candidates but not a lot of presidents.

As I discovered in this article, senators are more likely to run for president and are more likely to lose. For a variety of reasons, sitting senators appear to be relatively nonstrategic about the decisions to jump into presidential waters, at least when compared to governors and ex-senators. My update of the data through 2004 suggests that the pattern still holds. From 1960 to 2004, sitting senators made up roughly 1/3 of all presidential contenders, but comprised only 1/4 of all nominees and just 8% of all winners. Governors, in contrast, made up more than 20% at all three stages. Ex-senators do pretty well too, many because they served as vice-presidents. My advice to a senator who wants the presidency: resign your seat early as Bob Dole did in 1996 and then seek the presidency without constraint.

Posted by Barry Burden at 10:39 PM | Comments (5)

13 October 2005

Once upon a time...

A good well-told story is a powerful frame. Adam Berinsky of MIT and Donald Kinder of the University of Michigan show this in a paper forthcoming in the Journal of Politics entitled, "Making Sense of Issues through Media Frames: Understanding the Kosovo Crisis." Here's the abstract:

How do people make sense of politics? Integrating empirical results in communication studies on framing with models of comprehension in cognitive psychology, we argue that people understand complicated event sequences by organizing information in a manner that conforms to the structure of a good story. To test this claim, we carried out a pair of experiments. In each, we presented people with news reports on the 1999 Kosovo crisis that were framed in story form, either to promote or prevent U.S. intervention. Consistent with expectations, we found that framing news about the crisis as a particular story affected what people remembered, how they structured what they remembered, and the opinions they expressed on the actions government should take.

Berinsky and Kinder use an experiment where they have subjects read "newspaper articles" (which they have written) about the Kosovo Crisis. The articles convey the same information and are almost exactly the same in regard to the text. The only thing the authors manipulate are the organizational structure of the articles and their section headings. They find that the narrative structure of the article is a powerful frame and can in turn structure the attitudes of a reader.

I like the article a lot. The experiment is well thought out and they actually do it twice -- once before 9/11 (in Michigan) and once after 9/11 (in New Jersey).

A few thoughts:

  • Berinsky and Kinder make the (key) point that people organize information around a "good story." Is there no distinction between a frame and the "good story"? The "good story" hypothesis seems more complex than the usual frames that I think of. It also seems to be structural as well as literal. Upon my initial reading, I thought they were drawing a distinction between the frame and the "good story," which I thought made sense.
  • They (at least implicitly suggest that this type of framing doesn't occur in newspapers. Are their findings generalizable to other types of media (especially TV).
  • What about the duration of the framing effects? Since they're interested in cognition and opinion formation and not just a one time question wording effect, I was left wondering how long the frame stayed with the subject. If the treated groups were asked later would they still retain the frame? If they read the control article, would they retain the frame?

I hope the contrived newspaper articles are included in the appendix of the journal article. I spent serveral minutes comparing the three conditions, convinced that they were not the same.

Posted by Andrew Reeves at 11:25 AM

Abramowitz on Culture Wars

The next session of PPBW features a paper by Alan Abramowitz titled "Is Polarization a Myth?" The paper takes on Fiorina et al. and to some degree the Ansolabehere et al. paper we discussed here. It's clearly going to be a while before we settle this debate. Here's the abstract:

This paper uses data from the American National Election Studies and national exit polls to test Fiorina’s assertion that polarization in America is largely a myth concocted by social scientists and media commentators. Fiorina argues that “we [ordinary Americans] instinctively seek the center while the parties and candidates hang out on the extremes? but the evidence indicates that the high level of polarization among political elites reflects real divisions within the electorate. Moreover, contrary to Fiorina’s suggestion that polarization turns off voters and depresses turnout, the evidence indicates that polarization energizes the electorate and stimulates political participation.

Posted by Barry Burden at 10:12 AM

12 October 2005

The Value of Turning Research Questions Upside Down

I was struck by two recent articles that turned old research questions on their heads. What happens when we allow the independent and dependent variables to switch places?

First, do electoral systems determine the number of parties? Following Duverger's logic, the answer had always been yes. But in a recent Political Studies article Josep Colomer argues instead that parties choose electoral systems, thus turning Duverger "upside down." Taking advantage of cases where a country changed its electoral system, Colomer finds that the adoption of proportional representation is as much the product of multipartyism as its cause.

Second, do tough primaries hurt candidates' general election chances? Although the literature has produced mixed findings, the evidence tilts in favor of the idea that divisiveness does cause nominees to do poorly in the fall election. Jeffrey Lazarus argues in a Legislative Studies Quarterly piece that in fact anticipation of general election outcomes causes primaries to be competitive, thus reversing the causal arrow. In particular, challengers are more likely to enter primaries when the incumbent appears vulnerable.

Although neither of these questions is entirely settled, both Colomer and Lazarus have managed to provoke researchers into challenging long-standing assumptions about what causes what. I argue that we ought to be doing more of this.

Posted by Barry Burden at 10:00 AM

11 October 2005

McDermott Paper

The Political Psychology and Behavior Workshop hosted Monika McDermott of the University of Connecticut who presented a paper (pdf) about the effects of AFL-CIO endorsements on Congressional candidates (see Barry's post for more details).

This paper uses data from an experiment as well as data from the American National Election Survey (ANES) and reaches similar findings in both analyses. The most notable findings are that an AFL-CIO endorsement activates liberal cues and increases the likelihood that a liberal will vote for the Democrat. Conservatives, on the other hand, don't appear to be influenced by a labor endorsement of the Republican candidate.

One question raised by an audience member was about the role of information. Aren't those who classify themselves as "liberal" or "conservative" likely to be high interest / information voters? If so, it only seems natural that they would receive and correctly process an AFL-CIO endorsement.

I was personally curious about whether or not the AFL endorsement is an ideological cue or whether it's really just another partisan cue. Two things suggest it is in fact an ideological cue: 1) in the experiment, the party of the candidate is given; and, 2) McDermott says ideology (interacted with the endorsement) does a better job of predicting vote choice than party id (interacted with the endorsement). In any event, I think the paper would benefit by addressing this issue more directly.

Others raised the question of how endorsements were distributed geographically. Are the endorsements focused in certain regions (for instance, Michigan or Illinois, and not in the South). Are races where there are endorsements somehow different from races where the AFL has not endorsed a candidate?

Another interesting point was made about whether or not respondents know what the AFL-CIO is or stands for. One possibility is that respondents view any endorsement as positive. What if, instead of the AFL-CIO, the candidate was endorsed by the ABC-DEF? Could high information liberal Democrats be driving the positive relationship McDermott finds between ideology and the endorsement?

Again, the nice thing about this paper is that there in an experiment and some external validation of that experiment by the ANES so McDermott's findings are not so easy to dismiss.

Posted by Andrew Reeves at 9:01 AM | Comments (1)

8 October 2005

Swing Voters or the Base?

A new report released by Third Way, a centrist Democratic group, made some news this week. The Politics of Polarization essentially recommends that Democrats appeal to moderates in 2008 rather than focus on turning out the liberal base. If correct, it suggests that the Democrats should not try to mimic the Republican approach in 2004. Apparently polariztion serves the GOP well but not the Dems. Anyone know of research (beyond what's in this report) that would support or challenge the Galston and Kamarck argument?

Posted by Barry Burden at 10:20 AM

7 October 2005

A House Divided Will Stand by Guest Columnist Michael Fortner

In his September 26 article, “Black Voters, No Longer a Bloc, Are Up for Grabs in Mayor's Race,? New York Times reporter Manny Fernandez describes the supposed political diversity of New York’s black community. His anecdotal reporting finds black homes split between Mike Bloomberg and Fernando Ferrer: cousin against cousin, brother against brother, father against son, and wife against husband. Craig Livingston, a Caribbean-American real estate developer who has never voted Republican and a member of the newly formed African-Americans for Bloomberg, told Fernandez: “This is the first time that I know of in my lifetime where the black electorate has been this receptive to the Republican agenda, and in New York City that is huge.? Despite the political heterogeneity the article describes, several questions remain: Who are these black Bloombergites? Do they dispel the notion of a politically homogeneous black community? Does race still matter in black political behavior?

Oddly enough, I’m somewhat familiar with this new vanguard of black Republicanism. Through a New York-based email list of black, young professionals that graduated from Harvard, Yale Princeton, Columbia, U Penn, Wesleyan, Georgetown and the like, I received an invitation from African-Americans for Bloomberg to attend an info session/rally for the Republican mayor. Hastened by a Billionaire-subsidized open bar, rare musical performances and a chance to schmooze with New York’s black par nevu petite bourgeoisie, I went.

Standing in a gentrified Harlem among gentrified African Americans and West Indians, I too observed the “diversity? of the black vote: middle-class, old-age pensioners—mostly former blue-collar workers—living out their halcyon years; young lawyers—poor ones from the ADA’s office and rich ones from the city’s top law firms—living no life but work; and, young and middle–age black entrepreneurs struggling to start new enterprises or keep old ones afloat in this new New York—a city without crime, rent-control or pervasive racism. They were black like me: mahogany and caramel faces, jocund spirits and green pockets.

The place was packed. After bumping into the spitting image of Fredrick Douglas, I joked: “I haven’t voted Republican since 1864.? My friend laughed. Then, with as much earnestness as she could muster, she responded: “Bloomberg isn’t really a Republican.? And there’s the rub.

The crowd disassociated Bloomberg Republicanism from Bush Republicanism. In one breath they bemoaned Bush’s response to Hurricane Katrina, calling it racism, and, in the next, they lauded Bloomberg’s outreach to New York’s black community and his effective governance of the city. If this new, local, black republican ideology has any sinew and consistency, it’s about good government—not big government or small government. This new ideology is also about the race. These black Republicans support the safe, crime-less streets of strong Republican rule because it benefits the race. They support pro-business economic policies because it benefits black enterprise. They support faith-based initiatives because it is black non-profits that will receive the grants and render the services. At the same time, they do not support the dismantling of the welfare state. They support school choice and harsh testing regimes because, in their view, black students will benefit. Though not New Dealers, they believe that certain segments of the community still deserve a safety net.

Additionally, I didn’t see a huge divide on cultural issues. These weren’t values voters sipping on gin martinis and snacking on crab cakes: no one spoke of gay marriage or abortion. I would have to attend the late-night rival at the store-front, Pentecostal church down the block to hear such concerns. Even then, the pastor may have sermonized against sodomy but supported Ferrer. Other black religious leaders have endorsed culturally-liberal mayor. It seems access and the new patronage of faith-based programs have made strange bedfellows. As this split among New York's black religious class reveals: black politics is undoubtedly complex now. But it's still racially-oriented.

Despite the prognostications of Fernandez’s article or the wishful thinking of African-Americans for Bloomberg, do not expect a huge vote for the pseudo-Republican mayor. Al Sharpton and other traditional black politicos support Ferrer and will mobilize their networks of activists and associations on his behalf. Furthermore, if Bloomberg wins a substantial proportion of the black vote (around 30%), it will not be a victory for new-age conservatism. If anything, it will be a win for old-style, good-government progressivism. And, according to these black do-gooders, it will be a victory for the race.

Posted by Traci Burch at 8:43 AM

6 October 2005

The Genetic Bases of Political Attitudes

In the past decade or so, an interesting discussion about the emotional and rational dimensions of political behavior has appeared in the literature . Spurred by the alleged inability of rational choice approaches to explain various aspects of political participation, some have attempted to provide an account of the role that the emotions might play in our decision-making. Finally disintegrating the caricature of "emotional behavior" as constisting solely of rash, impulsive, and distinctly irrational acts, works such as Jon Elster's Alchemies of the Mind - in my view unparalleled among contemporary social science books for brilliance and erudition - have driven home the point that emotions are central to cognition. In a similar vein, works such as Marcus et al. 's Affective Intelligence and Rose McDermott's 2004 Perspectives article have begun to show how advances in neurosciences might be of use in explaining some aspects of political behavior. But, a recent article by Alford, Funk, and Hibbing (APSR, May 2005) - which also challenges some of the conventional wisdom about the origins of political attitudes and the sources of political behavior - has received a lot more press (NYT, APSA press release) and is perhaps destined to generate quite a bit more controversy than any of the aforementioned works.

Already in 2004, Alford and Hibbing had published an article ambitiously titled "The Origin of Politics" in Perspectives. In that article, the authors endeavored to sketch out a "theory of the genetic origins of political behavior." More recently, Alford, Funk, and Hibbing (AFH) published a somewhat more modestly tited empirical article in the APSR - "Are Political Orientations Genetically Transmitted?" - which uses the results of twin studies to distinguish the environmental determinants of political attitudes from hitherto ignored inherited traits. Their arguments are greatly appealing: if we really do inherit behavioral as well as more "structural" traits - as biologists from Darwin onwards have widely believed and as recent advances in psychology seem to suggest - why not attitudes that may affect our political preferences?

The basic mechanism posited by AFH is that genetic factors shape behavior in interaction with environmental influences: particular genotypes do not determine behavior; however, our genes affect the extent to which an individual is sensitive to environmental influences on a varieties of behaviors. There is no single gene, moreover, that is singularly responsible for a particular attitude; rather, it is the complex web of chemical reactions in the brain shaped by our genetic make-up that may translate into particular predispositions.

It is quite obvious that the most difficult challenge the authors face is that of showing that specifically political attitudes are to a certain extent genetically determined. For instance, the dominance of the Left-Right dimension of party competition across so many world regions - dubiously cited by the authors as preliminary evidence for the fact that there is a "genetic component to political ideology" - is decidedly a recent phenomenon. As Caramani (2004) shows, the functional Left/Right cleavage - at least in Europe - is the result of 19th century party competition, not a constant feature of politics throughout human history. Distinctly political attitudes such as those that make up a leftist or rightist ideology have only been relevant for the past 100 years or so - definitely too short a period for natural selection to have shaped such attitudes. Rather, the link between genetics and political orientations is only plausible if such attitudes can be traced back to larger personality traits subject to evolutionary forces.

This seems precisely the line of attack chosen by AFH. "The heritability of social attitudes," in fact, is in their view "likely derivative of the heritability of various personality traits" - such as "general openness" - which may affect a host of political predispositions. The evidence marshaled by AFH generally supports the notion that political attitudes are to a varying extent inherited. Monozygotic (MZ) twins (who share 100% of their genetic make-up) have consistently more similar positions on a multitude of topics than Dizygotic (DZ) twins (who on average share only 50% of their genes). The authors do their best to exclude rival explanations dealing with different patterns of childrearing that may potentially confound the results; on this point, their evidence and arguments seem to be sufficiently persuasive. Nonetheless, some skepticism remains when one examines one of AFH's main assumptions - that any difference between MZ and DZ twins included in the study is entirely due to genetics. In other words, the two groups do not differ systematically on any other characteristic affecting political behavior.

It goes without saying that this assumption may potentially run into trouble - in the form of unbalanced samples - in the absence of random selection, but the authors pay little attention to that. Even more importantly,though, whether the correlations found may indeed be indicative of an underlying causal relationship remains questionable in the absence of a clear causal mechanism linking particular sets of genes with particular sets of attitudes. The process by which specific attitudes are shaped by genetics, therefore, remains largely a black box. I am aware that these criticisms may be perceived to be part of a "moving the goal post" strategy. If anything, though, my personal bias is in favor of the authors' arguments, as I consider this to be quite a promising research avenue. Far from rejecting the argument based on any instinctual (genetic?) aversion to the claims or abhorrence for their implications, I am simply wary of the article's key assumption, especially in the absence of a precise causal story. It goes without saying that asking AFH to account, in a single article, for the genes responsible for each of the attitudes they consider would be absurd. However, the absence of any such mechanism renders this article a promising beginning, rather than conclusive demonstration that any political orientation is genetically determined.

Posted by Federico Ferrara at 2:05 AM | Comments (1)

3 October 2005

Do Group Endorsements Matter?

This Friday, October 7, PPBW features a paper by Monika McDermott of UConn titled "Not for Members Only: Group Endorsements as Electoral Information Cues." Maybe the most interesting finding is that labor endorsements can help reinforce stereotypes about Democratic candidates but provide no information about Republicans. Here's the abstract:

Endorsements by groups in American politics have typically been studied as voting cues only for members of the given organization. Using both the formal theoretical and low-information cognitive voting literatures this paper argues for a broader electoral role for group endorsements. Specifically, if groups that have clear ideological or policy preferences endorse candidates, these endorsements should provide all voters with ideological or issue information about the endorsed candidates. This inferred information should then impact voters’ behavior, especially in low-information scenarios. Using both an experimental test and an actual electoral test of the hypothesis and the case study of labor union endorsements, this paper analyzes elections to the U.S. House of Representatives. It finds that when the AFL-CIO endorses Democratic candidates, voters behave as though a liberal message has been sent – liberals are significantly more supportive while conservatives are significantly less supportive than they are when no endorsement is given, regardless of whether they are union members. At the same time, however, the analysis finds no support that endorsements of Republicans have any ideological impact on voting.

Posted by Barry Burden at 2:08 PM

2 October 2005

Slight Change in Schedule

The November 4 session of PPBW featuring Tammy Frisby's work on term limits is being moved to October 28. November 4 will now feature a talk by John Aldrich, cosponsored with the Institutional Development Initiative.

Posted by Barry Burden at 12:14 PM

1 October 2005

Race in the NYC Mayor's Race

A recent Marist College poll shows Mayor Michael Bloomberg with a comfortable 53-38% margin among likely voters over opponent Fernando Ferrer. Given that the Democrats have just wrapped up their primary campaign, and that Bloomberg has been blanketing the airwaves for months, this is not too surprising. What may be of more interest is that Bloomberg is leading among likely African American voters by 50-42% in the same poll.

Some months back, Ferrer got off to a rocky start in the campaign when he was quoted as saying that the police officers involved in the shooting of Amadou Diallo - an unarmed black immigrant - were "overindicted." While Ferrer, of Puerto Rican descent, was able to forge a black-Hispanic coalition four years ago to make it into a runoff with Mark Green, it seems that his early missteps may be hurting him among black voters this time around.

Bloomberg benefited from the racial tensions in the Ferrer-Green fight in 2001, during which, ironically enough, Ferrer lambasted Green for not calling the Diallo shooting a "crime." Green eventually won the nomination narrowly, but the party had a hard time uniting around him after such a divisive primary. It seems as though history may be repeating itself, but this time with Ferrer as the one in trouble. While still early, the idea that a white, Republican, billionaire mayor could capture anything close to 50% of the black vote is certainly something that bears watching.

Posted by Ian Yohai at 12:22 PM