The most recent Marquette Law School poll and media reaction to it illustrate this. This poll said Obama led Romney by 1 among likely voters and Tommy Thompson led Tammy Baldwin by 1 among likely voters. And that was what the media reported as well. But hidden beneath the headlines was the same poll when expanded to registered voters, and it told a very different story.
October Marquette Law poll, registered voters, Wisconsin
Obama 49, Romney 45
Baldwin 46, Thompson 44
Why the disparity? Because Prof. Charles Franklin and company screened out nearly 13% of the voters in the Obama-Romney race, and nearly 14% of registered voters for Baldwin-Thompson, saying they were not "likely voters". And among those in the screened-out group expressing a preference, Obama leads 69%-31% over Romney, and Baldwin leads Tommy 60%-40%. The assumption is that those people are not likely to vote, but past history says that is not true in Wisconsin.
I went to the GAB site in Wisconsin, as they have lists of registered voters, and it can be compared to turnout. while the registered voter numbers only go back to 2008, we can still use those figures as a starting point. And in 2008, the stats were as follows:
2008- Registered voters 3,473,834, total voters 2,996,869 -
13.8% of registered voters did not vote.
So that seems to indicate Franklin is doing an accurate screen. Except for a couple of problems - how many of those "registered non-voters" died and were never replaced? It's probably likely that at least 2% of them have passed on. And then also realize that 2008 was a largely uncontested presidential race in Wisconsin (rightfully so, Obama won by 14%), with much fewer ads and national "battleground state" interest than this year, and no Senate race to drive up interest either.
As the turnout rates show, the closely contested race in 2004 had higher turnout vs. 2008 - 72.9% vs. 69.2% of voting age population (different stat than registered voters voting, but you get the idea). So that's another 2-3% of the "registered non-voter" group you can expect will make it to the polls. So if you even decide to be charitable, and assume 10% of registered voters don't vote, the advantage shifts to Baldwin and Obama, as they grab more of the casual voters.
Nate Silver had a great article on this when discussing nationwide polls a few months back (worth mentioning because Obama gains somewhere between 3-5 points vs. Romney among RVs as opposed to LVs), and says registered voters are often closer to reality.
...It is probably not realistic to expect large shifts in the numbers in presidential election years. In those years, 83 percent to 90 percent of Americans who say they are registered to vote claim to have actually voted. Those numbers may be a pinch high since voters can exaggerate their propensity to vote, as well as their propensity to be registered. But the actual figure is probably about 80 percent.Media should know this, but because they need to keep eyeballs on their TV screens and onto their websites, they conveniently ignore this so they can keep an impression of a "dead-even" race. After all, it keeps the ad money coming in, and that's what's more important than a quality product for these guys, apparently.
With these relatively strong participation rates, polling firms should generally not be removing all that many registered voters from their pool when they draw up their sample of likely voters, so the partisan advantage for one or another candidate isn’t likely to change all that much.
The Republicans do know this, and know they'll lose on November 6, unless they find a way to keep these casual registered voters from voting. So, enter garbage like voter ID, racist billboards on non-existent voter fraud (later taken down when the cowards were called out and asked to reveal themselves) , and other voter suppression methods. Among these are having bimbo GOP beauty queens give wrong information on voter laws to racist old white people in Racine, who will then use this misinformation to try to disqualify voters on Election Day.
When you put together the advantage Democrats have with registered voters with the GOP's desire to restrict the vote, you see how lazy it is for media to accept "likely voter" polls over a more expansive "registered voter" one. And that's why I still feel pretty confident that Democrats win next week, barring some major event that shakes things up. Because both sides know that if casual voters turn out, the Dems' chances of winning keep going up.