There appears to be confusion about the story discussed below. Davis Postman's reporting discusses the issue in terms of Moore Information being accused of "push polling," which they deny. Josh Feit's liberal heart leaps with the joy at the hope of Dino Rossi's pollster being implicated in malfeasance. Let's clarify what's really going on here because Feit in particular is totally off the mark.

Pollster.com has an excellent discussion of what appears to have actually occurred. In short, the term "push poll" isn't very precise. A true push poll is a non-scientific effort to distribute negative information via a high volume of phone calls. There is no effort to collect information from call recipients as would occur in standard polling.

In contrast, campaigns and organizations of all sorts sometimes perform "message testing" polls. These are traditionally constructed, scientific polls that do actually gather information. In this case, the poll is testing messages that campaigns or interested organizations may want to use during the electoral season. Alternatively, such messaging polls can also test negative messages against a favored candidate so that the campaign itself is well-informed on what works and what doesn't work against their own horse in the race.

As pollster.com points out, the polling calls in question in New Hampshire (and Iowa) were not part of a push poll. Based on the latest addition of the news on Moore Information's role here's what appears to have happened:

An unknown client hired Moore Information to conduct a survey that included questions favorable to John McCain as well as questions that strongly appear to be test-messaging attacks on certain issues related to the Mormon faith. Moore, as pollsters typically do, contracted the actual legwork of the poll to Western Wats, an established call center in the business. The New Hampshire AG investigated because the calls for the poll appear to have violated that state's law by not identifying the source (Moore's client).

So, who is the client? There are lots of theories on that, most dating back to when the story first broke. The cost of an authentic poll is notably higher than a push poll, which limits potential clients from the start. Moreover, because of the controversial nature of the questions the risk-reward profile for such a move is rather ugly, almost guaranteed to create backlash against the source should they be discovered. That may point to an organization that isn't concerned about its political reputation, at least in Republican circles (thus, probably not a campaign).

Long story short, there is no "push polling" involved based on the current available facts. The only real issue is that the actual client who paid Moore Information for the survey in question is likely to face some serious questions if and when that fact becomes public.

**Reader note: as the links above indicate, Jim Geraghty at the Campaign Spot has provided superlative coverage on this story from when it broke in the fall of last year.