In (some) defense of Doug Schoen

For those who missed it last week, Doug Schoen released a fascinating poll consisting of in-person interviews of n=200 Occupy Wall Street protesters.  The topline report can be found here and is worth reading in its entirety.  It provides a glimpse into some of the real lunatic fringe elements that are part of (but not the whole of) the Occupy Wall Street protests.

That “part of” bit is important, because Schoen took a little beating over imprecise language in his discussion of the poll and because of one of those silly errors that are the bane of running a real-world polling operation.  He took a beating mostly for disagreeing with liberals, but he didn’t help his own cause.

You see, Schoen made the all-too-rational argument that the Occupy Wall Street protesters might not be the kind of gang that Democratic leaders want to get too close to.  To a conservative, that’s about as controversial as saying that you don’t want to go throw trashcans through windows at an anti-IMF protest.  But the left wing of the Democratic Party is so desperate to anoint Occupy Wall Street as its own version of the Tea Party movement that they treated Schoen as an apostate.

For a summary of the anti-Schoen position, take a look at Andrew Gelman’s very subtly titled piece, “Hack pollster Doug Schoen illustrates a general point: The #1 way to lie with statistics is…to just lie!”  Gelman is a smart guy and his blog is a regular read in our firm, so I’m not singling him out here—the same arguments and rancor appeared elsewhere across the progressive blog universe.

The complaints about Schoen are basically three:

1) Schoen says that the Occupy Wall Street protest “believes in radical redistribution of wealth, civil disobedience and, in some instances, violence…”

To “debunk” this, the liberals go to an open-ended question about what protesters want to achieve.  They point out that influencing the Democratic Party the way that the Tea Party has the GOP (35%) is the number one objective of the protesters and only one other objective (breaking the two-party duopoly, 11%) even scores in double digits.  They further point out that only 4% want to dissolve the American system of government or the capitalist economic system and only another 4% want “radical re-distribution of wealth.”

The first point is fair enough, slightly more than one-third of the protesters may be fairly reasonable people who just want to pull the Democratic Party to the left.  If you add in the 9% who say they want to engage and mobilize progressives, it’s fair to grant that a plurality of the protesters are just run-of-the-mine activists who want to influence their Party and get more liberal Democrats elected.

But what about those couple of “only 4%” groups in there?  Here are the actual numbers:

Seven percent (7%) of protesters want “direct democracy,” which probably is fine if you’re sitting around your freshman seminar daydreaming about the tempeh burger you’re going to eat later and the game of hacky sack you’ve got on for the afternoon.  But in the really real world that means…oh I don’t know…something like overthrowing the American system of government.

Then there’s that 4% who just outright say they want to dissolve representative democracy and the capitalist system.  To translate that from progressive-speak to English, that’s Marxist revolution or something quite like it.

And there’s that other 4% who want to radically redistribute wealth.  I’m not sure if that’s Marxism or just looting, but it doesn’t sound good.

Plus, as anyone who has worked with re-coded open ended data before knows, there are probably a handful of responses that couldn’t be coded together with anything else but were likely just as crazy.

So one in six (15%) protesters want to overthrow the American system of government, the American economy, or the notion of private property.  Plus there are an unknown number of other kooks hiding in the un-coded data.  That’s not exactly a mainstream movement anywhere other than perhaps Berkeley or Amherst.

Then there’s that violence point.

According to Schoen’s data, almost one in three (31%) protesters said they would engage in violence to achieve their goals.  I don’t have the crosstabs, but I’m guessing the ones who want to influence the Democratic Party, promote a national conversation, or energize Progressives aren’t the ones who said they would use violence.

So what we’re talking about here is a not-insignificant number of these protesters (somewhere between one-sixth and one-third) who are advocating the violent overthrow of the United States Government, the violent disruption of the economic system, or the violent theft of private property.  There are words we usually use for all of those things and none of them are good and none of them are at all mainstream.

And does anyone else think this might have received a bit more attention from the  Mainstream Media if between one-sixth and one-third of Tea Party activists wanted to overthrow the government?

2) Schoen said in writing and on TV that “Sixty-five percent {of protesters} say that government has a moral responsibility to guarantee all citizens access to affordable health care, a college education, and a secure retirement—no matter the cost.”

A liberal blogger with access to the data posted the topline report showing that the question that was actually asked doesn’t include the “no matter the cost” clause and so Schoen was misleading people.  The liberal claim is that Schoen was lying about his data here and making the sentiment seem more extreme than it was.

Here’s the problem—if you look at the topline report posted on Schoen’s own website (and linked to above), it has the “no matter the cost” language.  So we’re faced with a he-said/he-said situation.  Either the liberal blogger altered the report to embarrass someone he disagrees with, Schoen altered his own report to align with his analysis, or something else happened.

We probably will never know the truth. But rather than someone outright lying here, I suspect what really happened is that the version of the topline report leaked to the blogs had an error.  Quality control in a polling firm that is “putting up big numbers” (running a lot of projects at the same time) is a constant effort.

From time to time things will get put into pre-production that are based on an earlier version of a much-edited survey instrument or that have a silly typo in them and get caught before final release.  From time-to-time one even gets sent out with an error in it—I recall a humorous CBS News poll from 2008 that repeatedly referred to Hillary Clinton with male pronouns.

If the leaked version of the poll wasn’t the final release, it wouldn’t shock me if it had question wording that wasn’t the final wording.  The fact that Schoen’s written analysis, TV analysis, and actual topline use precisely identical language suggests to me that the version actually fielded included the “no matter the cost” language and that the liberal blogosphere is pouncing on what amounts to a mid-stream quality control hiccup as evidence of malfeasance.

3) Schoen says that the protesters support raising taxes on the wealthiest Americans.

Here the liberal bloggers simply point out that so do a large number of Americans in some surveys.

This is a fair point.  In the current environment, with concerns about an exploding budget deficit and major stresses on their own pocket books most Americans are willing to support the notion of “tax the other guys” when asked what solutions they do or do not support.  That’s a little different than just wanting to tax the rich more on general principle, but it’s a fair argument that the context would change the interpretation of the Schoen poll somewhat.

But, it’s also a quibble over interpretation.

Without the first two complaints to support it, a failure to put this question in context with other data hardly amounts to being a “hack pollster” or “just lie!”ing.

It seems to me that after looking at the evidence, the worst that Doug Schoen is guilty of in this case is running a slightly leaky ship with some mid-stream quality control issues and failure to contextualize and caveat every single piece of data in the way a professional academic would (but which also takes weeks longer to produce—time Schoen didn’t have here).

Hardly seems worth the war chant of “lies, lies, lies” rising up from the liberal blogosphere.  But then again, it’s not my fantasy of a grass-roots progressive movement to equal the authentic wave that was the Tea Party that he’s debunking.

Get Alerts