Eh, here’s a little project: count the number of gay couples on prime time television shows. Then count the number of black-white interracial couples on prime time television shows. Try it again with soap operas.
I think the Marx question would be pretty silly, since the vast majority of American Democrats have never read a word of Marx (though they may have heard something about "the opiate of the masses"), and are probably about as averse to the name as most American Republicans.
But I do think the racial questions are important because, while interracial marriage may not be up for debate, other issues related to race are, and it's not so bad to know how people stand on race. And interracial marriage is a pretty good weather vane.
It might help to read the other comments by that same employee, Dustin Ingalls. Relevant to the discussion we've been having, he notes that they asked the interracial marriage question of everyone, and that the questions in the report were not the only questions asked. This is puzzling (why didn't they include the other questions in this report? Are the results coming out in future reports?), and it may mean the survey was long, but it also answers one of the objections Mark and others have raised.
On the age issue, he also gives info on the margin of error for the young age group. This doesn't completely answer the concerns, but it helps a bit.
By the way, since a consistent theme from some commenters has been that this was a hit job, I thought this comment from the PPP blog by one of the PPP employees might be of interest:
BTW, we did ask this interracial marriage question of everyone, not just Republicans, but we'll be releasing the full results on that at a later date.
Eh, I'm not sure support for gay marriage and support for interracial marriage are the same thing. I also think it's interesting that, though it's within the margin of error, the highest % of respondents with favorable ratings of Palin was among the youngest age group, suggesting that they are in fact quite conservative.
Oh, I agree, but I don't intend it as the sole data point. Instead, it's another piece in a chain of evidence in a chain that supports the Mississippi poll, or at least suggests it's not an anomaly.
And let's be honest, I doubt anyone from the south is all that surprised. I mean, I've been to Mississippi, and while I love the state (I really do, and not just because it gave us this), racism is much more on the surface there than it is in much of the country in 2011 (parts of South Carolina, Mobile, and parts of Louisiana are the only places I've seen it so readily displayed in the last 15 years). You don't have to go looking for it, even among young people.
That a substantial percentage of African Americans believe that the U.S. government plays some role in the AIDS epidemic -- whether it created the disease, is withholding a cure, is using the disease against black people specifically, or a handful of other conspiracy theories -- is pretty well known, and completely irrelevant. But way to go with the, "I know you are but what am I" response.
It's not about a law, but it is about interracial marriage. And it's a few years old, but not old enough to not be relevant. Notice that nationwide, support for interracial marriage among whites over 50 is around 64%, suggesting that 36% either disapprove, don't know, or don't have an opinion. It's not unreasonable to expect that number to be higher in Mississippi or any of the former confederate states, and to be even higher among conservatives in that age group.
That poll doesn't help to explain, and perhaps even casts doubt on the youth results from the Mississippi poll, but since it's a southern state exclusively, and only polls conservatives, I don't think those results are outrageous.
Let me just add, before Heidegger writes 1200 words about how I must be a commie radical who's bombed some government office in the 60s because I think all conservatives are racist, that I don't, in fact, think all conservatives are racists. Or at least, I don't think all conservatives are any more racists than the average person (and I think to some degree everyone's a racist, including myself; the attitudes are too deeply ingrained in our culture for us not to be). In fact, I don't think most conservatives are any more racist than the average. But I do think you're likely to find more racism among conservatives, and in particular more overt racism. And of course, there's data to back me up on that.
I will say that the age issue (which has been discussed here already; good eye, Tom) is somewhat puzzling, but as people in that age group do tend to vote Democrat, it doesn't seem unreasonable to speculate that young Republicans tend to be more... let's say, fervent.
As Mark mentions in this post, there's independent evidence that suggests the poll isn't inaccurate.
By the way, if you look at political surveys over the phone, you'll often find policy questions thrown in like that. In general, the pollsters (and the people who hired them) want a quick look at an issue or two, and a bunch of information about potential voting patterns. In short polls like that (and it is short, even if it reads long -- probably took fewer than 5 minutes to complete), they usually put them near the end before the demographic stuff. It may lead to inaccuracies, but there's no inherent reason to think it does. As someone who, many years ago, used to do political surveys over the phone (I was in college and needed a job that would let me work nights, sue me), I can honestly say nothing about that survey structure looks odd.
This question was a bit odd in that, as Mark's original post (quoting the other blog post, maybe?) mentioned, interracial marriage isn't an election issue in Mississippi right now. However, I don't think that renders the results suspect. If anything, the unexpectedness would likely make people more honest, as they don't have the opportunity to think about what they should answer, instead of answering with what they really think. That's a common tactic in social psychology research, in fact. As long as the question was asked clearly, there's no reason to suspect the results are skewed because of some surprise factor (which I've never heard used as an argument against a political survey).
According to these numbers, 29% of the voters in the 2008 Democratic primary in Mississippi were over 60. Those numbers are likely to be off a little, because 13% of the voters in that primary were self-identified Republicans. By comparison, 44% of the voters in the 2008 Republican primary in Missisippi were over 60 (and as futher contrast, only 3% of Republican primary voters self-identified as Democrats).
Yeah, I know that study. I believe I wrote it up once or twice, a few years ago, in fact. There are more recent studies on dissonance and politics as well (e.g., this one).
Still, neither of those studies, or the many others, test your empirical claim.
Eh, sample size issues are generally, and this is a general case, only a problem for people who don't do statistics or social science research. The size of the sample is pretty much irrelevant after a certain point (and that point is well below 400), so long as the sample is representative and certain methodoligical issues are avoided (e.g., when I used to do survey research, I always did a follow-up non-response survey, though they take time so political surveys rarely use them).
I should add that the pdf to which both you and Res State link does not contain the words "weighted sampling." It mentions "weighting," but doesn't give context, and looks like part of a boilerplate disclaimer. A quick review of other PPP releases show that this is true.
Just so it's clear, "weighted sampling" is distinct from weighting results. Weighted sampling select subgroups (in this case, likely age and race demographics) based on some criteria (in this case, population, whether that population be the population of Mississippi or Mississippi Republican primary voters). It's used in almost all polls that us samples, which is almost all polls.
By the way, speculating that people might have misheard the question is one of the more unique responses to a poll result that I've seen. It's little more than speculation. I'd love to see real reasons for doubting these results, because let's face it, the results are disturbing. However, you haven't provided any such reasons.
You'll both enjoy them. I immediately put With Fire and Sword on my favorite books list, and The Deluge may be even better than it.
When you're done, all three books have pretty good Polish movie versions. Sienkiewicz is pretty much the Polish literary royalty, sort of like Pushkin, Tolstoy, and Chekov in Russia. You can't make it through a Polish education without reading his books.
Well, that depends on what you mean by "public sector." In the last two years, and particularly in the last few months, state and municipal employees have had real issues with job security.
Butanyway. Ordinarily, engineering and technical workers accept a tradeoff when they sign up to work for the government. They know that, relative to the private sector, their pay is going to be poor. In compensation, they get fantastic job security.
If you add in retirement benefits (the back end stuff) and health care, you're pretty much talking about all government employees. Since, in addition to the drop in job security, states and perhaps the feds are going after pensions and benefits, there is going to be very little incentive for anyone who can get a job in the private sector to work in the public sector, soon.
*Comment archive for non-registered commenters assembled by email address as provided.
On “Revisiting the Mississippi Interracial Marriage Poll”
Eh, here’s a little project: count the number of gay couples on prime time television shows. Then count the number of black-white interracial couples on prime time television shows. Try it again with soap operas.
"
Oh, I don't doubt that they're highly correlated. I'm rather saying that the relationship is non-linear, particularly at the tails.
On “Do a Plurality of Mississippi Republicans Want to Ban Interracial Marriage?”
I think the Marx question would be pretty silly, since the vast majority of American Democrats have never read a word of Marx (though they may have heard something about "the opiate of the masses"), and are probably about as averse to the name as most American Republicans.
But I do think the racial questions are important because, while interracial marriage may not be up for debate, other issues related to race are, and it's not so bad to know how people stand on race. And interracial marriage is a pretty good weather vane.
On “Revisiting the Mississippi Interracial Marriage Poll”
It might help to read the other comments by that same employee, Dustin Ingalls. Relevant to the discussion we've been having, he notes that they asked the interracial marriage question of everyone, and that the questions in the report were not the only questions asked. This is puzzling (why didn't they include the other questions in this report? Are the results coming out in future reports?), and it may mean the survey was long, but it also answers one of the objections Mark and others have raised.
On the age issue, he also gives info on the margin of error for the young age group. This doesn't completely answer the concerns, but it helps a bit.
"
By the way, since a consistent theme from some commenters has been that this was a hit job, I thought this comment from the PPP blog by one of the PPP employees might be of interest:
BTW, we did ask this interracial marriage question of everyone, not just Republicans, but we'll be releasing the full results on that at a later date.
http://www.blogger.com/comment.g?blogID=2799451770086337664&postID=8659061201993601699
"
Eh, I'm not sure support for gay marriage and support for interracial marriage are the same thing. I also think it's interesting that, though it's within the margin of error, the highest % of respondents with favorable ratings of Palin was among the youngest age group, suggesting that they are in fact quite conservative.
"
Oh, I agree, but I don't intend it as the sole data point. Instead, it's another piece in a chain of evidence in a chain that supports the Mississippi poll, or at least suggests it's not an anomaly.
And let's be honest, I doubt anyone from the south is all that surprised. I mean, I've been to Mississippi, and while I love the state (I really do, and not just because it gave us this), racism is much more on the surface there than it is in much of the country in 2011 (parts of South Carolina, Mobile, and parts of Louisiana are the only places I've seen it so readily displayed in the last 15 years). You don't have to go looking for it, even among young people.
"
That a substantial percentage of African Americans believe that the U.S. government plays some role in the AIDS epidemic -- whether it created the disease, is withholding a cure, is using the disease against black people specifically, or a handful of other conspiracy theories -- is pretty well known, and completely irrelevant. But way to go with the, "I know you are but what am I" response.
"
James, ask, and ye shall receive:
http://www.gallup.com/poll/28417/most-americans-approve-interracial-marriages.aspx
It's not about a law, but it is about interracial marriage. And it's a few years old, but not old enough to not be relevant. Notice that nationwide, support for interracial marriage among whites over 50 is around 64%, suggesting that 36% either disapprove, don't know, or don't have an opinion. It's not unreasonable to expect that number to be higher in Mississippi or any of the former confederate states, and to be even higher among conservatives in that age group.
That poll doesn't help to explain, and perhaps even casts doubt on the youth results from the Mississippi poll, but since it's a southern state exclusively, and only polls conservatives, I don't think those results are outrageous.
Let me just add, before Heidegger writes 1200 words about how I must be a commie radical who's bombed some government office in the 60s because I think all conservatives are racist, that I don't, in fact, think all conservatives are racists. Or at least, I don't think all conservatives are any more racists than the average person (and I think to some degree everyone's a racist, including myself; the attitudes are too deeply ingrained in our culture for us not to be). In fact, I don't think most conservatives are any more racist than the average. But I do think you're likely to find more racism among conservatives, and in particular more overt racism. And of course, there's data to back me up on that.
"
I will say that the age issue (which has been discussed here already; good eye, Tom) is somewhat puzzling, but as people in that age group do tend to vote Democrat, it doesn't seem unreasonable to speculate that young Republicans tend to be more... let's say, fervent.
"
As Mark mentions in this post, there's independent evidence that suggests the poll isn't inaccurate.
By the way, if you look at political surveys over the phone, you'll often find policy questions thrown in like that. In general, the pollsters (and the people who hired them) want a quick look at an issue or two, and a bunch of information about potential voting patterns. In short polls like that (and it is short, even if it reads long -- probably took fewer than 5 minutes to complete), they usually put them near the end before the demographic stuff. It may lead to inaccuracies, but there's no inherent reason to think it does. As someone who, many years ago, used to do political surveys over the phone (I was in college and needed a job that would let me work nights, sue me), I can honestly say nothing about that survey structure looks odd.
This question was a bit odd in that, as Mark's original post (quoting the other blog post, maybe?) mentioned, interracial marriage isn't an election issue in Mississippi right now. However, I don't think that renders the results suspect. If anything, the unexpectedness would likely make people more honest, as they don't have the opportunity to think about what they should answer, instead of answering with what they really think. That's a common tactic in social psychology research, in fact. As long as the question was asked clearly, there's no reason to suspect the results are skewed because of some surprise factor (which I've never heard used as an argument against a political survey).
"
According to these numbers, 29% of the voters in the 2008 Democratic primary in Mississippi were over 60. Those numbers are likely to be off a little, because 13% of the voters in that primary were self-identified Republicans. By comparison, 44% of the voters in the 2008 Republican primary in Missisippi were over 60 (and as futher contrast, only 3% of Republican primary voters self-identified as Democrats).
On “How Responsible Are You for Where Your Taxes Go?”
Ugh, I'm in moderation because I linked to stuff (my stuff, which makes it a genuine offense).
Anyway, cool. Do you have any studies that test your empirical claim, though? The one I quoted in my comment above?
"
Yeah, I know that study. I believe I wrote it up once or twice, a few years ago, in fact. There are more recent studies on dissonance and politics as well (e.g., this one).
Still, neither of those studies, or the many others, test your empirical claim.
"
and it’s something that afflicts political partisans more than most people
That's an interesting empirical claim, and one that would be remarkably difficult to test. What makes you think that's the case?
On “Do a Plurality of Mississippi Republicans Want to Ban Interracial Marriage?”
I like Collin's questions, because they show his biases more than those of any respondent.
"
Eh, sample size issues are generally, and this is a general case, only a problem for people who don't do statistics or social science research. The size of the sample is pretty much irrelevant after a certain point (and that point is well below 400), so long as the sample is representative and certain methodoligical issues are avoided (e.g., when I used to do survey research, I always did a follow-up non-response survey, though they take time so political surveys rarely use them).
"
I should add that the pdf to which both you and Res State link does not contain the words "weighted sampling." It mentions "weighting," but doesn't give context, and looks like part of a boilerplate disclaimer. A quick review of other PPP releases show that this is true.
So where did Red State get "weighted sampling?"
"
This is a great comment, perfectly devoid of substance.
"
Just so it's clear, "weighted sampling" is distinct from weighting results. Weighted sampling select subgroups (in this case, likely age and race demographics) based on some criteria (in this case, population, whether that population be the population of Mississippi or Mississippi Republican primary voters). It's used in almost all polls that us samples, which is almost all polls.
By the way, speculating that people might have misheard the question is one of the more unique responses to a poll result that I've seen. It's little more than speculation. I'd love to see real reasons for doubting these results, because let's face it, the results are disturbing. However, you haven't provided any such reasons.
On “Epictetus, Freedom, and Autonomia”
I did think of starting one with the words: “Are the Democrats or the Republicans responsible for the budget crisis? Read to the end to find out!”
Then you could end the post with, "Yes."
"
I read it. And hated it!
No, I kid.
It did remind me of the title of one of Montaigne's essays, "That to Study Philosophy Is to Learn to Die."
On “A Confession of Bias, Followed by a Bunch of Stuff You Should Probably Ignore”
You'll both enjoy them. I immediately put With Fire and Sword on my favorite books list, and The Deluge may be even better than it.
When you're done, all three books have pretty good Polish movie versions. Sienkiewicz is pretty much the Polish literary royalty, sort of like Pushkin, Tolstoy, and Chekov in Russia. You can't make it through a Polish education without reading his books.
"
Well, that depends on what you mean by "public sector." In the last two years, and particularly in the last few months, state and municipal employees have had real issues with job security.
"
Butanyway. Ordinarily, engineering and technical workers accept a tradeoff when they sign up to work for the government. They know that, relative to the private sector, their pay is going to be poor. In compensation, they get fantastic job security.
If you add in retirement benefits (the back end stuff) and health care, you're pretty much talking about all government employees. Since, in addition to the drop in job security, states and perhaps the feds are going after pensions and benefits, there is going to be very little incentive for anyone who can get a job in the private sector to work in the public sector, soon.
*Comment archive for non-registered commenters assembled by email address as provided.