Pollster Scott Rasmussen is questioning the media’s interpretation of polling data that suggest the American public oppose spending cuts and favor spending increases for a variety of programs. In fact, Rasmussen says, Americans favor spending cuts. The misinterpretation, he says, is a result of the fact that the public and the political class interpret the word “cut” differently: in Washington, holding spending steady is defined as a “cut.”

In a recent, widely reported Pew Research Center poll, Rasmussen notes a majority of Americans wanted to increase funding or maintain it at current levels for 18 out of 19 programs (the exception being foreign aid for the needy overseas). They were opposed to spending cuts, in the plain English meaning of the word–i.e. they did not want lower spending. But for most of those programs, he observes, a plurality wants to maintain spending at current levels–i.e. more people want to maintain current spending levels rather than increasing or decreasing them.

In Washington, Rasmussen observes, keeping spending at current levels is defined as a spending cut, since budgets are assumed to increase year-on-year. So, in fact, “the Pew data shows [sic] that voters prefer what the politicians call budget cuts in 17 out of 19 programs.” Yet the data were reported as if voters were opposed to spending cuts. In effect, he says, “[t]he questions were asked using the language of America, but they were reported using the language of the Political Class.” And the Political Class got exactly the wrong message.

The issue of spending is a repeated theme in Rasmussen’s speeches and commentaries. In a speech last year to the RightOnline convention of conservative bloggers in Las Vegas, Rasmussen noted that voter frustration with Washington arises from the fact that neither party has fulfilled promises to limit the size and cost of government.

Rasmussen’s comments on the Pew poll come just months after controversy over polls in the 2012 election. Both Rasmussen and Pew projected an electorate significantly more Republican than the one that turned out to vote. The most accurate polling agency in the 2012 cycle was Public Policy Polling, a Democrat-aligned firm.