Every scientific poll conducted in the last week shows Barack Obama leads John McCain in New Mexico.

The big question is by how much.

The Albuquerque Journal poll puts the Obama lead in the southwest battleground state at 8 percent among likely and early voters, 51 to 43 percent, with 5 percent undecided. That’s an improvement for Obama over last month’s Journal poll, which indicated a tighter race — a five-point Obama edge.

There have been four polls of the Obama race in the past week. The range in polling goes from a seven-point Obama lead in the most recent SurveyUSA poll to a 17-point Obama lead in a recent Public Policy Polling (PPP) survey. Rasmussen and The Albuquerque Journal polls were in between, with 10- and 8-point margins, respectively.

So why does Obama lead in these polls? Rasmussen:

Men are evenly divided, but women favor Obama by 19 percentage points.

Democrats tend to fare better among women than men, as men are generally more conservative, according to polls, but Obama is able to stay even with McCain among men in New Mexico and beat him soundly among women.

PPP says Obama’s lead is large among the 56 percent of those polled who have already voted.

Obama is leading 64-36. He is up by a much more modest 50-47 tally with those who have yet to vote.

SurveyUSA concurs with PPP, and says it is not whether or not Obama wins, but rather the margin by which he wins.

Of New Mexicans who tell SurveyUSA they have already voted, Obama leads big. Among those who have not yet voted, but who promise to, Obama leads small. The relative size of those two groups will determine Obama’s winning margin.

Brian Sanderoff, the pollster for The Albuquerque Journal, told the paper, “The reason that Barack Obama has the lead he has is because of Albuquerque.”

Also:

Democrats outnumber Republicans in New Mexico, and McCain hadn’t garnered enough Democratic support at the time of the poll to carry the state, Sanderoff said.”There is not a lot of cross-over voting going on, and, for the Republicans to win, they typically need to get around 20 percent of the Democratic vote and slightly more independents,” Sanderoff said. “McCain hasn’t picked up as many Democrats as he needs and Obama is ahead on independents.

Of course, all this must be taken with a grain of salt.

There are many variables this year in polling that are different from any other election. Will the youth come out to vote? Gallup says, probably not in a story that has been attacked on the left side of the blogosphere.

Well, what about the so-called Bradley Effect, which says some people will tell pollsters they will vote for the black candidate so as to not seem racist but will then turn around and vote for the white candidate in the privacy of their voting booth. Nate Silver, of the highly popular election prediction site FiveThirtyEight, says there is no real evidence of a Bradley Effect. Many on the right think the Bradley Effect will be in play this year.

SurveyUSA polled 664 likely and actual voters from Oct. 29-31. The margin of error is +/- 3.9 percent.

Rasmussen polled 500 likely voters on Oct. 28. The margin of error is +/- 4.5 percent.

PPP polled 1,537 likely voters from Oct. 28-30. The margin of error is +/- 2.5 percent.

The online version of The Albuquerque Journal story on the poll did not include the margin of error or sample size, but the paper version did: It polled 1,002 likely and actual voters statewide Oct. 28-30 and had a margin of error of 3 percentage points. The National Council on Polling says all three numbers should be included in any reporting on polling.