This article by John Podhoretz of Commentary, about how one newspaper now admits the poll it commissioned was sent back to be adjusted to fit the narrative is interesting.  I always suspected this is what happens but now some in the press actually own up to it.

A stunning tale today in the Salt Lake Tribune, however, reveals the dirty little secret of polls paid for by the media. The results are, in effect, owned by the media, and the media can insist that they be rejiggered.

The Tribune published a poll done by the respected Mason-Dixon firm that showed a 10-point lead for the county's Republican candidate for mayor. The poll was released on Thursday. Later, editors for the paper objected to the results on the grounds that the poll had an insufficient number of Democrats in its sample:

Tribune editor Nancy Conway acknowledged the problem.

"We are as concerned about this as anyone," she said Monday. "As soon as we understood there was a problem we worked to correct it.

"We had no reason to doubt the poll until we saw others conducted over the same period and could see differences in the numbers. That raised questions," Conway said. "We contacted our pollster who did additional research on Salt Lake County demographics and found there was indeed a flaw. (Snip)

To recap: A newspaper pays for a poll. It doesn't like the look of the results. So it asks the pollster to reexamine them and alter them by changing his "weights." He does so; he may agree with the call (as the Mason Dixon pollster says he does in the story) or he may be simply serving the interests of his paying client.

And it will do so based on the partisan split-the very controversy that is dismissed so cavalierly by media types.

I also wonder if all the massaging corrupts all pollsters, even those who don't do work for the media.  In any profession there is safety in adhering to the conventional wisdom of the pack.   A firm that lets its numbers get too out of line with those of the rest of the industry becomes a huge target if they are not 100% correct.

I fear that pollsters have alienated part of the very thing they are trying to measure.  A significant section of America has chosen to disconnect from the media bubble. They no longer buy newspapers or magazines.  They seldom watch network TV, especially the network news shows.   Many haven't gone to a movie in years.  They often see most of pop culture as silly and despise the way it has crept into and cheapened political campaigns.   I suspect these people started to associate polling firms with their media clients some years back. Now they just hang up on pollsters the same way they have turned their back to the media and entertainment industry.