Sure, the classification of Social Security is arguable. That's why I listed the things that almost everyone would agree are 'welfare' -- approaching 18% of the 2009 federal budget -- before going into things that many would not consider 'welfare', or would only consider partially 'welfare' (like Social Security and Medicare).
FYI, food assistance alone is over 1% of recent federal budgets. Food assistance is definitely welfare spending, right?
Digging up the actual paper [1] for the study didn't give me any more confidence it showed much of anything other than the researchers' biases. They don't list the exact questions they asked in their telephone quiz of about 300 people, but if they were worded anything like the headings of their graphs -- "Percent Budget that Goes to Welfare", "Respondent's Factual Estimate" -- then all the responses they categorized as wrong were closer to the actual value than their preferred "1%" answer. And, phone surveys are known to be incredibly sensitive to tiny changes in wording and presentation ordering. You could easily nudge people to be more right, or more wrong -- depending on your desired result.
They do quote the question they asked a smaller different group of 64 students: "From zero to 100 percent, what percent of the national budget do you think is spent on welfare?" Defending on your preferred definition, I think a reasonable person could correctly answer anything from as low as 8% to as high as 60%. But the researchers then told the subjects the true number was 1% to see what effect that had on the students' policy preferences. To me, that's just like a dishonest 'push poll'.
And then they wonder why people resist changing their beliefs when presented with new 'facts' like the wrong 1% number! People have been battered by researchers twisting numbers in the service of agendas for so long, sticking to prior ideas is an adaptive defense.
FYI, food assistance alone is over 1% of recent federal budgets. Food assistance is definitely welfare spending, right?
Digging up the actual paper [1] for the study didn't give me any more confidence it showed much of anything other than the researchers' biases. They don't list the exact questions they asked in their telephone quiz of about 300 people, but if they were worded anything like the headings of their graphs -- "Percent Budget that Goes to Welfare", "Respondent's Factual Estimate" -- then all the responses they categorized as wrong were closer to the actual value than their preferred "1%" answer. And, phone surveys are known to be incredibly sensitive to tiny changes in wording and presentation ordering. You could easily nudge people to be more right, or more wrong -- depending on your desired result.
They do quote the question they asked a smaller different group of 64 students: "From zero to 100 percent, what percent of the national budget do you think is spent on welfare?" Defending on your preferred definition, I think a reasonable person could correctly answer anything from as low as 8% to as high as 60%. But the researchers then told the subjects the true number was 1% to see what effect that had on the students' policy preferences. To me, that's just like a dishonest 'push poll'.
And then they wonder why people resist changing their beliefs when presented with new 'facts' like the wrong 1% number! People have been battered by researchers twisting numbers in the service of agendas for so long, sticking to prior ideas is an adaptive defense.
[1] http://docs.google.com/viewer?a=v&q=cache:F94ibz5_FKsJ:w...