Drew’s Collection of Things to Read on the Interwebs

Tuesday, July 10th, 2012

Where can you find thousands of dollars being poured out a window, birthday parties, hidden cameras, blood-stealing laws, out-of-control bureaucrats and the guy who voiced Darth Vader? In these links:

  • A “Code Enforcement Official” in Columbia County, Georgia, broke into the house of a sleeping woman and yelled at her for not mowing her lawn.
  • In Tennessee, a new law required eight drivers to give blood samples against their will last week because they were suspected of drinking and driving.
  • Text messages sent to friends and family are not private, according to a recent Washington state court ruling.
  • If you want birthday cake and guns at the same place, head to Texas.
  • Political candidates are getting a little creepy in their efforts to record their opponents’ daily lives.
  • In case you missed it last week, a (particularly dopey) subgroup of the Occupy Seattle movement protested campaign finance laws by throwing $5,000 out of a hotel room window. The bills had statements stamped on them. Apparently the protesters were unaware that copies at Kinko’s are only nine cents.
  • A lot of the folks who like James Earl Jones‘ corny movies like The Sandlot, Patriot Games and Field of Dreams are probably Tea Party types, but he still thinks they’re racists, anyway

Now that was a thoughtful and enjoyable way to kill a few minutes, wasn’t it?

Drew Johnson


Digg it |  reddit |  del.icio.us |  Fark

65 Responses to “Drew’s Collection of Things to Read on the Interwebs”

  1. #1 |  Christopher Swing | 

    Other Sean, did you actually read anything about that survey? Because it looks like they responded to your variety of obfuscation:

    http://depts.washington.edu/uwiser/racepolitics.html

    Multivariate Analysis of Racial Resentment and Selected Civil Liberties Among Whites

    Since the public has become aware of the data, several people have come forward to challenge our initial findings, specifically, that supporters of the Tea Party appear racially intolerant. A principal charge, one not without intellectual merit, is that the observed relationship between support for the Tea Party and racial resentment is more about the relatively conservative politics of Tea Partiers than racism. Indeed, conservatives tend to believe in a small government, one that doesn’t do much to help people who, they believe, should make an effort to do for themselves. This is certainly a legitimate view; it’s one to which many Americans have adhered from the beginning of the Republic. In short, some of our critics charge that, instead of the racism we observe associated with support for the Tea Party, we’re merely observing Tea Partiers’ conservatism at work. In other words, support for the Tea Party, they suggest, is simply a proxy for conservatism.

    To address this issue, we turn to regression, a statistical technique that allows analysts to tease out how one variable affects another. This is important because it permits us to account for the presence of other variables that may also affect the outcome while isolating the impact of the effect of the variable of interest on the result. So, in this case, if support for the Tea Party is truly a proxy for conservatism, the relationship between racial resentment and support for the Tea Party should evaporate once we control for conservatism. Otherwise, there’s something else going on with support for the Tea Party; it’s not just conservatism. To make things a little easier, we combined all of the items (questions) that comprise racial resentment, making them into a scale.

    As the figure shows, even as we account for conservatism and partisanship, support for the Tea Party remains a valid predictor of racial resentment. We’re not saying that ideology isn’t important, because it is: as people become more conservative, it increases by 23 percent the chance that they’re racially resentful. Also, Democrats are 15 percent less likely than Republicans to be racially resentful. Even so, support for the Tea Party makes one 25 percent more likely to be racially resentful than those who don’t support the Tea Party.

    Similar results obtain for racial profiling and the ability for authorities to detain people without putting them on trial. Again, controlling for ideology (conservatism) and partisanship, support for the Tea Party increases the probability that individuals agree that it’s okay to “racially profile someone on account of their race or religion” by approximately 27 percent. Support for the Tea Party also increases the probability, by 28 percent, that the authorities should have ability to detain individuals without being charged, for as long as authorities like. Of course, in both cases, conservatism also matters: increasing the likelihood that people will agree with racial profiling and indefinite detention by 30 and 33 percent, respectively.

  2. #2 |  Other Sean | 

    Christopher,

    Just listen to yourself (or rather, listen to the source from which you pasted that).

    They say things like “after controlling for ideology…”, and they speak in terms of factors that increase by “23 percent the chance that [respondents] are racially resentful.” If only such things could be measured. If only such variables could be controlled. Then the cargo cult that is social scientism would actually be social science!

    Here’s just one example of what happened when they tried to quantify racism and measure ideology. The survey asked respondents if they would disagree with the following statement: “Over the past few years blacks have gotten less than they deserve.” If you disagreed, you were listed as “racially resentful”, presumably because you think blacks got as much or more than they deserve.

    So what you end up with is a massive equivocation:

    1) Racism (defined in common parlance as the systemic oppression of a group based on stereotypes attributed to all members of that group) is bad.

    2) Many members of the Tea Party oppose affirmative action, believe blacks are getting about what they deserve from politics in recent years, and are willing to allow racial and religious profiling at airports.

    3) Many members of the Tea Party are racist (defined as people who oppose affirmative action, believe blacks are getting about what they deserve from politics in recent years, and are willing to allow racial and religious profiling at airports).

    The definition of racism changed drastically between 1) and 3). Naturally the people who conducted that survey are aware that few people will ever read its methods description. They know the results will be mentioned in brief, and pasted around the web. They also know that a typical reader will swap in the common parlance definition of racism and thereby reach the conclusion that:

    4) Many members of the Tea Party are racist (defined as people who support the systemic oppression of a group based on stereotypes attributed to all members of that group).

    But the survey doesn’t show that at all. Nothing was published in those results to meet the burden of supporting the assertion in statement 4).

    Don’t you think that’s just a bit misleading?

  3. #3 |  Christopher Swing | 

    Other Sean:

    So basically you hadn’t read that before, and instead of admitting what you were wrong when you tried to bullshit us about knowing his methodology, you go on to demonstrate you just don’t understand statistics.

    Cool story, bro.

  4. #4 |  Other Sean | 

    Christopher,

    If I was mistaken in describing the survey’s methods, tell me where.

    I didn’t really make any points that hinged on an understanding of statistics, but if you think I’ve made an error in that respect, tell me where.

  5. #5 |  Christopher Swing | 

    Other Sean – You said:

    “If a self-identified Tea Partier said he disliked affirmative action in the survey, by that virtue alone he was counted as a specimen of racist influence in the movement.”

    Oops. It wasn’t just one thing. “To make things a little easier, we combined all of the items (questions) that comprise racial resentment, making them into a scale.” (this has links at the original.) Never mind the list of other questions at the end of the snippet I posted.

    What the fuck, OS? You think you can lie to us in your previous comment and we’ll forget all about it when you spin more bullshit after you’re called on it? And then demand to be shown where you were wrong when it’s right there to be read? GTFO.

  6. #6 |  Other Sean | 

    Christopher,

    Well…that’s about one notch better than accusing me of a spelling error. Of course I know the survey had multiple questions. I never suggested otherwise. If I thought there was only one question, I wouldn’t have been using the word “survey” all this time. As long as these comments already are, we can’t afford to go listing entire questionnaires, now can we?

    The phrase “by that virtue alone” means just what it said. If a person said he opposed affirmative action, that response was considered racist without need for further evidence on that point. Meaning: they did not ask respondents why they opposed affirmative action, they just asked if, and when a respondent said yes, the answer was deemed racist “by that virtue alone”. Meaning: no more than a yes answer was required for the response to count as racist.

    I suppose I could have begun that paragraph with the words “to name just one example…”, but I didn’t. I could have also ended the paragraph by saying “and a similar problem applies to the other questions as well”, but I didn’t, because …frankly no one could have mistaken my meaning unless he were engaged in a desperate search for some trivial error to derail the discussion away from its main line of travel. Which reminds me:

    All the survey questions in the “measuring racism” silo are subject to the same criticism I directed toward the affirmative action question. None of the survey questions managed to measure racism directly or with selectivity. They all hinged on the assumption that certain political positions are racist inherently (or, as one might say, racist “by that virtue alone”).

    All my key criticisms remain unanswered. You haven’t defended the study, you’ve just accused me of miscounting the number of questions, and I didn’t even do that.

    But if that’s what you’re down to, if you have no more than that to say, then why are you so enthusiastically committed to your position?

    I’m not some troll who wandered in off the street. If you have a proper defense, I’d love to hear it, and I’d give it my most careful attention. Sadly, it seems you are not interested in testing that promise.

  7. #7 |  c andrew | 

    Thank you, Other Sean.

  8. #8 |  Christopher Swing | 

    “Of course I know the survey had multiple questions. I never suggested otherwise.”

    No, you just said, AGAIN:

    “If a self-identified Tea Partier said he disliked affirmative action in the survey, by that virtue alone he was counted as a specimen of racist influence in the movement.”

    *You said only one question determined if someone was racist or not in the survey.*

    That’s a hell of a lot more than a spelling error.

    And now you’re just trying to change your answer; which is still discounted by the researcher’s response to it. *Because you still don’t understand statistical regression.* Which would answer your “criticism,” if you were smart enough to understand it.

  9. #9 |  Other Sean | 

    Christopher,

    I’m sure you’re familiar with the phrase “garbage in, garbage out”? If the questions in the survey fail to isolate the variable they’re supposed to measure, then it doesn’t matter what statistical methods you use after that. Your results will be meaningless.

    In this case, all the regression proved was that there are two types of what the authors call “conservatives”: A) those who stayed at home in 2010, and gave more moderate answers to the survey questions, and B) those who took to the streets as Tea Partiers in 2010, and gave less moderate answers.

    But who needed a survey to tell them that? Of course people who identified themselves with the Tea Party were going to give a more ideologically robust set of answers from those of mainstream conservatives. That’s the whole point of the Tea Party: it was a rebellion against the political center and the two party consensus that unites the Mitch McConnells with the Claire McCaskills.

    So when the researchers claim to have used “regression analysis to control for conservatism and party affiliation”, it boils down to that. They discovered an extremely banal fact that everyone already knows: Tea Partiers are often more outspoken in their opinions than Paleo or Neo Conservatives. No Shit?

    But really, I should stop using the term “researchers” at all, because that’s not an accurate descriptor. What we have here are a couple of guys who set out with their conclusion already in hand, and then added a veneer of pseudo-science to make it look like academic inquiry. What John Stewart says with humor, what Keith Olbermann says with solemnity, they are saying with tables and graphs. But there is nothing behind those tables and graphs save for a long list of insufficiently selective questions.

    I’ve probably been way too generous so far, in letting you pretend that survey was anything other than a political opinion piece parading around in the cloak of scholarly research.

    Let your mind wander back to the origins of this discussion. If Mr. Holding The Fire had simply said: “I think the Tea Party is racist, but damn, I can’t prove it”, that would have been one thing. I don’t happen to agree with that statement, but because the opinion is presented as an opinion, I would not feel compelled to reply.

    I don’t know you, but it seems very likely that you do believe the Tea Party is racist, for reasons best known to you. But why pretend this “research” is one of those reasons? We both know its not, because people just don’t form opinions that way. No one sits around, neutral, waiting for the “U.W. Institute of Race & Gender Studies” to tell them what to think about the Tea Party. They decide on their own, based on a priori considerations, and refine their opinion as it goes.

    They only turn to scientism when an argument breaks out, thinking “Aha! Now here I have a silver bullet. This survey, so carefully designed to RESEMBLE the type of hard science one might find in The Lancet, this will win the argument for me. This will make my opinions look, sound, and smell like the product of dispassionate research.”

    But why bother with all that? I’m sure you have some real reasons for thinking the Tea Party is racist, and I’m sure they are more interesting than that survey with its shoddy methods and its pre-ordained results. Why not just offer those reasons up for discussion, and see what happens?

  10. #10 |  Christopher Swing | 

    Other Sean: Do you really think throwing more words at it will make it better? Now you’re just making *excuses*. And attacking the source; and I’m afraid I’m still going to trust them more than random libertarian commenter on the internet. Is your Asperger’s flaring up so badly again that you can’t concede that maybe this study and other information might have something to them?

    Dude, let it go.

    Now (speaking of libertarian cluelessness) for real entertainment, we can turn from the tea people to this gem:

    https://www.youtube.com/watch?v=C_ngQgcRJxI

    Racist? No. Funny? YES. And it’s not so much funny because you have two middle-aged white guys haplessly trying to explain why there are so few non-white libertarians. It’s funny because they could have decided to make that video any time they wanted, with any preparation they wished. And yet they apparently couldn’t find any non-white libertarians to talk to or include in the video. Or hell, apparently non-white people *period.* XD

  11. #11 |  Other Sean | 

    Chris,

    “I’m still going to trust them more than random libertarian commenter on the internet.”

    Actually, you should distrust both equally. When it comes to information, the idea of “trust” always implies something dangerous: a claim that some people should be exempted from criticism or scrutiny because of who they are, where they work, how long they stayed in school, what jargon they use, etc.

    As for that video, I’ll tell you what I saw: a couple of guys coming to grips with a pretty obvious and unpleasant truth, and admitting that they can’t quite explain it (though Welch had an interesting theory).

    You know, the fact the 95% of black America votes with clockwork reliability for the Democratic party is not a good thing for black Americans. It gives them very little bargaining power in the electorate. They are roughly equivalent to the anti-abortion element on the Republican side. Candidates court them good and hard in the primaries, then hide them away during the general, and after the election they get nothing but rhetoric…because hey, where are they going to go? Meanwhile, white middle class swing voters and the elderly are getting paid off.

    Imagine what could have happened if politicians had to COMPETE for the black vote between 1968 and now. Do you really think things would still be this bad?

  12. #12 |  Christopher Swing | 

    Other Sean:

    “Actually, you should distrust both equally.”

    Hahahahahhahahahaa

    “As for that video, I’ll tell you what I saw: a couple of guys coming to grips with a pretty obvious and unpleasant truth, and admitting that they can’t quite explain it.”

    Of course, you don’t get it either, which is also funny. :D

  13. #13 |  Other Sean | 

    Christopher,

    It’s interesting that you think Gillespie and Welch should have dug up a token black libertarian for the sake of putting a false face on their movement. Why not applaud them for answering the question honestly? Since we’ve all rolled our eyes at it a thousand times, shouldn’t they both be commended for NOT resorting to the “some of my best friends…” argument?

    What they said about the Nader faction is true: almost no blacks to be found there. The same probably goes for the readership of this website. The Occupy Movement has a very small contingent of blacks, given what one would expect from its claims to represent the indebted and the unemployed. A few months back I went to an ACLU conference on the analysis of racial profiling data, and the only black person in a room with 35 people was a cop sent (no doubt with some thought for strategy) to speak for his department.

    I don’t know why it should be so, but you just don’t find significant numbers of black people taking part in ideological movements or politically driven fads.

    When a shop opens up in the inner city with a sign reading “Organic Fair Trade Coffee and Vegan Desserts”, everyone knows straight away that they are seeing a symptom of white blight and gentrification. The same goes for hybrid cars, bicycle sharing programs, you name it.

    If you want be fair, you need to apply that same criticism to all sides, because evidently libertarians, conservatives, and movement progressives are all alike in doing something that chronically alienates 16% of the American public.

    And it raises an interesting question, because no one seems to know quite what that something is.

  14. #14 |  Christopher Swing | 

    Other Sean:

    “It’s interesting that you think Gillespie and Welch should have dug up a token black libertarian for the sake of putting a false face on their movement.”

    Never said any such thing.

    Of course, you *still* don’t get it, but you can hardly be expected to.

  15. #15 |  Other Sean | 

    Chris,

    You got me there. I’ve forgotten how to parry the “you just don’t get it” attack. It’s been so many years since middle school…