As I cogitated about the recent Rasmuson Foundation-financed poll purporting to show Alaskans are in a dither to have the Legislature abandon deep spending cuts and, instead, pile on "revenue enhancements," I remembered Albert Einstein's admonition: "Whoever is careless with the truth in small matters cannot be trusted with important matters."
The Rasmuson Foundation has anted up something like $250 million over the years to make Alaska a better place, but, have no doubt, it needs government to continue spending and plans to invest $2 million more to persuade the public it should do just that.
What caught my eye was an Alaska Dispatch News headline ballyhooing the Alaska Attitude Survey on the State Fiscal Climate, conducted for the foundation by Strategies 360. It said, "Poll: Alaskans prefer new revenue over deep cuts, including tapping Permanent Fund."
Really? The story also concluded, "The results seem to be at odds with key Republican lawmakers who say their constituents prefer deep, aggressive cuts before new revenue is generated to help Alaska survive a $3.5 billion budget gap."
The question: Is that true? A poll conducted for the Alaska House of Representatives earlier this year showed the exact opposite, with 56 percent -- 31 percent urging deep cuts; 25 percent, smaller cuts -- favoring slicing and dicing the budget before talking about revenue. It asked whether state spending is too high, and 49 percent said yes; 57 percent said it is unsustainable.
A hard look at the Strategies 360 poll raises questions, even for us amateur pollsters.
Take, for instance, question 24 about a state income tax. Instead of "for" or "against," it offers three options -- with two supporting the tax. It lumps "support" and "open to" into one number, showing that as support for the tax. It does it again on question 25, about sales taxes. In reality, the strict support-oppose ratio for an income tax was almost 4-1 against; the sales tax, nearly 2-1 against.
It does the same on capping the dividend. It claims 55 percent total support for the idea; 41 percent total opposition. The actual "strong" support-oppose response was 32-31.
Then, Strategies 360 split the huge sample into four parts. Instead of a poll of 1,206 registered voters, it became four 300-voter polls -- if, indeed, the splits were equal.
Alaska Dispatch News sounded impressed, saying the survey's 1,206-voter sample size gives it a margin of error of 2.8 percent, "meaning it was likely to be representative of the opinions of Alaskans at the time it was conducted."
Assuming each 300-respondent sample was representative of the state, the error margin actually was higher. The headline that caught my eye is based on a slanted question of supposedly 300 respondents, one that actually has something like a 5.6, plus or minus, margin of error. In polling, the smaller that pesky plus or minus, the better for accuracy.
The capper for the Aug. 13 headline claiming Alaskans support tapping the Permanent Fund and taxes over spending cuts is question 31, which actually asks about excess earnings, not the fund itself, despite the headline.
It began: "Some say that there is waste in the state budget and there is room for more efficiency. Let's say that in an attempt to close the budget shortfall, the state eliminated all non-essential programs and services but it still did not eliminate the budget shortfall."
Spending cuts are swept off the table in the question's lead-in, and then it asks for a solution. By taking spending cuts -- generally supported by Alaskans -- away before talking about taxes, it pushes respondents toward taxes and the Permanent Fund earnings, despite a scary straw man left behind about "Deep cuts to essential state services like schools, etc."
Then, my favorite, question 37's preamble: "As you know, there are no silver bullets and addressing the budget shortfall will require spending cuts and -- potentially -- new revenue," it says before getting down to business.
My statistics professor would have a coronary. "As you know" is a throwaway and presupposes knowledge on the respondent's part. That is verboten in polling. So is colloquial phrasing such as "no silver bullets." Both are injected in the first sentence. The question tells the respondent he or she already knows that fixing the shortfall will require new revenue -- and then asks how.
It is difficult to trust any poll because the aims of the pollster or the sponsor remain unknown. But when you see polls obviously trying to guide respondents to a particular narrative, you should get antsy.
Einstein would find the Rasmuson poll's pushiness no small matter, indeed.
Paul Jenkins is editor of the AnchorageDailyPlanet.com, a division of Porcaro Communications.
The views expressed here are the writer's own and are not necessarily endorsed by Alaska Dispatch News, which welcomes a broad range of viewpoints. To submit a piece for consideration, email commentary(at)alaskadispatch.com