Post/ABC poll shows electorate leaning in Dems’ direction

November 4, 2007

Following up on yesterday’s item about the latest Newsweek poll, new data from the Washington Post/ABC News poll is also worth considering. Apparently, Americans aren’t terribly pleased with the status quo.

One year out from the 2008 election, Americans are deeply pessimistic and eager for a change in direction from the agenda and priorities of President Bush, according to a new Washington Post-ABC News poll.

Concern about the economy, the war in Iraq and growing dissatisfaction with the political environment in Washington all contribute to the lowest public assessment of the direction of the country in more than a decade. Just 24 percent think the nation is on the right track, and three-quarters said they want the next president to chart a course that is different than that pursued by Bush.

Overwhelmingly, Democrats want a new direction, but so do three-quarters of independents and even half of Republicans. Sixty percent of all Americans said they feel strongly that such a change is needed after two terms of the Bush presidency.

Republican pollster Neil Newhouse said, “It is a political environment pretty heavily tilted toward the Democrats.” I think that’s a safe assumption.

The best thing about Post/ABC polls is that the Post publishes the internals online, offering a helpful look at the details. Today’s release is chock full of interesting details.

Among the results that stood out:

* A majority of Americans (51%) have a favorable opinion of the Democratic Party. For Republicans, that number is 39%.

* On the issues, Americans trust Dems over the GOP to handle Iraq (50% to 34%), healthcare (54% to 29%), the economy (50% to 35%), taxes (46% to 40%), and immigration (42% to 35%). The parties are about tied on handling terrorism, with Republicans leading 42% to 41%.

* In the race for the Democratic nomination, Hillary Clinton is still way out in front, though not by quite as much as a Post/ABC poll in September. She now leads with 49% (down from 53% in September), followed by Barack Obama with 26% (up from 20%), and John Edwards with 12% (down from 13%). All other candidates are below 5% nationally.

* On character traits, Clinton excels among Dems on questions involving who is the strongest leader, the best capable of handling the Middle East, and the most electable in a general election. She fares much worse, however, on the question about which candidate is the most trustworthy.

* In the race for the Republican nomination, John McCain is on the upswing. Rudy Giuliani still leads the field with 33% support (down from 34% in September), followed by John McCain with 19% (up from 12%). Fred Thompson is third at 16% (down from 17%), Mitt Romney is fourth with 11% (unchanged), and Mike Huckabee is fifth with 9% (up from 8%). All other candidates are below 5% nationally.

* In hypothetical general-election match-ups, Clinton leads Giuliani (50% to 46%), McCain (52% to 43%), Thompson (56% to 40%), and Romney (57% to 39%).

* A majority of voters (54%) want Dems to keep the congressional majority.

* When respondents were asked whether we “need to elect a president who can set the nation in a new direction” or if “we need to keep the country moving in the direction Bush has been taking us,” it’s a landslide — 75% want a new direction, 23% want to stay the course.

* And the poll also included an interesting question about gay rights: “Do you think homosexual couples should or should not be allowed to form legally recognized civil unions, giving them the legal rights of married couples in areas such as health insurance, inheritance and pension coverage?” It was surprisingly one-sided: 55% support legally-recognized gay relationships, while 42% do not. Support for civil unions has jumped 10 points since June 2006, which is pretty impressive.

I’m not the least bit surprised that Hillary is cleaning Romney’s clock. Survey USA’s recent polls show that in a hypothetical matchup, Hillary would win every Southern state, and even win Montana, Indiana and South Dakota. A hypothetical Edwards-Romney matchup has Edwards winning every state but Utah.

Go, Mitt, go …

  • Why don’t the Dems, from the DNC on down, unite on a single theme, such as, “We KNOW things are going in the wrong direction. And this is what WE propose to do”.

    They should also CLEARLY state why they can’t end the war. They have to explain, clearly, in terms the most cretinious Republican can understand, why they don’t have the numbers in Congress.

    The Dems are their own worst enemy in terms of getting the correct message out.

  • I noticed that there is no detailed description of the polling methodology in the WaPo article. Did they call people with landline telephones only, or also cellular phone users as well? How did they generate a “random national sample”?

    It is arrogant and ridiculous that this lack of critical thinking is foisted upon the America People as meaningful.

    The “just trust us” methodology even extends to “public-opinion research.” The Corporate Military Industrial Media complex is doing a heckuva job keeping us line and voting the way that they want us to vote.

    Not this bumpkin.

  • JKap…

    Are you implying that (suddenly) standard polling techniques are no longer valid?
    Do you discount this poll because you disagree?

    Would that poll be acceptable if it were attributed to the RNC? The Times?
    Ann Coulter?

    Seems like a slight-of-hand argument to draw attention elsewhere or to dispel a harsh view of the current mess.

  • I’ll be shocked if we’re not inaugurating Hillary in January 2009.

    I recently came across this fellow who theorizes that presidential election campaigns are largely irrelevant, and that the outcome of the popular vote is largely set in place before the campaign begins. His system (which he insists on referring to as “the thirteen keys” to the presidency) has correctly predicted the popular vote winner in every election since it was created, and retrospectively, accurately predicts the popular vote winner in every election all the way back to the civil war.

    The specifics of the system interest me less than the idea that elections are largely a referendum on the previous performance of the ruling party, and that if that performance falls short (in a set of fairly objectively quantifiable ways) then the challenger party’s candidate will win. The particular characteristics of the challenger, and the campaign (s)he runs, are essentially meaningless. If the conditions are right, virtually any challenger can win. If they’re wrong, no challenger will be able to survive. (To those bemoaning the notion that “Hillary can’t win,” I think you’re being silly. Of course she can win. And in this particular election season, she is very likely to win. The 13 Keys system already predicts her victory.)

    One other implication of the system is that modern presidential campaign coverage, focused on the irrelevant “horse race,” is simply a massive exercise in missing the forest for the trees. I find this notion both plausible and deeply emotionally satisfying.

  • These results are encouraging for Democrats and progressives in the sense that a sweep in 2008 is possible, but I think very discouraging when you look at what this country has become – deeply divided.

    “On the issues, Americans trust Dems over the GOP to handle Iraq (50% to 34%), healthcare (54% to 29%), the economy (50% to 35%), taxes (46% to 40%), and immigration (42% to 35%). The parties are about tied on handling terrorism, with Republicans leading 42% to 41%.”

    The key word above is “trust” which I think is a misnomer. Americans want completely different things. The word “trust” implies we’re all in this together, but that Americans think Democrats are more likely to get us “there” than Republicans. But there is no “there” there. (Very loosely) half want peace, universal health care, progressive taxes, firm action on global warming, a new energy policy away from fossil fuels, restored environmentalism, an active, competent government in domestic, social programs, etc. etc. etc. But the other half doesn’t want that. They want a Norquist sized government, dog eat dog social Dawinism, unfettered capitalism and a militant, imperialistic and aggressive foreign policy.

    How can we survive as a nation when we are so fundamentally different as a people?

    So yes, 75% think Bush was horrible, but for millions that’s because he didn’t cut out all the social programs and because he didn’t lower taxes enough on the rich and because he didn’t wage war competently in Iraq and because he hasn’t attacked Iran and because he didn’t deregulate enough.

    In the last thirty years 40% of the American people have become war mongering, mad and mean as hell, fundamentalist right wing extremists, and that’s a disaster, a tragedy
    for which I see no solution. Only more so in the future.

  • Re: Barrett Wolf @ #4

    The current state of the American political system is corrupted, everyone realizes this to some degree. The Bush Disaster is almost in its last throes. Yes, I get the sense that the majority of the people that I interact with are not part of the No Peace, More War Movement.

    But I am not taking a side on any position from the poll per se, but as you articulated, I am questioning the polling methodology and the lack of a detailed description in the article.

    It’s like they want us to “trust them” because they’re Washington Post/ABC News, imagine that.

    The “standard polling techniques” you refer to should be scrutinized, since they are so integral to the American political system, the two-party system that Rudolf W. Giuliani says “has served us well.” Many people in my experience base their vote on “who is ahead in the polls.”

    Well, I’d like to know, since many people have “went wireless” –how have polling methodologies changed to reflect this reality? What about the advent of the internet as a means of gathering information from a diverse demographic? Why are there no “scientific” online polls conducted?

    Unlike the poll in question, many polls purportedly target “likely voters.” Well, what about the 50% of Americans who do not vote and forgo representation? They are not polled. Shouldn’t we be looking for ways to expand democracy in our own country? When 50% of the people are omitted from “public-opinion research,” aren’t those people then locked out of the system to some degree? Isn’t inclusion better?

    I know that it can be an exercise in futility, but I am trying to call everyone to a higher standard.

  • JKap, to some degree the answers will only be available on a forward-looking basis from the results: if we start seeing electoral outcomes that are wildly divergent from the polled predictions, that is what will most undermine the credibility of polling.

    I do not think that anything has happened recently that repeals the basic principles of statistics (and I say this not just in response to you – there have been a lot of attacks on polling results lately on these boards, including attacks on sample size). But the reality is that so far there have been few examples of the polls being wildly wrong. Yes, the exit polls have not seemed to match the outcomes in 2000 and 2004. First, those were close elections and the divergence was largely within the stated margin of error. Second, there is good reason to believe the divergence was caused by fraud – that is, the polls were actually more accurate in reflecting the electorate than the “legal” outcome.

    Your biggest concern with methodology appears to be that “cord cutters” are undersampled. I have yet to see a reliable study that puts cord cutting at more than 10% nationally; the best studies suggest around 6-8% right now are cord cutters. Many of those are young — and over a long, long history, regardless of conditions, reglardless of candidates, the young choose against voting in bigger numbers than other age groups. As a result, undersampling of cord cutters causes less problems in polling accuracy. And then the question remains whether cord cutters as a class are statistically different in their politics from landline users of the same demographics who do get sampled; there is no evidence I am aware of that this is the case — and given how small the cord cutting group is, the political differences would have to be really really extreme to impact the validity of the statistics.

    There are two ways to mitigate the concerns you have about polling in something closer to real time, however: (1) compare to real money electronic prediction markets, which are both less scientific in that the “sample” is self-selected but because money is at stake the participants have incentive to factor in all available information, and (2) look to poll aggregator sites which, by aggregating numerous polls, tend to “balance out” the impacts of different methodologies, bringing down the error in the aggregate. Several of these poll aggregators were exceedingly accurate in 2006.

    My guess is that if you compare prediction markets and aggregators to a given poll, the results will largely be the same. If I am right, there is not much cause to believe that suddenly we have entered a political era fraught with inaccurate or misleading polls. People have always disliked the results and importance of polling, and have always looked to attack the methodology as a result, but over time polls have proven to be fairly reliable. The underlying statistical principles prove themselves in non-political-poll settings day in and day out; they are by now pretty solid.

  • I have yet to see a reliable study that puts cord cutting at more than 10% nationally; the best studies suggest around 6-8% right now are cord cutters.

    I allege that reliable studies on this subject are not in the offing predominantly.

    Yes, I know, Zeitgeist. The system has “served us well.”

    Like I said, exercise in futility.

  • Wireless devices have surpassed landline phones as the favored voice calling system in the United States.

    Surveys of U.S. users conducted earlier this year by Mediamark Research Inc. (MRI) reveal what the New York analysis house calls a “milestone” in telecom history. The percentage of Americans in cell phone-only households now exceeds the landline-only counterparts, according to the recently issued, seven-page white paper, “The Birth of a Cellular Nation.”

    Cell Phone-Only Rises

    Based on survey responses between September 2006 and April 2007, Mediamark says cell phone-only households now represent 14 percent of the population, while landline-only households dropped to 12.3 percent. In a March 2006-October 2006 survey, wireless-only and landline-only households were pegged at 12.4 and 14.5 percent, respectively.

    […]

    I haven’t bothered to scrutinize the veracity of this article. But the point is scrutinize the status quo.

    Look at what the status quo has delivered for us: The Bush Disaster.

  • It is not, as one would suggest, an “exercise in futility” to conduct a poll using hardwired phones.

    I don’t know of many teenagers walking around with great big plastic telephones in their pockets; but I do know many who have cellphones provided by their parents. Would these not-even-old-enough-to-vote “cordcutters” be targeted for a “random sampling?”

    How would you eliminate them from the sample?

    Also, I’m not aware of a lot of people who turn off their hardwired telephones; most cellular customers do on a regular basis at one time or another. Most cellular customers also demonstrate the penchant for using voicemail—and you cannot conduct a poll with a voicemail system. It’s almost as fallible as trying to gather real-time data via email, and I am one person who does not spend my entire day checking my email for “important news.” I already dump twenty to thirty blaring messages from the Ron Paul people on a daily basis; why would I want to reply to them?

    Finally, I do know several people who have gotten “survey calls” on their cellphones. Each and every one of them do the exact same thing—they hear the words “survey” or “poll”—and they push the little “disconnect” button. Or they’ll see that the call is coming from an unfamiliar number—or an unidentified (anonymous) number—and they won’t even pick up the call.

  • JKap, you are conflating apples and oranges. Just because the two party system doesn’t work well does not mean statistical principles don’t work well. Put differently, when I explain why polling is probably fairly accurate, to accuse me of defending the two party system is a non sequitur.

    Yes, there are some surveys suggesting that cord cutters are as much as 12% (that 14% you posted is by far the highest I have seen – and by the way numbers who “primarily” communicate by wireless are meaningless – it is just a minutes of use study – as most people have both cellular and landline phones). The FCC, in their most recent annual Report on CMRS Competition did what you admit you have not: they had technical staff look not only at available surveys but at how they were structured. The FCC’s best estimate was the 6-8%; I have seen financial and industry analysts give numbers as low as 4-5%. But the number could actually go fairly high before it would matter because of the other points I made in my post which you didn’t address, choosing instead to argue that your distaste for the two party system must somehow mean facts are no longer facts.

  • Let’s just keep it clear that the “new direction” half the Republicans want is a far right fascist theocracy – “Christofascism” to confront “Islamofascism” and hurry on the Apocalypse.

  • To answer JKap’s question above, without even looking I can tell you that the percentage of cell-phone-only households included in this Post/ABC poll or any other professional telephone is 0%. That’s because federal law prohibits mass dialers from calling numbers that may be charged a fee for incoming calls. So cell phone exchanges are strictly off limits to them.

    As for how the random-digit dialing (RDD) lists are compiled, it’s fairly standard procedure that varies very little from one pollster to the next. There’s an industry organization called the Public Opinion Research something or other that provides a set guidelines that that virtually all of the major companies follow in constructing their RDD lists. I believe the ABC/Post polls actually farm out their field work to another company that specializes in that kind of thing.

    The cell phone issue is a growing concern to people in that business but to date has not proven to be a huge factor in practice, at least with regard to election polls. Note that in the article JKap linked to, the comparison was between cell phone only and wired phone only households, whereas the vast majority of American households now use both. The number of US households with no phone service of any kind is pretty small, around 5%, so if you add the 14% of cell-phone-only households from JKap’s article — which is the higher than other estimates I’ve seen but is probably credible — you still have 80% who are reachable by wired phone.

    In the last few elections we’ve also seen that polling done in the final days before elections have been remarkably accurate predictors of actual election results. Most of the time the difference in percentage shares for the major candidates in the actual election falls within the expected margin of error of the difference in their shares in the final polls. So there’s little to suggest so far at least, that in terms of voting preferences the 20% of US households who can’t be reached by wired phone differs significantly form the 80% who can.

  • The Republican party was going to become irrelevant in a couple of decades due to demographics changes in the U.S. However, the incompetence of the Bush Administration has just sped the timetable up for a couple of decades.

    The real question is what will the U.S. be like as a one party state where the Democratic Primary is the only relevant election. In such a scenerio virtualy all elections will be meaningless.

  • I agree wholeheartedly with the comments of hark and of Tom Cleaver. the polling question I’ve heard a number of times, variants of ‘are we headed in the right direction?’ is fairly useless. many of the respondents who answer ‘no’ think the last 7 years have not been far enough to the Right. They think we are pussyfooting around in Iraq and should worry less about civilian casualties, they think WE should use nukes in Iran, they are disappointed that more has not been done versus gays and women’s reproductive rights, etc. etc. but I would add a couple of big issues that I think could use some more attention for the average rank and file Republican’s anger at Bush — numero uno: IMMIGRATION — he is way too lax in their minds. two: GAS PRICES — it’s our god given right to have unlimited cheap gas for our behemoth cars.

    • Comments are closed.