By the Numbers
By the Numbers
©2016 Ross Williams
Sociology is
a soft science. Similar to psychology, economics and
history. Hard Sciences are biology, chemistry, physics and geology. The hard sciences are "hard" due,
according to Carl Sagan, to the "rigorous standards of evidence and
honesty." In other words, when a
physicist adds 2 and 2 and arrives at 22 he is wrong, and will be ever after
considered a quack. Unless, of course,
he can sell his 2+2=22 idea to Hollywood and have them place the notion of nuclear winter in the collective psyche
of the masses, in which case he'll be forever seen as a visionary.
Other such visionaries are the neo-malthusian Paul Ehrlich, the pre-neo-luddite Rachel Carson, and the doubly ethically blind Stephen Schneider. Being proven wrong is not necessarily an impediment in the hard sciences.
In the soft sciences, however, being proven wrong is typically requisite for the soft scientist to attain and maintain professional credibility. Just ask Paul Krugman, acolyte of the also-proven wrong John Maynard Keynes. There's always an excuse for why being wrong was actually right, usually having to do with the entire world refusing to accept the soft scientist's egoistic declarations and discounting reality.
And vice versa, for what it's worth. Being right often ends up wrong. The "stop and frisk" model of civil order grew out of the "broken window theory" of economics. If police want to reduce future major crime before it happens, they should address more thoroughly the minor crimes as they occur: the broken windows and graffiti. New York City implemented this notion and major crime was indeed reduced. Politicians wanted to reduce major crime even more, so they implemented further and further tweaks to their policies, incrementally reducing major crime along the way, and voila! they arrived at stop and frisk. Never mind that it operates at right angles to the legitimate power of government in this constitutional republic, and imposed a police-state. Right became wrong.
Other such visionaries are the neo-malthusian Paul Ehrlich, the pre-neo-luddite Rachel Carson, and the doubly ethically blind Stephen Schneider. Being proven wrong is not necessarily an impediment in the hard sciences.
In the soft sciences, however, being proven wrong is typically requisite for the soft scientist to attain and maintain professional credibility. Just ask Paul Krugman, acolyte of the also-proven wrong John Maynard Keynes. There's always an excuse for why being wrong was actually right, usually having to do with the entire world refusing to accept the soft scientist's egoistic declarations and discounting reality.
And vice versa, for what it's worth. Being right often ends up wrong. The "stop and frisk" model of civil order grew out of the "broken window theory" of economics. If police want to reduce future major crime before it happens, they should address more thoroughly the minor crimes as they occur: the broken windows and graffiti. New York City implemented this notion and major crime was indeed reduced. Politicians wanted to reduce major crime even more, so they implemented further and further tweaks to their policies, incrementally reducing major crime along the way, and voila! they arrived at stop and frisk. Never mind that it operates at right angles to the legitimate power of government in this constitutional republic, and imposed a police-state. Right became wrong.
Right becoming
wrong played a significant role in this last election cycle, as well. Donald Trump won, as I knew — and said [ref:
"None of the Above"] — he would.
A number of others including, interestingly, the moron Michael Moore,
"had a feeling", and several trend-watchers predicted the outcome
based on "enthusiasm" and other non-quantifiables. But the pollsters, almost to a bean-counter,
missed and they missed wide left.
This makes the second time in a year that, worldwide, pollsters got reproved by the reality they don't fully understand how to measure. European pollsters declared that British voters would vote overwhelmingly to stay in the European Union; they did not … overwhelmingly. And US pollsters declared that Hillary Clinton would occupy the White House with ex-President and First Philander Bill. And they shall not. Considering where the polls predicted we would end up and the reality that arrived in its place, the pollsters missed by a significant double-digit margin.
I previously hinted at why and how this occurred: a significant portion of those who ended up voting for Trump in the primaries were discounted by pollsters as "not likely to vote". That was the biggest reason that Trump came in ahead of the standard republican party selections in defiance of the polls.
Opinion research polling is a major sub-discipline in the field of sociology; sociology is what I did in grad school. Analyzing and quantifying the results of polls was a large part of the practical course work involved; the impractical course work was mainly given over to learning how to construct arguments as to why reality is wrong when it differs from sociological theory. Despite refusing — outspokenly — to supplant reality with sociological theory, I maintained a straight-A GPA.
Among the knowledge I carried away from grad school was how pollsters determine who is likely to vote and who isn't. If anyone has answered the phone from August to early November in an election year and answered twenty minutes of questions from a bored mumble-mouth making less than minimum wage, you will probably recall that after all the questions relating to the candidates, the issues, and their relative importance to you, you were given a final series of "…and now for statistical purposes" questions.
This makes the second time in a year that, worldwide, pollsters got reproved by the reality they don't fully understand how to measure. European pollsters declared that British voters would vote overwhelmingly to stay in the European Union; they did not … overwhelmingly. And US pollsters declared that Hillary Clinton would occupy the White House with ex-President and First Philander Bill. And they shall not. Considering where the polls predicted we would end up and the reality that arrived in its place, the pollsters missed by a significant double-digit margin.
I previously hinted at why and how this occurred: a significant portion of those who ended up voting for Trump in the primaries were discounted by pollsters as "not likely to vote". That was the biggest reason that Trump came in ahead of the standard republican party selections in defiance of the polls.
Opinion research polling is a major sub-discipline in the field of sociology; sociology is what I did in grad school. Analyzing and quantifying the results of polls was a large part of the practical course work involved; the impractical course work was mainly given over to learning how to construct arguments as to why reality is wrong when it differs from sociological theory. Despite refusing — outspokenly — to supplant reality with sociological theory, I maintained a straight-A GPA.
Among the knowledge I carried away from grad school was how pollsters determine who is likely to vote and who isn't. If anyone has answered the phone from August to early November in an election year and answered twenty minutes of questions from a bored mumble-mouth making less than minimum wage, you will probably recall that after all the questions relating to the candidates, the issues, and their relative importance to you, you were given a final series of "…and now for statistical purposes" questions.
The answers
to these are used to calculate whether the pollster will throw away your
responses or add them to the final tally they publish. For better or worse, those over the age of
30-35, whites, males, middle class who have a history of consistently voting
are going to be most likely to vote this time.
Those who are younger, or non-white, or female, or poverty class, or super-rich,
or do not consistently vote, you are less likely to vote this time. Furthermore, depending on several topical
factors, being democrat or republican alters the equation, and independents are
always less likely to vote. How much less likely is the relevant
question.
Donald Trump
consistently outperformed his polls in the primaries. And I explained why. The republican party has operated for the
last generation in a manner dedicated to spurning a large and growing portion
of its historic voter base — the fiscal conservatives — in favor of the
minority ideologue religious tight-asses.
Fiscal conservancy deplores "free trade" bargains that benefit
other nations and ends up selling US jobs to second- and third-world countries;
allowing foreigners to come to the US illegally and siphon any amount of US government largesse provided by taxpayers or [most
likely] China, who will one day demand to collect on our debt to it; and tax
policies that incentivize US companies to move to second- and third-world
countries so they might, then, be able to profit from our "free
trade" agreements that screw the American pooch.
These issues are what Trump spoke to, and the fiscal conservative republicans who vacated the republican party since the mid-80s heard him and came back in the primaries to vote for him. This ended, sometimes, decades of voting drought. But the pollsters, upon learning that the guy they were polling hadn't voted since 1992, or even 2000, threw away the response which read, "I am voting for Trump because he wants to allow me to get my decent job back; he wants to stop selling my job to India; he wants to stop my company from moving to Mexico." Trump voters were underrepresented in the results of the polls − by design, but not out of malice or political imperative. And because large wads of inconsistent voters voted in the primaries, Trump outperformed his numbers.
These issues are what Trump spoke to, and the fiscal conservative republicans who vacated the republican party since the mid-80s heard him and came back in the primaries to vote for him. This ended, sometimes, decades of voting drought. But the pollsters, upon learning that the guy they were polling hadn't voted since 1992, or even 2000, threw away the response which read, "I am voting for Trump because he wants to allow me to get my decent job back; he wants to stop selling my job to India; he wants to stop my company from moving to Mexico." Trump voters were underrepresented in the results of the polls − by design, but not out of malice or political imperative. And because large wads of inconsistent voters voted in the primaries, Trump outperformed his numbers.
Trump
outperformed his numbers in the general election as well, and inconsistent
voters voting were a small part of the reason for it. The bigger, and probably the biggest, reason is, like the Brexit vote in
the UK, explained by one of the undergrad exercises Sociology Departments undertake
at — I've got to believe — every college and university in the world.
I wasn't in the undergrad program so I never got to participate in the exercise, but I got to hear all about it in my graduate seminars when we discussed the pitfalls and pratfalls of public opinion measurement. During one of the many university festivals taking place each year, the Sociology Department set up — under the cover of a fictitious ethnic or cultural diversity program — a food stand that gave away an unlikely sample of street food from a foreign culture. In the exercise that I got to hear about, the street food was Laotian or Cambodian, and they gave away two samples of ethnic food: a mini pork kabob, or a mini dog kabob. Because it was free, each customer was allowed to pick just one.
Those who lined up for the free food were funneled in at one end, and had to funnel out at the other. In between, they were regaled by carefully choreographed protests of animal cruelty staged by other members of the Sociology Department who were not working the kabob-stand. Sometimes the staged protests were joined by other students, unaware that it was a con. And of course the University administration — well aware of the entire production — received many complaints about the group of animal abusers serving dog meat. The University would dismiss them, saying "Thank you for bringing this to our attention, we will look into it."
But there was nothing to look into; it was all pork grilled on a hibachi. It didn't matter if you asked for the free pork kabob or the free dog kabob: you got pork. Those giving away the kabobs kept track of how many of each they handed out. It was after you got your free food and you were funneling out the far end of the free food stand where the sociological experiment reached its conclusion. You were asked one question and one question only, right in front of the carefully choreographed protesters: did you get the pork or the dog? …and you had to answer the question or else they'd hold you up and not let you out.
Ninety-five percent of those who went in for free food reported getting the pork.
However, twenty-five percent of the food given away was dog.
…which means that four out of five people getting the free dog kabob lied about it.
Why would people lie? Apart from having masses of presumptuous ideologues screaming in their ears about how horrible they are for the choices they make, even if the screaming is staged … I can't think of a single reason.
But that brings us back to the inaccurate polls and the Trump victory: being told you're racist when all you want is to not lose your job to China; being told you're a homophobe when you don't want to have your tax dollars being handed — in any amount — to people who refuse to follow immigration laws in coming here when you do your damnedest to be law-abiding; being told you're islamophobic when you don't want your company to move to Mexico and take your job with it; being informed by the smug and sanctimonious that you are deplorable and bigoted for wanting to provide for yourself and your family and be left alone to do it … is not simply presumptuous, factually insupportable, and serves no purpose apart from stoking the self-righteous egos of the faux-pious who make those assertions. It gets in the way of understanding what people in a democratic society want, need and expect from that democratic society, and the government which runs it … as measured by the pollsters who try to find it out.
If a democratic society refuses to understand what its people want, need and expect because sneering contempt and demonizing denunciation masks those wants, needs and expectations, it will cease being democratic. Crack a history book.
I wasn't in the undergrad program so I never got to participate in the exercise, but I got to hear all about it in my graduate seminars when we discussed the pitfalls and pratfalls of public opinion measurement. During one of the many university festivals taking place each year, the Sociology Department set up — under the cover of a fictitious ethnic or cultural diversity program — a food stand that gave away an unlikely sample of street food from a foreign culture. In the exercise that I got to hear about, the street food was Laotian or Cambodian, and they gave away two samples of ethnic food: a mini pork kabob, or a mini dog kabob. Because it was free, each customer was allowed to pick just one.
Those who lined up for the free food were funneled in at one end, and had to funnel out at the other. In between, they were regaled by carefully choreographed protests of animal cruelty staged by other members of the Sociology Department who were not working the kabob-stand. Sometimes the staged protests were joined by other students, unaware that it was a con. And of course the University administration — well aware of the entire production — received many complaints about the group of animal abusers serving dog meat. The University would dismiss them, saying "Thank you for bringing this to our attention, we will look into it."
But there was nothing to look into; it was all pork grilled on a hibachi. It didn't matter if you asked for the free pork kabob or the free dog kabob: you got pork. Those giving away the kabobs kept track of how many of each they handed out. It was after you got your free food and you were funneling out the far end of the free food stand where the sociological experiment reached its conclusion. You were asked one question and one question only, right in front of the carefully choreographed protesters: did you get the pork or the dog? …and you had to answer the question or else they'd hold you up and not let you out.
Ninety-five percent of those who went in for free food reported getting the pork.
However, twenty-five percent of the food given away was dog.
…which means that four out of five people getting the free dog kabob lied about it.
Why would people lie? Apart from having masses of presumptuous ideologues screaming in their ears about how horrible they are for the choices they make, even if the screaming is staged … I can't think of a single reason.
But that brings us back to the inaccurate polls and the Trump victory: being told you're racist when all you want is to not lose your job to China; being told you're a homophobe when you don't want to have your tax dollars being handed — in any amount — to people who refuse to follow immigration laws in coming here when you do your damnedest to be law-abiding; being told you're islamophobic when you don't want your company to move to Mexico and take your job with it; being informed by the smug and sanctimonious that you are deplorable and bigoted for wanting to provide for yourself and your family and be left alone to do it … is not simply presumptuous, factually insupportable, and serves no purpose apart from stoking the self-righteous egos of the faux-pious who make those assertions. It gets in the way of understanding what people in a democratic society want, need and expect from that democratic society, and the government which runs it … as measured by the pollsters who try to find it out.
If a democratic society refuses to understand what its people want, need and expect because sneering contempt and demonizing denunciation masks those wants, needs and expectations, it will cease being democratic. Crack a history book.