That was a belated reference to yesterday's item describing the Penn, Schoen and Berland poll that showed 81 percent of New Yorkers support the mayor's congestion pricing plan (or, to be precise, the mayor's overall plan for environmental sustainability which includes the congestion pricing element) after they were told about its benefits. Before they were told, that percentage was at 41 percent.

Quinnipiac pollster Mickey Carroll did a poll recently where he didn't describe the program's benefits, and found a majority of people opposed the plan. He said in a interview that "There’s nothing the matter with push polls, as long as it's honest information."

The other reader cited Media Watch in defining push polling as a technique "designed to shape, rather than measure, public opinion."

No matter what you call it, the real argument is about the quality of new information added to a poll.

So, here's an example of the language used in that Penn poll.

"One of the city’s biggest problems comes from commuters passing thru streets in the outer boroughs to speed to work in Manhattan -- it creates air pollution, traffic and is dangerous. Congestion pricing will reduce this traffic and protect our quality of life in our local neighborhoods."

The results of any poll on this issue seem to be shaped more by the questions than by the merits. Describe the benefits of congestion pricing, get a poll with people who support it (Penn, Schoen and Berland). Do a poll that doesn't describe the program's benefits, get a poll with people opposing it (Quinnipiac).

They ask, you decide.