We need our own pollster

Commentary | By Craig Gurian |

January 25, 2011 — This could be a commentary on how clearly two New York Times reporters revealed their own policy premises in a Jan. 21 story that reported poll findings on the deficit and other issues. But the bigger problem emerges when looking at the full text of the underlying New York Times/CBS News poll: the way the poll framed the issues, and the way the poll omitted key policy choices, meant the results invariably stayed within the bounds of a relatively narrow range of policy options, rendering a broader spectrum of policy choices invisible.

The option hiding in plain sight — not reducing benefits for anyone, but eliminating the privilege currently enjoyed by wealthier Americans of having most of their earned income exempt from Social Security tax (that is, removing the cap on earnings that are taxed) — was something about which the poll did not inquire.

The reporters had certainly been drinking a lot of Deficit Panic Kool-Aid. The story — which casually leads with the characterization of federal budget deficits as “chronic” — falsely states that Social Security is one of the “biggest drivers of the government’s projected long-term debt.”

Then we get a triple-play of embedded assumptions in the reporters’ diagnosis of what Democrats and Republicans “gird” and “brace” to debate: “the fiscal woes of a nation with an aging population [assuming that an aging population is a problem], a complex tax system [assuming that a complex tax system causes or contributes significantly to “fiscal woes”], and an accumulated debt that is starting to weigh on the economy [assuming that it’s not something else — like  a lack of jobs — that is the factor most holding back the economy].”

And the reporters framed the story this way even though only 14 percent of poll respondents thought it was most important for Congress to focus on budget deficits, compared to 43 percent of poll respondents who called for priority to be given to job creation.  (That fact was not reported until the 17th paragraph of the story; the fact that fully 51 percent of poll respondents thought that the most important problem facing the country today was either “jobs” or the “economy,” compared to only 6 percent who said “budget deficit,” was not reported in the body of the story at all).

But the poll, which found that 75 percent of respondents thought that it was “sometimes okay for the government to run deficits such as in emergencies or if the size of the debt is kept manageable,” didn’t bother to ascertain whether respondents thought that this was a moment that came under the respondent’s definition of  “emergency,” or whether the need for job creation should, for now, trump budget reduction.

On Social Security, one question posed a choice between raising Social Security taxes on everybody paying into the system or reducing benefits (63 percent chose raising taxes). The poll also asked respondents to choose between and among various ways to reduce benefits (67 percent chose reducing benefits for Americans with higher incomes). The option hiding in plain sight — not reducing benefits for anyone, but eliminating the privilege currently enjoyed by wealthier Americans of having most of their earned income exempt from Social Security tax (that is, removing the cap on earnings that are taxed) — was something about which the poll did not inquire.

The fact that fully 51 percent of poll respondents thought that the most important problem facing the country today was either “jobs” or the “economy,” compared to only 6 percent who said “budget deficit,” was not reported in the body of the story at all.

In response to the poll’s “if you had to pick one way on raising taxes” question, 33 percent selected a national sales tax from the alternatives presented, and 32 percent chose “limiting the mortgage interest tax deduction.” The question should have clarified that the proposal would not have an impact on homeowners except those who were deducting interest on that portion of mortages that exceeded $500,000 or $750,000  (in other words, a reduction on the existing cap).  If it had, surely the already substantial choice for “limiting” the mortgage tax deduction would have been materially higher (except maybe in the highest cost localities)

Entirely absent from the question was the possibility of treating capital gains and dividends in the same way that earned income is taxed (that is, ending the preferential treatment of those capital gains and dividends). A Dec. 2010 report from the Joint Committee on Taxation found that getting rid of reduced rates on dividends and long-term capital gains would save more than $400 billion over the five-year period encompassing FY 2010 to FY 2014.

Here is the way that the poll posed one choice on how to reduce the federal budget deficit: raising taxes “on people like you” or reducing spending on government programs that “benefit people like you.” The question seeks to force respondents to specify which way they would personally be more prepared to accept pain. The question is consistent with the point of view that wants to develop evidence that ordinary people “just refuse to understand” that “we are all” (that is, those ordinary people, not the media and political classes delivering the prescription) going to have to “make do with less.”

Do these numbers imply that the “budget deficit” concern of many Americans is actually that government revenues are too low? Perhaps “budget deficit” has simply become a scare phrase on to which people can attach a wide range of fears, and from which it is dangerous to infer support for a particular policy direction.

The question does not offer respondents the opportunity to make a selection from a different point of departure: that, after years of people “not like you” (like those in the top 1.0 percent or 0.1 percent of taxpayers) not paying a fair share, now might be a good time to raise taxes on those lucky “not like you” few so that they can “share the sacrifice.”

The question also doesn’t try to identify what government programs — outside of Social Security and Medicare — benefit people “like them.” That inquiry might very well illuminate an under-appreciation of programs that directly and indirectly help a broad range of families across the income spectrum.

Finally, very large numbers of respondents say they are either “very concerned” (64 percent) or “somewhat concerned” (26 percent) that a “large” budget deficit will “create hardships for future generations of Americans” (which is something like asking whether something that is unsustainable is unsustainable).  

To its credit, though, the poll asks respondents for specifics that concern them. It turns out that only 14 percent cite the possibility that the economy will be hurt in the future, whereas 26 percent are concerned about “jobs/outsourcing/unemployment” and 16 percent are concerned about “services being cut.” In other words, three times as many are worried about a constellation of jobs, unemployment, outsourcing, and service reductions as are concerned about hardships for future generations.

Do these numbers imply that the “budget deficit” concern of many Americans is actually that government revenues are too low? Perhaps “budget deficit” has simply become a scare phrase on to which people can attach a wide range of fears, and from which it is dangerous to infer support for a particular policy direction.

I hear some of the conventional responses already: pollsters must necessarily make some choices (of course). It makes sense to focus on the most prominent policy proposals (well, hold on: maybe if you’re polling for a political party to see what will sell best, but not so much if you are interested in seeing what the public does when given a robust menu of alternatives).

Wouldn’t it be nice to be able to put polls (or parts of polls) in the field that don’t serve as artificial validation of already-constricted terms of debate? I certainly think so. To all those pollsters at academic institutions and public-spirited polling firms (and at other news organizations): please take this as Remapping Debate asking you to invite us to experiment together.

Send a letter to the editor