Will the Real Public Opinion on ObamaCare Please Stand Up

In the battle over ObamaCare the president's supporters point to numerous polls showing majorities endorsing stronger, more inclusive government-subsidized healthcare.[1] Opponents meanwhile stress Town Hall meetings in which hundreds of otherwise docile, mild-mannered Americans noisily and sometimes belligerently confront hapless legislators.  For ObamaCare fans, this explosive fury is far less "public opinion" than manufactured angry mobs ("Astroturf") who, allegedly, resemble Nazi storm troopers or the KKK.

Merits of ObamaCare aside, the debate over the "real public opinion" reflects two very different understandings of "public opinion" and this distinction goes a long way in explaining the vitriolic frustration felt by many opponents. 

Some history: "Public opinion" long predates modern surveys and while academics and media pundits deem modern scientific polling the gold standard, this "new and improved" instrument hardly supersedes older manifestations of popular sentiment. The older popular outcries were self-initiated; views were not solicited by paid interviewers with somebody else's questionnaire. Vox populi thus required a modest threshold of prior engagement -- those who did not pay taxes were far less inclined to riot over a tax hike than those on the hook. Second, expressing one's views was more forceful than just a "yes/no" verbal endorsement, so absent the pollster who gathered the data from passive respondents, public opinion was manifested in boisterous rallies, vandalism, or even dumping tea into the harbor. The "public" in this "public opinion" thus differed profoundly from those receiving unexpected phone calls.  Third, and critically, the spontaneous behavioral nature of traditional public opinion could prove troublesome. No member of Congress is terrified by bad poll numbers from a random sample though he or she certainly prefers upbeat numbers. But, a Roman Senator informed that public opinion now opposes an ill-fated military campaign might go into hiding. No wonder these Senators respected vox populi-it was nearly synonymous with the angry mob.

Modern scientific polling can thus manage, even manipulate, a potentially bothersome public, at least for those who can afford costly surveys. Whereas traditional public opinion was events driven, its modern incarnation can very well create, not just measure, political reality. An ambitious billionaire can now "prove" that millions of Americans believe that George W. Bush is a war criminal though not one in 10,000 ever thought about this before and not one in a 100,000 could accurately define "war criminal." Now, in an instant, ephemera enter our national consciousness. And if the numbers disappoint, highlight only the good news or just don't release anything since all poll data are sponsor owned. Nor must the billionaire put his or her name on the poll. 

Thus understood, "public opinion" via the modern, allegedly scientific survey is not the spontaneous outpouring of public concern; it is the result of an expensive, technically sophisticated enterprise that is "democratic" only in the sense that anybody with the money can pay for a poll. Vox populi humbug; vox prodigus is more like it.

The pollster's manipulative power goes well beyond timing and agenda shaping. Question writing is highly flexible and even the most professional, non-partisan questionnaire drafters enjoy ample leeway here.  This is not about "loaded" questions or cherry picking respondents; it is all about above board word usage, context, proffered alternatives and all else that goes into the sausage-making. Questionnaire writers enjoy god-like power in supplying information, and since realism adds expense and reduces all-important completion rates, simplify everything. To appreciate this immense power, try "being sophisticated" when telephoned for "one's opinion." You will be told to just answer the questions as presented and if you refuse, your erudite responses will be relegated to a meaningless "other" category or, more likely, an uncooperative attitude will end the interview. So much for an unfiltered "voice of the people."  

This god-like power grows dramatically since randomly-generated phone polls deliver respondents by the truckload scarcely interested in pollster-determined issues. The involuntary conscripts are far more likely to be influenced by the "minor" details of the sponsor-controlled survey, for example, linking a particular policy with a popular incumbent president. Not so for those with a stake in the matter-they stick to their guns regardless. Imagine those under 40 glibly responding to a query about taxing social security benefits versus the same questions asked to retirees. Not even the most skilled pollster could manipulate the codgers while indifferent younger respondents might be easily swayed by including terms like "pay their fair share." The national random sample inevitably provides ample clay to be molded by those crunching the numbers.  This is not craven dishonesty; it just comes with all surveys.       

Consider, for example, a July 2009 Kaiser Health Tracking Poll that asked about expanding the number of Americans covered by health insurance (http://www.kff.org/kaiserpolls/healthpoll.cfm). The option given was creating a government administered public health insurance option similar to Medicare to compete with private health insurance plans. It was strongly favored by 26%, favored somewhat by 33%, a clear (59%) endorsement of ObamaCare. To demonstrate just how this typical questions is riddled with assumptions and missing information, envision a knowledgeable respondent handling this item. He might ask for more information about who, exactly, lacks insurance coverage and what proportion is voluntary or just reflects non-medical priorities. Or even how many are illegal aliens? He will certainly inquire about projected costs, how they are calculated, and how this "free" government help is to be funded. And what about that word "compete"? Will government under-cut private firms ("making medical care more affordable") and thus drive them into bankruptcy? Will voluntary non-participants still have to pay taxes for those enrolled? Who are these "Americans"? Will illegal immigrants and non-citizens be eligible, and how will government administer this sorting if they show up at the local ER?  No doubt, the typical telephone interviewer will be totally overwhelmed by our well-informed, sophisticated respondent, especially since the Procrustean Bed questionnaire lacks adequate space for non-scripted respondent input. The interviewer (who may be paid by the completed interview) will probably hang up and welcome the next randomly dialed respondent who sheepishly accepts the questionnaire designer's over-simplified policy framework. In a sense, polling is closer to ventriloquism than giving voice to what thoughtful people actually think.   

No inherent ideological right/left bias infuses opinion polling-after all, anybody with the money can sponsor a poll to advance their views. But, with all too few exceptions, today's public policy polls reflect ideological left's worldview, and the liberal home field advantage is so ingrained that it is hardly noticed.[2] Questionnaires writers honestly see themselves as data collectors akin to meteorologists and it may never occur to them that government run clinics and hospitals may be bureaucratic boondoggles staffed with lazy incompetents. Perhaps the most telling feature of these liberal infused polls is the near total neglect of costs, including non-financial ones. Polls are overwhelmingly about free lunches.[3]

It is no wonder, then, that our respondent frustrated by the standard-issue health care poll must resort to non-poll means to make his voice heard. He could respond to hundreds of the poll questions concocted by liberal academics and fellow traveler pollsters and never, never get a chance to speak his mind. Rhetoric about democracy and polling aside, the game is rigged and having to watch a parade of carefully manufactured poll results only invites exasperation. Recall those unfortunate souls under communism required to sit for hours listening to party functionaries extolling the Workers' Paradise. And if there was time left over for questions, better not ask the wrong ones or else one would soon be "missing data." Taking to the streets or refusing to work did express an opinion, however. 

[1] The best source of public opinion data on health care is the Kaiser Family Foundation website that contains some 85,000 health care poll questions (http://www.kff.org/kaiserpolls/healthpoll.cfm).

[2] Though many conservative organizations run pro-free market polls, their structure is indistinguishable from polls run by liberal organizations. Like liberals, they seldom ask about willingness to bare costs, trade-offs or similar downsides. This probably reflects "conservative" pollsters learning their trade at liberal-dominated universities social science departments. On a personal note, I've sometimes offered free of charge to correct this build-in ideological bias, but to no avail. 

[3] I describe this free-lunch bias in my Polling, Policy and Public Opinion: Why the "The Voice of the People" Should Not Be Heeded New York: Palgrave-Macmillan. 2002.
In the battle over ObamaCare the president's supporters point to numerous polls showing majorities endorsing stronger, more inclusive government-subsidized healthcare.[1] Opponents meanwhile stress Town Hall meetings in which hundreds of otherwise docile, mild-mannered Americans noisily and sometimes belligerently confront hapless legislators.  For ObamaCare fans, this explosive fury is far less "public opinion" than manufactured angry mobs ("Astroturf") who, allegedly, resemble Nazi storm troopers or the KKK.

Merits of ObamaCare aside, the debate over the "real public opinion" reflects two very different understandings of "public opinion" and this distinction goes a long way in explaining the vitriolic frustration felt by many opponents. 

Some history: "Public opinion" long predates modern surveys and while academics and media pundits deem modern scientific polling the gold standard, this "new and improved" instrument hardly supersedes older manifestations of popular sentiment. The older popular outcries were self-initiated; views were not solicited by paid interviewers with somebody else's questionnaire. Vox populi thus required a modest threshold of prior engagement -- those who did not pay taxes were far less inclined to riot over a tax hike than those on the hook. Second, expressing one's views was more forceful than just a "yes/no" verbal endorsement, so absent the pollster who gathered the data from passive respondents, public opinion was manifested in boisterous rallies, vandalism, or even dumping tea into the harbor. The "public" in this "public opinion" thus differed profoundly from those receiving unexpected phone calls.  Third, and critically, the spontaneous behavioral nature of traditional public opinion could prove troublesome. No member of Congress is terrified by bad poll numbers from a random sample though he or she certainly prefers upbeat numbers. But, a Roman Senator informed that public opinion now opposes an ill-fated military campaign might go into hiding. No wonder these Senators respected vox populi-it was nearly synonymous with the angry mob.

Modern scientific polling can thus manage, even manipulate, a potentially bothersome public, at least for those who can afford costly surveys. Whereas traditional public opinion was events driven, its modern incarnation can very well create, not just measure, political reality. An ambitious billionaire can now "prove" that millions of Americans believe that George W. Bush is a war criminal though not one in 10,000 ever thought about this before and not one in a 100,000 could accurately define "war criminal." Now, in an instant, ephemera enter our national consciousness. And if the numbers disappoint, highlight only the good news or just don't release anything since all poll data are sponsor owned. Nor must the billionaire put his or her name on the poll. 

Thus understood, "public opinion" via the modern, allegedly scientific survey is not the spontaneous outpouring of public concern; it is the result of an expensive, technically sophisticated enterprise that is "democratic" only in the sense that anybody with the money can pay for a poll. Vox populi humbug; vox prodigus is more like it.

The pollster's manipulative power goes well beyond timing and agenda shaping. Question writing is highly flexible and even the most professional, non-partisan questionnaire drafters enjoy ample leeway here.  This is not about "loaded" questions or cherry picking respondents; it is all about above board word usage, context, proffered alternatives and all else that goes into the sausage-making. Questionnaire writers enjoy god-like power in supplying information, and since realism adds expense and reduces all-important completion rates, simplify everything. To appreciate this immense power, try "being sophisticated" when telephoned for "one's opinion." You will be told to just answer the questions as presented and if you refuse, your erudite responses will be relegated to a meaningless "other" category or, more likely, an uncooperative attitude will end the interview. So much for an unfiltered "voice of the people."  

This god-like power grows dramatically since randomly-generated phone polls deliver respondents by the truckload scarcely interested in pollster-determined issues. The involuntary conscripts are far more likely to be influenced by the "minor" details of the sponsor-controlled survey, for example, linking a particular policy with a popular incumbent president. Not so for those with a stake in the matter-they stick to their guns regardless. Imagine those under 40 glibly responding to a query about taxing social security benefits versus the same questions asked to retirees. Not even the most skilled pollster could manipulate the codgers while indifferent younger respondents might be easily swayed by including terms like "pay their fair share." The national random sample inevitably provides ample clay to be molded by those crunching the numbers.  This is not craven dishonesty; it just comes with all surveys.       

Consider, for example, a July 2009 Kaiser Health Tracking Poll that asked about expanding the number of Americans covered by health insurance (http://www.kff.org/kaiserpolls/healthpoll.cfm). The option given was creating a government administered public health insurance option similar to Medicare to compete with private health insurance plans. It was strongly favored by 26%, favored somewhat by 33%, a clear (59%) endorsement of ObamaCare. To demonstrate just how this typical questions is riddled with assumptions and missing information, envision a knowledgeable respondent handling this item. He might ask for more information about who, exactly, lacks insurance coverage and what proportion is voluntary or just reflects non-medical priorities. Or even how many are illegal aliens? He will certainly inquire about projected costs, how they are calculated, and how this "free" government help is to be funded. And what about that word "compete"? Will government under-cut private firms ("making medical care more affordable") and thus drive them into bankruptcy? Will voluntary non-participants still have to pay taxes for those enrolled? Who are these "Americans"? Will illegal immigrants and non-citizens be eligible, and how will government administer this sorting if they show up at the local ER?  No doubt, the typical telephone interviewer will be totally overwhelmed by our well-informed, sophisticated respondent, especially since the Procrustean Bed questionnaire lacks adequate space for non-scripted respondent input. The interviewer (who may be paid by the completed interview) will probably hang up and welcome the next randomly dialed respondent who sheepishly accepts the questionnaire designer's over-simplified policy framework. In a sense, polling is closer to ventriloquism than giving voice to what thoughtful people actually think.   

No inherent ideological right/left bias infuses opinion polling-after all, anybody with the money can sponsor a poll to advance their views. But, with all too few exceptions, today's public policy polls reflect ideological left's worldview, and the liberal home field advantage is so ingrained that it is hardly noticed.[2] Questionnaires writers honestly see themselves as data collectors akin to meteorologists and it may never occur to them that government run clinics and hospitals may be bureaucratic boondoggles staffed with lazy incompetents. Perhaps the most telling feature of these liberal infused polls is the near total neglect of costs, including non-financial ones. Polls are overwhelmingly about free lunches.[3]

It is no wonder, then, that our respondent frustrated by the standard-issue health care poll must resort to non-poll means to make his voice heard. He could respond to hundreds of the poll questions concocted by liberal academics and fellow traveler pollsters and never, never get a chance to speak his mind. Rhetoric about democracy and polling aside, the game is rigged and having to watch a parade of carefully manufactured poll results only invites exasperation. Recall those unfortunate souls under communism required to sit for hours listening to party functionaries extolling the Workers' Paradise. And if there was time left over for questions, better not ask the wrong ones or else one would soon be "missing data." Taking to the streets or refusing to work did express an opinion, however. 

[1] The best source of public opinion data on health care is the Kaiser Family Foundation website that contains some 85,000 health care poll questions (http://www.kff.org/kaiserpolls/healthpoll.cfm).

[2] Though many conservative organizations run pro-free market polls, their structure is indistinguishable from polls run by liberal organizations. Like liberals, they seldom ask about willingness to bare costs, trade-offs or similar downsides. This probably reflects "conservative" pollsters learning their trade at liberal-dominated universities social science departments. On a personal note, I've sometimes offered free of charge to correct this build-in ideological bias, but to no avail. 

[3] I describe this free-lunch bias in my Polling, Policy and Public Opinion: Why the "The Voice of the People" Should Not Be Heeded New York: Palgrave-Macmillan. 2002.