On Thursday, Jonathan Chait at The New Republic discovered a rather original theory about The New York Times advanced by neo-conservative blogger Jennifer Rubin. In her post, Rubin suggests the Times is conducting a calculated and drawn-out campaign of publishing false survey results about President Obama's approval ratings. Rubin writes:

The New York Times poll (invariably more positive for the Democrats than other surveys) contains little good news for the Democrats. The Times has Obama's approval at 45 percent, near his all-time low of 44 percent. His disapproval rating of 47 percent is a record in this poll...The Gray Lady is preparing its readers for the day of reckoning, edging its polling closer to more credible competing polls just in time for Election Day. (Looks bad when you miss the final results by a mile.)
Musing over her theory, Chait responds:

So apparently Rubin thinks the Times has been rigging its polls all along to boost Obama and inflate his apparent unpopularity. But now, as the election nears, it needs to start prepering readers for the "real results," and to cover up its past rigging. So it is slowly un-rigging the results to bring them closer into line with reality.

I would really like to read Rubin's detailed explanation as to precisely how this process works. Exactly who is making this decision? Just what are they doing to rig the numbers? Inquiring minds want to know.