by Scott Winship
I had hoped to end the week with my fifth Data Day post in a row, but I discovered late last night that I can actually analyze the raw data from the survey of Dean activists that I mentioned Wednesday. I think it will be worth it for me to play around with that over the weekend rather than just summarize the original study today. Somehow, you’ll just have to make it through the weekend….
In other news, Strategist editor Stan Greenberg has released a Democracy Corps memo [pdf] with James Carville assessing Democrats’ position in the upcoming elections. By the way, you can get on their website and subscribe to their email list to stay up to date on their work.
Greenberg and Carville note that President Bush’s approval ratings have inched up a bit, but the number of strong supporters or opponents hasn’t changed in the past month. Democrats have a fragile lead in the overall congressional vote. Majorities of voters will either vote Democratic or are basically looking for a reason not to vote Republican. One in five 2004 Bush voters say they will vote for the Democratic congressional candidate.
Dissatisfaction with Bush is rooted in his perceived indifference to the economic squeeze on the middle class, his siding with business over everyday people, his mishandling of the deficit, his lack of a strategy in Iraq, the corruption and catering to special interests within the Administration, and his failure to put the concerns and security of Americans first.
Greenberg and Carville outline a strategy for Democrats to capitalize on the voters’ desire for change, but to find out about that, you’ll have to check out the paper yourself….
And finally, for those who have only recently discovered us – and those who want to rapturously re-read every post we’ve done – I want to point out the new Archive links to the left on the main Daily Strategist page. We’ve been going for about a month now – do let us know how we’re doing! Have a great weekend.
The Democratic Strategist
by Scott Winship
At some point, I am going to rue the day that we decided to call this blog “The Daily Strategist.” Maybe we should have gone with “The Occasional Strategist” or “The Sometime Strategist”. Perhaps “I’ve Got Your $*^#@ Strategy Right Here!”
Oh dear, pardon my French. Anywho, I promised I would return to my post from Monday examining the number of liberals and conservatives. This is going to be a longer and very Mystery Pollster kind of post today. But even if you don’t want the arcane stuff, there are some interesting findings here on public opinion in a number of areas.
The biggest question I received from folks was how I defined the four issue areas I created: security and foreign policy; values; economic and social policy; and fiscal policy. You may recall that I defined people as liberal or conservative on a number of individual items from the 2004 National Election Study, gave weights to the items depending on how well they predicted the presidential vote, and then defined people as liberal or conservative depending on whether their (weighted) liberal responses outnumbered their (weighted) conservative responses.
Here is a summary of the issue area components, with weights included and a breakdown of how many people answered one way or another on the individual items:
Security and foreign policy:
1. human rights (weight=1.4)
• liberal = promoting human rights is a very important foreign policy goal (43%)
• conservative = somewhat important or not important at all (57%)
2. support for the United Nations (weight=1.7)
• liberal = strengthening the U.N. is a very important foreign policy goal (48%)
• conservative = somewhat important or not important at all (52%)
3. democracy promotion (weight=1.3)
• liberal = promoting democracy is somewhat important or not important at all as a
foreign policy goal (78%)
• conservative = very important (22%)
4. neoliberalism (weight=1.4)
• liberal = promoting market economies is somewhat important or not important at all
as a foreign policy goal (76%)
• conservative = very important (24%)
5. soft vs. hard power (weight=4.2)
• liberal = prefer diplomacy and international pressure to threatening the use of force
(i.e., placed self as being between 1 and 3 on a 7-point scale) (39%)
• conservative = prefer threatening the use of force (i.e., placed self as being
between 5 and 7) (39%)
6. military strength (weight=1.6)
• liberal = very or somewhat important for the US to have a strong military, or not
important at all (43%)
• conservative = extremely important (57%)
7. war on terror (weight=6.3, decreased from 23.7)
• liberal = disapprove of Bush’s handling of the war on terror (44%)
• conservative = approve (56%)
OK, let’s pause for some comments here. First let me say something about the weights. These were produced using what’s known as logistic regression. Essentially I predicted whether a person voted for Bush or for Kerry using these seven items. The weights are “odds ratios”. So the weight of 4.2 for “soft vs. hard power” indicates that the odds of voting for Bush rather than Kerry were 4.2 times higher if a person leaned toward military force than if she leaned toward diplomacy, controlling for the effect of the other six items. Intuitively, I’m giving items greater weight the more important they are in predicting how people voted.
I arbitrarily lowered the weight on the war on terror item to be 1.5 times the next-highest weight because otherwise this single question completely determined whether one was coded as a liberal or conservative on security and foreign policy. That’s worthy of reflection – the “effect” of a person’s response to this item on her vote is in some sense over five times the effect of the next-most important item.
The war on terror item itself isn’t ideal – it reflects both attitudes toward Bush’s policies and attitudes toward his competence. But this is the best I could do given the questions available, though I am exploring other approaches. As for other items, it is perhaps problematic to define the liberal and conservative positions on democracy promotion, but since the weight on this item is so small, it doesn’t really have any effect on the results.
Finally, note that some people are neither coded as liberal nor conservative on the hard vs. soft power item. That’s because they could give a neutral response (4 on a scale from 1 to 7). When a person didn’t have liberal/conservative scores on every item in an issue area, I compared weighted liberal and conservative responses for the items she did have a score on. I required that the person have scores on a majority of the items, otherwise she received no overall liberal/conservative designation for that issue area.
Next: values:
1. Constancy of moral values (weight=2.8)
• liberal = we “should adjust our moral views to a changing world” (agree strongly or
somewhat) (47%)
• conservative = should not (disagree strongly or somewhat) (43%)
2. Feminists vs. fundamentalists (weight=1.6)
• liberal = feel more warmth (on a scale of 0-100) toward feminists than toward
Christian fundamentalists (35%)
• conservative = feel more warmth toward fundamentalists (41%)
3. Legality of abortion (weight=1.6)
• liberal = “a woman should always be able to get an abortion as a matter of personal
choice” (36%)
• conservative = a woman should only be able to get an abortion in the case of rape,
incest, or danger to her life (if at all) (46%)
4. Public funding for abortion (weight=1.8)
• liberal = favor a state law providing public funds for poor women to have abortions
(38%)
• conservative = oppose (62%)
5. Gay marriage (weight=2.4)
• liberal = “same-sex couples should be allowed to marry” (or get civil unions if this
option was volunteered by the respondent) (36%)
• conservative = “should not be allowed to marry” (61%)
6. Job protection for gays (weight=1.0, increased from 0.9)
• liberal = favor “laws to protect homosexuals from job discrimination” (74%)
• conservative = oppose (26%)
7. Patriotism (weight=1.5)
• liberal = love for country is very, somewhat, or not very strong (45%)
• conservative = extremely strong (55%)
8. Capital Punishment (weight=5.2)
• liberal = oppose the death penalty (29%)
• conservative = favor (71%)
9. Gun control (weight=2.1)
• liberal = “the federal government should make it more difficult to buy a gun” (55%)
• conservative = should not (45%)
Note that one can be neither liberal nor conservative on the abortion item if she believes that abortion should be allowed in cases other than rape, incest, or danger to the mother’s life, but only once the need for an abortion has been established. I increased the weight on the gay job discrimination item because I didn’t want any of the weights to be less than one.
Economic and social policy:
1. Government activism vs. individualism (weight=3.1)
• liberal = the federal government “should see to it that everyone has a job and a
good standard of living” (34%)
• conservative = it “should just let each person get ahead on their own” (47%)
2. Jobs vs. environment (weight=1.9)
• liberal = we “should protect the environment even if it costs some jobs or reduces
living standards” (scored self as 1-3 on a 7-point scale) (45%)
• conservative = “protecting the environment is not as important as maintaining jobs
and living standards” (scored self as 5-7) (27%)
3. Illegal immigration (weight=1.6)
• liberal = “controlling and reducing illegal immigration” is only somewhat important
or not at all important as a foreign policy goal (42%)
• conservative = very important (58%)
4. Unions vs. corporations (weight=5.5)
• liberal = feelings toward labor unions are warmer than feelings toward big business
(on a scale from 0 to 100) (44%)
• conservative = feelings toward big business are warmer (36%)
5. Poor people vs. business people (weight=1.5)
• liberal = feelings toward poor people are warmer than feelings toward business
people (41%)
• conservative = feelings toward business people are warmer (26%)
6. Opportunity (weight=6.0)
• liberal = agree that “one of the big problems in this country is that we don’t give
everyone an equal chance” (strongly or somewhat agree) (49%)
• conservative = disagree (strongly or somewhat) (34%)
7. Civil rights enforcement in employment (weight=2.8)
• liberal = the federal government “should see to it that black people get fair
treatment in jobs” (54%)
• conservative = it “is not the federal government’s business” (43%)
8. Outsourcing (weight=1.3)
• liberal = the federal government should discourage U.S. companies from “hiring
workers in foreign countries to replace U.S. workers” (66%)
• conservative = it should encourage them or do nothing (34%)
9. Health care system (weight=1.4)
• liberal = a government insurance plan should “cover all medical and hospital
expenses” (placed self as 1-3 on a 7-point scale) (47%)
• conservative = all such expenses should be “paid by individuals through private
insurance plans” (5-7 on a 7-point scale) (34%)
10. Social security privatization (weight=3.1)
• liberal = oppose allowing people to “put some of their Social Security payroll taxes
into personal retirement accounts” (25%)
• conservative = favor (43%)
11. School vouchers (weight=1.9)
• liberal = oppose having the government “give vouchers to low-income families so
their kids may attend private or religious school instead of local public
school” (67%)
• conservative = favor (31%)
Note that respondents can be neither liberal nor conservative on most of these items.
Finally, fiscal policy:
1. Expansion of government (weight=1.7)
• liberal = “government has become bigger because the problems we face have
become bigger” (59%)
• conservative = government has become bigger “because it has gotten involved in
 things that people should do for themselves” (41%)
2. Current role of government (weight=1.5)
• liberal = “there are more things that government should be doing” (58%)
• conservative = “the less government the better” (42%)
3. Taxes vs. spending vs. deficit reduction (weight=1.5)
• liberal = against cutting domestic spending, against tax cuts, and for increasing
domestic spending (25%)
• conservative = either a) against increasing spending and for either cutting spending
orfurther tax cuts, or b) for leaving taxes, spending, and the deficit as is (31%)
4. Spending vs. services (weight=2.5)
• liberal = want more government services even if it means more spending (39%)
• conservative = unsupportive of more government services if it means more
 spending (61%)
5. Tax progressivity (weight=3.9)
• liberal = a) agree that either one’s own taxes or those of rich are too low and
 b) disagree that the taxes of the rich are too high and that the taxes of the poor
are too low (61%)
• conservative = a) disagree that one’s own taxes and those of the rich are too low
and b) agree that the taxes of the poor are too low, agree that either one’s own
taxes or those of the rich are too high, or supported the Bush tax cuts and agree
that one’s own taxes or those of the rich are about right (29%)
6. Spending on social programs (weight=3.3)
• liberal = increase spending on public education, child care, and aid to the poor
(41%)
• conservative = decrease spending on one of the three or leave it as is on two of the
three (38%)
These items are perhaps the most arbitrary, and a couple of them leave many people with no liberal/conservative designation. Incidentally, I separated fiscal policy from economic and social policy because people may, for instance, take conservative positions on the latter but be in favor of progressive taxes or increased spending. Or they may philosophically favor liberal positions on economic and social policies but in practice be unwilling to pay for them (or may disagree about how to distribute the costs).
Once I had everyone coded on these four issue areas (with some people missing one or more codes due to nonresponse), I compared the weighted number of liberal and conservative codes for the four areas to get an overall classification as being operationally liberal or conservative. The weights were again based on logistic regression and again are odds ratios. Security and foreign policy received the highest weight – 10.1, which was actually reduced from 14.5 to equal 1.5 times the next-highest weight. That next-highest weight was for economic and social policy (6.7). Values came next (5.4) followed by fiscal policy (2.7).
And that’s all I have to say about that. Simple, huh? I am certain that this won’t be the last set of estimates I produce, so I would gladly accept feedback and suggestions for improvement. Go nuts….
by Scott Winship
Today I come to you from the eastbound orange Metro line. I promise to get back tomorrow to my earlier post examining the prevalence of liberals and conservatives, as many of you shouted from the rooftops, “Arcane methodological details, dammit!! We want more arcane methodological details!!” While I hope this blog’s primary contribution ultimately won’t be to unite the bickering community of progressive public-transit-riding data geeks, I’ll happily satisfy the public.
As you hopefully recall, yesterday I found that the Democratic netroots community was indeed large enough to impact the electorate in terms of raw numbers. I estimated it at 1.6 million adults – bigger than the civil liberties and gay activist communities, as big as the feminist and minority activist communities, smaller than the group of environmental activists, and much smaller than the labor movement. I want to qualify these conclusions today and to consider what we know about the make-up of the netroots.
First, a sheepish admission – I should have noted the uncertainty in my estimate. “Sampling error” is Statistics 101. Drawing another random sample would very likely produce some estimate other than 1.6 million adults, just because different people would be randomly picked. I should have qualified my estimate by noting that there is a 16 percent chance that the true number of Democratic netroots activists – by my definition – is less than 950,000 and a 16 percent chance that it is more than 2.2 million adults. There is a 2.5 percent chance that the figure is less than 350,000 and a 2.5 percent chance it is more than 2.8 million. My estimates are just that, but we can be reasonably confident that the conclusions I stated about the netroots’ size relative to interest groups’ are accurate. Mea culpa – but note that I’m pointing out the oversight without anyone mentioning it to me first! I’d make a terrible politician.
Upon further reflection, however, I think my definition of the netroots actually overstates its size. By my definition – a liberal or Democrat who was minimally politically active and regularly gets news from blogs – I myself would be part of the netroots, though I rarely read the more stridently loud-and-proud blogs. And if one compares the 1.6 million figure I came up with yesterday to a more reliable estimate of the DailyKos audience by a leading company that tracks internet traffic – 212,000 unique visitors in April 2005 – the possibility that there are lots of people like me included in my definition becomes very real. The same company, comScore, found that while 34,000 people visited DailyKos.com the day before the November 2004 election, just 86,000 visited it on Election Day.
Being in academia right now, I will use this opportunity to bust out the most clichéd conclusion found in that rarefied world: Further research is needed.
Regardless of the size of the netroots community, if it is no different than the rest of the Party, then it wouldn’t really be influential in the sense of affecting Party positions, whether you think that would be a good thing (netroots community) or a bad thing (Chait et al.). So how does it look?
Well, the short answer is that we don’t know. Seriously. The Pew study I relied on yesterday is too small to be used for making meaningful statements about the netroots, even as I defined it. (That ol’ sampling error problem is a big problem when you start slicing and dicing the community.) There was a poll done last month by Chris Bowers of MyDD under the auspices of BlogPac. Bowers contacted a random sample of MoveOn.org members via email. This would be quite an interesting poll, except that just 7 percent of those contacted agreed to participate. You can almost guarantee those folks are different from MoveOn.org’s membership as a whole. I’ll link here, but really, these results shouldn’t be taken seriously. I didn’t bother reading further after getting to the response rate, which Bowers is to be commended for including.
Bowers also references an earlier (voluntary) Blogads survey, but just like his own, it can’t be trusted to be representative of the netroots as a whole. Again, I couldn’t tell you what it says because it’s not worth the time it would take to read it.
Then there’s a recent analysis of the DailyKos.com readership by comScore. Their methods lead me not to put great stock in their conclusions either. While I have more trust that they can track the overall readership of a site like DailyKos accurately, the challenges to ensuring that one has a representative snapshot of a site’s viewers are quite steep. comScore solicits panelists online and through promotions, which means they get a non-random sample to begin with. So they then must weight panelists’ responses so that in the aggregate they are representative of the population of interest. But comScore is unable to fully determine what makes their sample “non-random” – it may be known to be disproportionately male, but what if it consists of people who are less social than average or who are more likely to allow comScore to install tracking software on their computer? Furthermore, to create the weights, they must have accurate information on the “universe” of internet users. But one needs another survey to get this information, and that survey may itself be flawed (which would even affect their overall estimates of readership).
Here’s some sparse information on the comScore findings. Note that according to their methods, more senior citizens read DailyKos than do people between the ages of 18 and 34. This just seems highly unlikely to me, given that aggregate blog readership and internet use are skewed toward the young. Just 34 percent of adults age 65 or older are online, compared with 89 percent of those 18 to 28 years old. My tabulation of the Pew data from yesterday indicates that just 2 percent of senior citizens regularly relies on blogs for news, compared with 39 percent of those 18 to 29 years old.
In short, don’t believe the hype – we know very little about the netroots per se. Later this week I’ll look at a more reliable study of Dean activists that may be the best source of information we have on the Democratic netroots.
by Scott Winship
Over the past few weeks, the blogosphere has been debating the influence of the netroots on the Democratic Party, mostly inspired by the Lieberman/Lamont race. (For examples, click on any word that’s in this sentence.) At issue are two questions: how influential will the netroots be in elections, and will it help or hurt Democrats at the ballot box? I was going to follow up on yesterday’s post, but I couldn’t resist the urge to examine these questions with some data. Disclaimer: I’m not taking sides here and I claim no expertise on the netroots or the ways in which it exercises influence. Should snark, disdain, profanity, or sacrilege proliferate as a consequence of this post, I claim no responsibility.
I’ve located a few surveys of the netroots that are quite interesting. But the data geek in me began drooling when I learned that the Pew Internet & American Life Project lets anyone with access to statistical software download the raw data. I – like so many other people – spent yesterday evening creating crosstabulations on my laptop as I rode the bus home from my gym.
Moving right along, how influential can we expect the netroots to be? For my part in this debate, I’m going to just look at its size as one indicator. I’ll (mostly) leave it to others to elaborate on how my findings do or do not affect the influence the netroots wields. Using a post-election survey from 2004, I defined “the Democratic netroots” as those adults who “regularly” get “news or information” from “Online columns or blogs such as Talking Points Memo, the Daily Kos, or Instapundit” and who are either self-identified Democrats or liberals. Blogs were one of twelve media sources that were asked about, and each of the twelve was a separate question (so respondents didn’t have to choose between competing sources). Rather than answering that they consulted a source “regularly”, respondents could say that they did so “sometimes” or “hardly at all”. Everybody happy?
What does your gut tell you when you think of the percentage of adults that can claim membership in the Democratic netroots? The answer, according to this survey, is 1 percent. One percent of adults translates into 2.24 million people. At first glance, one percent may sound pathetic. But let’s provide some context. Since one strand of the blogosphere debate has compared the netroots with various special interest groups, it might be instructive to consider how large those groups might be. But first we need to isolate the activist subset of the Democratic netroots so that the comparisons below are apples-to-apples. Take a look:
• Democratic netroots members who either attended a campaign rally, donated money to a campaign, knocked on doors, or worked a phone bank – 1.6 million adults (0.7 percent of adults)
• Union members – 15.7 million
• NOW – 500,000 contributing members
• NARAL Pro-Choice America – 900,000 members of their “Choice Action Network”
• Sierra Club – 750,000 members
• National Resources Defense Council – over 1 million members
• ACLU – over 500,000 members
• Human Rights Campaign – nearly 600,000 members
It’s difficult to make comparisons because these groups do not include all activists in a given issue area. Plus there’s obviously substantial overlap among the groups. But it’s safe to say that there are more Democratic netroots activists than civil liberties or gay rights activists, at least as many as there are feminist activists (and hence probably minority activists), but fewer than there are environmental activists or (especially) union members. Given the influence these groups have had on the Party, it seems reasonable to conclude that the netroots really is a force to be reckoned with. On the other hand, these interest groups draw their strength from the popularity of their mission. I would argue that the netroots’ “mission” is to elect candidates who are as uniformly liberal as public opinion in the relevant electorate allows. If I’m right, then mobilizing popular support for an across-the-board liberalism is likely to be significantly more difficult than assembling support for a liberal position on a single issue.
I’ll try to address whether I’m right or not tomorrow. But no big promises that it’ll be possible.
by Scott Winship
Y’all ready for this? Another Data Day! (Admit it, when you read the first sentence that synthesizer-and-drum instrumental that they play at the baseball stadium between innings started playing in your head. And you kind of danced along.)
My vast legion of regular readers should know by now that I am dangerously obsessed with the distribution of liberals and conservatives in the U.S. The obvious first cut at this question is to look at polls that ask people how they identify. You should sit down if you’re not familiar with how these results turn out. Here is a representative set of findings:
• Adults, late 2004, based on my own analyses of the 2004 National Election Study: 35% liberal, 55% conservative (remainder are moderates, non-identifiers, or reported inconsistencies before and after the election)
• Adults, late 2004, Pew Research Center for the People and the Press: 19% liberal, 39% conservative (remainder are moderate)
• Voters, late 2004, based on my own analyses of the 2004 National Election Study: 33% liberal, 56% conservative
• Likely Voters, January 2006, Democracy Corps: 19% liberal, 36% conservative
My own analyses are different from the others in that I have two responses from each person – one before and one after the election – and because the NES tries to get as many people as possible to choose either liberal or conservative rather than moderate. Anyway, the bottom line is that when respondents can choose “moderate”, roughly twice as many people identify as conservative as call themselves liberal. If moderates are forced to choose, they split roughly evenly, leaving 55-60 percent more conservatives than liberals. And these statements hold whether one is looking at adults, voters, or likely voters.
OK, the response from those who don’t like these facts is invariably that a lot of people really are liberal, but the term has been made into a dirty word by conservatives. If you ask people about their policy preferences and values, liberals would be in the majority.
Of course, saying it doesn’t make it so, but this assertion could be true. To test it, I used the NES from 2004, first choosing questions from the survey related to values and values-laden issues; foreign policy and national security; economic and social policy; and fiscal policy.* Within each of these four domains, I created weights for each question based on how well it predicted the presidential vote. Then I categorized everyone as a liberal or conservative in each domain by seeing whether weighted liberal responses to the questions out-numbered weighted conservative responses. Finally, (de-glaze your eyes) I weighted the four liberal/conservative designations based on their predictive power and categorized everyone as an “operational” liberal or conservative.
Now the good stuff. Based on my weighting scheme, the country is evenly split between operational liberals and conservatives. Adults are conservative on foreign policy and national security (52 to 48) and values (62 to 38), but liberal on economic/social policy (57 to 43) and fiscal policy (60 to 40). Consistent with the idea that liberal is a stigmatized word, just 56 percent of operational liberals self-identified as liberal, while 30 percent self-identified as conservative. In contrast, 79 percent of operational conservatives said they were conservative.
I divided the electorate into five groups. The biggest group consists of self-identified conservatives who are also operationally conservative – 42 percent of the electorate. These folks are solidly conservative in all four policy domains, and solidly Republican. Self-identified liberals who are also operationally liberal constitute a smaller group – 27 percent of the electorate. They are the mirror image of their conservative counterparts.
Another 13 percent of voters say they are conservative but are operationally liberal. Forty-three percent say they are Democrats, while just 26 percent indicate they are Republican. Solid majorities voted for Gore in 2000 and Kerry in 2004. They are consistently liberal in the four policy domains, except that they are split down the middle on values. It’s unclear whether values trumps the other policy domains or whether these are the voters for whom liberal is a four-letter word.
Voters who say they are liberal but are operationally conservative amount to just 5 percent of the electorate. Most of these voters are independents. They gave Bush 49 percent of their vote in 2000, but 59 percent in 2004. Tellingly, they are conservative on foreign policy and national security, as well as on values. They split on economic and social policy and on fiscal policy.
Finally, 13 percent of voters do not consistently describe themselves as a liberal or a conservative. This is actually a diverse group. They lean slightly Democratic, but they gave Kerry a solid 59 percent of their vote. Over half are operational liberals. They split on foreign policy and national security, lean right on values, and lean left on economic and social policy and fiscal policy.
There’s much more I could write, which I’ll save for a future post. One point I will eventually expand on is that the fact that so many people identify as conservative even when they tend to prefer liberal policies may imply that they are voting on “character” rather than issues. The liberal/conservative gap in self-identified ideology means something. For now, I’ll just note a couple of take-home points for Democratic strategy.
First, consistent with conventional wisdom, attracting swing voters means emphasizing values and national security. These issues are crucial to improving performance among inconsistent identifiers and liberal-identifying conservatives. Values issues also appear key to keeping and improving performance among conservative-identifying liberals.
It is possible that an economic populist message would be effective among inconsistent identifiers, who appear primed for both economic and cultural populism. Populism doesn’t appear particularly likely to resonate among liberal-identifying conservatives, who became much more likely to support Bush between 2000 and 2004, during which time the al Qaida attacks seem to have pushed them toward Bush. Nor does it appear to be promising as a strategy aimed at conservative-identifying liberals who, after all, call themselves “conservative” mostly on the basis of their views on values issues.
Finally, increasing turnout could be successful, but I found that nonvoters had pretty much the same ideological distribution as voters did. So it wouldn’t necessarily yield a bumper crop of new Democratic votes.
*Space prevents me from going into details, but if you are interested in a memo summarizing my analyses and additional results, send an email to swinship-at-gmail.com and I will do my best to get it to you within a couple of weeks.
by Scott Winship
Mark Schmitt recently took issue with this op-ed by James Carville and Mark Penn asserting Hillary Clinton’s electability in the 2008 presidential election. Criticizing them for their lack of any empirical case, Mark cites approvingly the Strategist’s philosophy of “facts, not factions”. Matt Yglesias followed with a piece examining Clinton’s performance in New York relative to Chuck Schumer, Al Gore, and John Kerry in arguing that she would be a weak candidate. Now Garance Franke-Ruta has linked to a new Gallup poll that sheds additional light on this question. So let’s look at some facts.
While Garance’s post revolves around the views of Democrats toward Senator Clinton, what really matters for the electability question is how independents view her. According to a recent ABC/Washington Post poll, 48 percent of indies have a favorable opinion of her, while 46 percent view her unfavorably. (The rest are unsure.) This is remarkably close to John Kerry’s 49-48 margin among independents in the 2004 election. So an initial conclusion is that with Clinton heading the Democratic ticket, we will be dealing with another nail-biter in 2008. (Of course, much depends on the Republican ticket.)
On the other hand, Clinton’s favorability among Republicans – 26 percent – is significantly larger than Kerry’s performance among Republicans (a whopping 6 percent). Presumably she would end up getting substantially less than a quarter of the Republican vote in 2008, but it may be that she can attract enough Republican women to improve on Kerry’s performance.
That said, when respondents get the chance to say they are undecided as to Clinton’s favorability, just 11 percent of Republicans and 28 percent of independents are favorable toward her while 62 percent and 37 percent view her unfavorably. So, much of her support is tenuous.
That means a big question is whether Clinton’s popularity would go up or down over the course of a primary and general election campaign. Of course there is the possibility that many independents and Republicans who view her favorably will ultimately decide they do not want to relive the battle between the Clintons and the Vast Right-Wing Conspiracy. Many of them may not approve of having Bill Clinton back in the White House, particularly as First Gentleman.
I suspect, however, that Republicans would shy away from the sorts of attacks unleashed against the Clintons in the 1990s. One has to believe that the losses they sustained in the 1998 elections taught them that it is possible to go too far. Indeed, Hillary Clinton’s favorability ratings have fluctuated between 44 and 54 percent since the start of Bill’s second term…except during and after the Lewinsky scandal, when they jumped 8 points in the first days that it broke and remained at 59 percent through mid-1999 (when she revealed she was considering a Senate run). If the Republican nominee has had an extramarital affair, alluding to the Clintons’ marital dramas will prove risky too.
No, they will have plenty to work with without having to prime swing voters to remember the various Clinton scandals. For one, there is HillaryCare. With health care as a major campaign issue in 2008, I suspect that the Republicans would rather face Clinton than other Democrats (who will have their own health care plans and no baggage from 1994). But the Gallup poll Garance links to reveals a second promising front for Republicans.
Among independents, the most common reason given for disliking Clinton was, “Wavers too much on issues to her advantage/wishy-washy.” If the campaign began today, she would go into it with 12 percent of independents believing she’s Clintonesque and a flip-flopper. These are basically the same characterological flaws that Republicans used effectively in the past four Presidential elections. Twelve percent may not seem that high, but keep in mind that this is before the GOP slime machine kicks in. You can bet the 527 groups will be out in full force, if not Rove or his protégés.
Furthermore, Clinton will face strong pressure in the primaries to change her Iraq position and to move to the left generally. As Howard Dean’s candidacy did to Kerry, Russ Feingold’s will set Clinton up to be more effectively portrayed as a flip-flopper in the general election. Indeed, it may be worse for Clinton. She will have much more successfully portrayed herself as moderate going into the primaries (before she zigs) than Kerry did, and because the netroots are feeling far less accomodationist today than in 2003 (at least on Iraq), she will have at least as far to zag in the general election.
So is Clinton electable? Sure. Is it likely she’d be elected? Much less clear. The evidence above gives reason to think that with Clinton as the nominee, 2008 could be the third carbon-copy presidential election in a row for Democrats, which would leave them agonizingly short of victory again. But with the current 50/50 Nation, it’s impossible to say with much confidence what would actually happen.
Update: Regarding Mark’s main question as to whether Clinton is likely to attract strong support from women in general and married and non-Democratic women specifically, I’ve tabulated some evidence from the 2004 National Election Study. Women, but not men, rated Clinton higher than they did Kerry on a “thermometer” scale where 0 equals very cool feelings and 100 equals extreme warmth. The average for Clinton was 59, versus 54 for Kerry. Men rated both between 50 and 51. So Carville and Penn seem correct here. On the other hand, Clinton’s boost among married women was no larger than her boost among married men, and much smaller than among single women. Married women rated Clinton 53 and Kerry 50, compared with 48 and 46 for married men and 66 and 59 for single women. Among Republican women, there was no boost, and she was barely any more popular than among Republican men. The average score for Clinton was 31, compared with 32 for Kerry. The figures for Democratic women were 80 and 73; for Republican men, 28 and 29. Finally, Clinton averaged 63 among independent women, while Kerry averaged just 55. Among independent men, on the other hand, she averaged 52 to Kerry’s 54. So while there’s some evidence a Clinton presidency would energize independent women (including married independent women, which I don’t show here), there’s little to indicate that it would convert other married women or Republican women. Furthermore, these numbers would probably fall over the course of a campaign, given GOP smear tactics.
by Scott Winship
Lately hawkish Democrats have been disparaged for wanting to prioritize terrorism over other problems. Some object to privileging terrorism on substantive grounds. Other objections are strategic. I’d like to address both here with some data. But before I do, I’ll ask that we set aside the debate over Iraq for the purposes of this discussion. Terrorism was only tenuously related to Iraq before we invaded, in the sense that hostile regimes such as Saddam’s could potentially cooperate with jihadists to inflict damage on the U.S. One can oppose the Iraq War, favor redeployment, and still believe that terrorism ought to be the primary concern of the federal government.
First a brief substantive defense (since I’m supposed to be focusing on strategy). Lots of observers, bloggers, and pundits debate to no end how serious the terrorist threat is. But Foreign Policy magazine and the Center for American Progress just released findings from a poll of 116 experts. (Hat tip to Kevin Drum, who hat-tips Eric Martin.) One in three believes it is “likely” or “certain” that a 9/11-sized terrorist attack will occur in the United States by the end of this year. A majority of 57 percent believes we will experience a Madrid-sized attack by the end of the year. Four in five believe that by the end of 2011 another attack as devastating as 9/11 will occur.
While it is true that in terms of sheer capacity for destruction, the jihadist threat does not approach that of the Soviet Union during the decades of the Cold War, there is an important difference between the two cases. Mutually assured destruction can serve as a deterrent within the framework of nations acting as rational actors. But it will not deter stateless men who are committed to martyrdom. The same argument may be made when comparing terrorism to other foreign policy threats such as China.
Terrorism is also unique today in that it has the potential for violently disrupting the lives of civilians on American soil. I tend to agree with Newt Gingrich, who fears that another 9/11 would make the Patriot Act look like it was written by the ACLU. Don’t like federal snooping, sanctioned torture, extraordinary rendition, or enemy combatant treatment? Better hope that we manage to stop that single person it would take to detonate a dirty bomb in Washington D.C. and render it uninhabitable for several hundred years.
As for the strategic defense, many have pointed to recent public opinion polls showing that terrorism has declined in importance in the minds of voters. In a recent ABC News/Washington Post poll, terrorism ranked last behind Iraq, the economy, health care, immigration, and gas prices as a national priority. The last CBS News poll that asked this question was in May and was open-ended. Iraq, the economy, and immigration clearly were deemed more important than terrorism. Polls from NBC News/Wall Street Journal and Fox News come to similar conclusions.
On the other hand, a recent CNN poll asked respondents to rate the importance of a number of different issues rather than having them choose one as the most important. In this case, more people deemed terrorism very important than said the same of Iraq, the economy, gas prices, corruption, immigration, or spying. The Pew Research Center has asked adults a similar question for a number of years. Terrorism has ranked first every year since 2002.
Taken together, this evidence indicates that while terrorism may not consistently be the single most important issue for voters, it is consistently very important. Crucially, a Gallup poll from last month indicates that voters trust Democrats more than Republicans on all of the issues noted above…except for terrorism, where the advantage for Republicans is as large as the advantage for Democrats on Iraq [subscr. only].
In other words, Democrats have an advantage on most of the issues that are important to voters, except on terrorism, which is clearly a threshold issue for them. More to the point, terrorism was clearly the issue that drove the 2004 election, despite the infamously poorly-worded exit poll question that indicated moral values to be the main factor. In the National Election Study, when asked open-endedly about the single most important issue from the past four years, 43 percent of adults chose terrorism, followed by 19 percent who chose Iraq, and 12 percent who chose the economy.
When a number of issues are all important but the Republican advantage on one is relatively large, that single issue may dominate the others even if it’s not the most common issue mentioned when voters are asked for their most important issue. Joe Swingvoter may go to the polls thinking that education is his most important issue but feeling that the candidates are equally good (or bad) on it. But Joe may think that the Republican candidate is clearly better on terrorism, which is also important to him, and so he ends up voting for the GOP.
Uniting the substantive and strategic case for addressing terrorism more prominently, if we do have the misfortune of experiencing another terrorist attack, terrorism will surely reside unambiguously at the top of the agenda. And if a Republican is elected, he or she may not botch things the way our current commander-in-chief has. That could put us in the political wilderness for a generation.
by Scott Winship
We here at the Strategist pursue objectivity like Star Jones pursues TV cameras. The whole point of the magazine is to – as much as possible – use evidence to adjudicate between competing political strategies, putting our biases on the shelf. Obviously, we feel like there’s not enough objectivity out there among our fellow partisans and that evidence can persuade people to change their minds.
But what if putting our biases on the shelf is extraordinarily difficult? What if opinions are strongly resistant to change? Charles S. Taber and Milton Lodge of Stony Brook University examine these questions in their new paper, “Motivated Skepticism in the Evaluation of Political Beliefs” (American Journal of Political Science, July 2006). Their work is sobering, and should be particularly so for Democrats who advocate strategies of “leadership”, “education”, or other approaches intended to change public opinion rather than accommodate it.
Taber and Lodge report a number of fascinating results from experiments they ran using their students as subjects. Subjects were asked to evaluate the strength of different arguments in favor of or against affirmative action or gun control after answering a battery of questions on their political views, including their views on those two issues. The arguments were based on statements by interest groups involved in the policy debate over these issues and edited to ensure that the pro and con versions were equivalent in terms of structure and length.
There was a notable tendency to evaluate more positively arguments that were congruent with one’s position. Subjects spent more time reading the arguments that they were predisposed to disagree with than the arguments that were congruent with their position. And when asked to evaluate the arguments, they criticized the incongruent ones much more often than they did the congruent ones, which they tended to speak well of. The implication the authors draw is that all that extra time spent reading the arguments that contradicted the subject’s view was devoted to poking holes in the arguments.
Students were also asked to evaluate eight out of sixteen possible arguments for or against a policy, which they chose by clicking on one of sixteen boxes to reveal an argument. The only information they had on the arguments was their source – two being organizations in favor of the policy (affirmative action or gun control) and two being opposed organizations. The students were told to seek out information in an unbiased manner so that they could educate other students on the issue. Nevertheless, subjects were more likely to choose arguments (boxes) associated with a source that espoused their own view than arguments from an organization with the opposite view.
Finally, students answered a second battery of questions on affirmative action and gun control. Taber and Lodge found that people generally became even stronger supporters of their original position after completing the experimental tasks. That is, even though the tasks were neutrally presented, these subjects made choices that reinforced their beliefs rather than challenging them, and their beliefs grew even stronger as a result.
An interesting footnote to all of this research is that the patterns were strongest among those students who had the strongest initial beliefs and the most political knowledge. Results for subjects in the bottom third in terms of strength of beliefs or political knowledge were less pronounced and rarely statistically meaningful, though they usually followed the same pattern as for subjects in the upper third.
Taber and Lodge’s paper points to the strong resistance we all have to questioning our beliefs. Anyone who has ever had extended discussions with conservatives trying to get them to concede that people really don’t have equal opportunities in life has experienced this phenomenon first hand. To abandon a belief in equal opportunity would force a complete restructuring of economic conservatives’ politics, if not their entire identity. But liberals and centrists can be just as guilty of self-delusion.
More importantly, if people tend to seek out information, news, and friends who tend to confirm their own beliefs, it will be quite rare for anyone to change their political views in any profound way. That means that political strategies rooted in bringing public opinion around, changing minds, or “leading rather than following” face significant psychological hurdles. This all accords with my own gut belief (bias) that political elites rarely change public opinion; they instead exploit situations where policy doesn’t accurately reflect public opinion. Anti-abortion activists haven’t moved people to the right on abortion; they have groped for restrictions that bring abortion policy more in line with public preferences (e.g., no partial-birth abortion, no public funding, etc.). There remains solid majority support for the Roe v. Wade decision, and no amount of framing will alter that.
This question of how malleable attitudes are couldn’t be more important – if accommodating moderates continually moves the median voter ever rightward, then center-left strategies need to be rethought. But if parties don’t change public opinion, there’s no basis for the claim that Republicans can simply keep moving the “middle” in their direction. On this question itself, all of us must strive to overcome our own psychological barriers to seeing the world clearly.
by Scott Winship
There are a lot of things I don’t understand. (No, it’s true.) I’ve never understood the appeal of The Big Lebowski, for instance. I don’t understand how Cops could have had higher ratings than Arrested Development. And I don’t get how anyone can enjoy the taste and texture of coconut.
Similarly, I think a lot of progressives are confused about the values gap. (Not at all a strained transition there.) For example, many seem to believe that the answer to Democrats’ problems is for our candidates to find religion or fake it. This is frankly ridiculous. Religious beliefs are among the most personal we have. No one can be expected to change their beliefs out of electoral concerns. And good luck faking greater devotion. The bottom line is that Democratic politicians are disproportionately drawn from (relatively) secular areas and segments of the population compared with Republicans. Absent an active campaign at the party level to change this, it’s unlikely that Democrats will end up more outwardly religious in the future.
The question is: is this a problem? I could make an argument that it is not. The U.S. is among the most religious nations in the world, but faith remains a mostly private matter. Over three quarters of the population says that religion is an important part of their life, yet little more than half pray everyday, and only a third attend religious services weekly. The latter are roughly equally divided between Democrats, Republicans, and Independents. (This all comes from the National Election Study of 2004.)
On the other hand, ceding to the Republicans so many of those for whom religion is an important factor in their voting is quite a consequential decision in a nation so closely divided politically.
A number of Democrats are themselves devout, but few are comfortable with and effective at describing how their faith affects their life or policy orientation. Amy Sullivan has a must-read piece in Slate on Barack Obama’s recent speech on faith and politics. She clearly outlines how the speech shows the way for Democrats struggling to connect with religious audiences.
The problem is that Obama can reach the devout largely because he is one of them. Other than Jimmy Carter and Bill Clinton, who could both cite scripture effortlessly, no Democratic nominee going back over 40 years has conveyed that religious faith was an important influence on their politics.
This conundrum leaves few possible strategies for Democratic nominees running in places where religious faith cannot be ignored. I disagree with those who counsel “reframing” on the theory that values voters are just voting on “values” per se rather than on specific values they hold. It is as if the problem is what gets defined as an important values issue rather than the poor performance among voters for whom current values issues are important. I don’t deny that many religious Americans who oppose abortion and gay marriage also support greater generosity toward the poor. But the problem is that too large a fraction of them decide their vote on the basis of abortion or gay marriage instead of greater support for the poor. It is not that they don’t understand the parties’ positions on these issues – if redistribution is more important to them than banning abortion, they will vote for the Democrat.
At this point, you’re probably asking, “But what should we do, Scott? Show us the way, o ye car-less sage.” No? Well, let me offer my two cents anyway. Candidates who are themselves religious yet pro-choice or in favor of gay marriage should be prepared to discuss how they reconcile their faith and these relatively liberal positions. Bill Clinton is perhaps the master in this regard.
More-secular candidates should be up front about their views rather than trying to skirt these questions with poll-tested pablum (“Abortion is a decision between a woman and her doctor.” “I believe marriage is between a man and a woman.”). The data does not support those who think that conventional progressive absolutist positions on abortion and gay marriage command majority support. Nonetheless, secular candidates can pick and choose their fights more carefully, acknowledge the lack of consensus on controversial issues, promote incremental measures that move the country toward their goals, push for state experimentation, and adopt rhetoric and tone more consistent with public opinion. This is how the President deals with abortion, rather than espousing a loud-and-proud anti-abortion agenda.
Addressing the values gap doesn’t require that progressives clothe themselves in an entirely new moral wardrobe. But it does require that we wear the occasional conservative suit when we would rather be “fashion forward”, that we throw out those garments in our closet that never fit, and that we resign ourselves to bringing the rest of the country around rather than expect that traditionalists will immediately accept our haute couture.
by Scott Winship
A central line of social psychological research maintains that people have stable rank-ordered values, each of which inform behavior across diverse settings. On the other hand, political behavioral theory offers a number of reasons why voting might not be a simple matter of consulting one’s ordered values. Values may be in conflict, and choosing to prioritize one may come at the expense of others. Another possible complication is that voters may be indifferent, thus unable to rank two competing values. Or value rankings may be situational. Finally, the possibility that value ordering may be manipulable by political elites is highlighted by the current popularity of “framing”.
In his new paper, “Value Choices and American Public Opinion” (American Journal of Political Science 50(3), William G. Jacoby examines the question of how values are ordered. The survey Jacoby used first defined four values for respondents: liberty, equality, economic security, and social order. It then asked respondents to rank pairs of these four values presented one at a time so that each value was evaluated against each other value.
Jacoby found that economic security was ranked highest, followed by liberty, equality, and social order at the bottom. When he considered how many people produce a clearly ranked set of values from the pair comparisons (as opposed to a set of inconsistent choices), he found that four in five adults have a clear ordering. And when he looked at sets of three values, between 90 and 94 percent of adults ordered them consistently.
These figures overstate how many people truly have consistently ranked value preferences however. Some people who end up with ordered rankings really have non-ordered ones but for one or more pair comparisons, they couldn’t prioritize the two values and essentially chose one randomly. If people who are truly ambivalent or indifferent between a pair of values flip a coin when stating a preference, then the one-fifth of adults whose responses were inconsistently ranked would translate into one-third of adults actually having non-ordered preferences.
Another interesting finding is that adults with less education, political knowledge, and income are more likely to order the four values inconsistently. Jacoby persuasively argues that this is evidence that inconsistent rankings are primarily due to indifference arising from a lack of political information and, more generally, education and time.
Finally, Jacoby provides a creative test of whether framing affects policy preferences. If framing is effective, it should be the case that the influence of some values on support for policies increases while the influence of other values decreases under alternative frames. In contrast, Jacoby finds that the effects of liberty, equality, and economic security on support for government spending are statistically the same regardless of whether spending is framed as being for the poor, for minorities, or for the general public. Jacoby doesn’t seem to appreciate, however, that values could be operating through partisanship, which is also included in the models and which does have different effects depending on the framing.
Jacoby’s study shows that for most adults, there is a clear ordering of values. Apparently, choosing between competing or conflicting values is not a problem for most people in forming policy preferences. Americans value economic security above liberty, liberty above equality, and all three above social order. Where values are not clearly ordered, it is mainly due to low education levels. Those who are less educated consider questions of value ordering less than well-educated adults do, perhaps because they have less time, interest, or ability to do so. Unfortunately, Jacoby didn’t look at whether the policy positions of the least educated reveal similar indifference and whether the positions of the most educated are consistent. This question is crucial to interpreting survey responses on preferences for spending, tax cuts, deficit reduction, and other policies.
Update: Here’s a link to the paper.
