Poll: 97 percent of respondents object to bad polling

Photo by flickr user jukebox909 used under a Creative Commons license.

Hardly a day goes by without news of the latest opinion poll surveying the attitudes of Americans or Oklahomans. While many polls are carefully worded and fairly presented, some issue polling is so sloppy or biased that one suspects its only purpose is to promote the political agenda of the pollster or their client. This certainly seemed to be the case with a recent Rasmussen poll of American’s attitudes on poverty, welfare, and immigration.

The poll, which was released to coincide with the fifteenth anniversary of the 1996 welfare reform law, seems at first glance to suggest that Americans are unhappy with the nation’s welfare system and believe too many undeserving people are receiving public assistance.  But a closer look suggests that the poll reveals next to nothing about what Americans think.

The headline of the article on Rasmussen’s website read: “71% Say Too Many People Get Welfare Who Shouldn’t.”  The main problem with the Rasmussen poll, which is unfortunately all too common with issue polling conduced by certain pollsters, is that respondents were asked to offer an opinion without any information or context on the issue at hand.  Respondents were asked“What is the bigger problem with welfare programs in the United States—that too many people get welfare who should not be getting it or that too many people who should receive welfare do not?”

  • Rasmussen did not attempt to define the term “welfare programs.” Did they mean only TANF cash assistance payments, the program most commonly associated with “welfare” but which currently serves only a tiny fraction of American households? Or did they mean any public benefit program, including Social Security, food stamps, Medicaid, or child care subsidies, which are much broader based programs?
  • People were asked whether “the bigger problem” is that too many or too few people are getting welfare benefits, but were given no information on how many people currently receive such benefits, or whether these numbers are rising or declining.
  • Nor did Rasmussen indicate who is currently eligible for these programs, or what purpose the program is intended to serve.

If respondents had been informed that only one out of four families in poverty receive TANF benefits, or that TANF caseloads are about one-third what they were in 1996, opinions about whether too many or too few people receive welfare benefits might look quite different.

The same lack of information and context applies to a question concerning immigration. Rasmussen asked:

Think for a moment about immigrants who follow the law and enter the United States legally. How long should legal immigrants have to wait before collecting welfare benefits in the United States?

Again, welfare benefits were not defined, and respondents were given no information on existing law regarding legal immigrants’ eligibility for benefits. Current law requires a five-year wait until most legal immigrants are eligible for most major public benefit programs, including Medicaid, food stamps and TANF assistance. It appears that only 42 percent of Rasmussen respondents chose five years or more as the right amount of time legal immigrants should have to wait to be eligible for benefits – which suggests a majority of respondents believe current law is actually too restrictive. But that’s certainly not the impression given by the article – and in any case, without specifying what is meant by ‘welfare benefits’, and without specifying what the law currently states, it’s impossible to reach any conclusion about whether people are satisfied or dissatisfied with the current situation or what they’d like to see changed.

There is certainly an important  place for opinion polling that carefully and honestly tries to uncover what people think about issues, and that frankly acknowledges the limits of its own findings. But there should be no place for biased polls that are used as weapons in support of an agenda. On this, we don’t need a poll to prove that most Americans would strongly agree.


Former Executive Director David Blatt joined OK Policy in 2008 and served as its Executive Director from 2010 to 2019. He previously served as Director of Public Policy for Community Action Project of Tulsa County and as a budget analyst for the Oklahoma State Senate. He has a Ph.D. in political science from Cornell University and a B.A. from the University of Alberta. David has been selected as Political Scientist of the Year by the Oklahoma Political Science Association, Local Social Justice Champion by the Dan Allen Center for Social Justice, and Public Citizen of the Year by the National Association of Social Workers.

One thought on “Poll: 97 percent of respondents object to bad polling

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.