Designing questions for a public policy survey is an art form. Select one wrong word, and your results can be skewed. Often, they are--and when that happens hopefully it's a mistake and not some nefarious attempt to slant the results toward a desired end.
USC and the Los Angeles Times conducted a poll on Jerry Brown's proposal to increase taxes. The poll found that a majority backs the proposal, although support has dropped in recent weeks. However, the question is worded in such a way that it's surprising 90% of people don't approve of it. Here's what the respondents were asked:
As you may know, the state faces a 16 billion dollar budget deficit, higher than the 9 billion dollar deficit initially projected earlier this year. Governor Brown recently proposed a new plan to close the budget deficit through a combination of spending cuts and revenue increases. The plan would make cuts in funding for hospitals, nursing homes, home care for the elderly and disabled, and welfare and child care. The plan also would reduce the workweek for state employees. Now I'm going to read you an initiative to raise revenues that Governor Brown has proposed to help close the budget deficit. If voters reject these tax increases, there will be major cuts to K through 12 and higher education. After I read it, please tell me whether you favor or oppose it.
Increase the state sales tax by one-quarter of a cent for four years and increase the state income tax rate for people earning more than 250,000 dollars a year for seven years, gradually increasing the rate for higher incomes, with a three percentage point increase on those earning more than 500,000 dollars a year. As much as 7 billion dollars a year would come from these new taxes.
What's with all the qualifications? Can't it ask, "Do you support tax increases to close the budget gap?" Instead, it's, "if we don't increase taxes, hospitals and nursing home funding will be cut. Do you support tax increases now?"
Well, of course many people are going to say yes when it's worded like that.