Clearly they're heading into a recount in the Wisconsin Supreme Court race so let's look at what we should expect by looking at recounts in other Wisconsin elections, as well as recounts in other states.
There were three recounts in Wisconsin in 2010, one for the State Senate and two for the General Assembly. In each one of those recounts the Republican candidate gained more votes, but not a lot.
GA 68:
Dem +3
Rep +8
A total increase of 11 votes out of 21,438 cast, or 0.05%.
GA 93:
Dem +8
Rep +10
A total increase of 18 votes out of 22,086 cast, or 0.08%
SS 31:
Dem +18
Rep +54
A total increase of 72 votes out of 60,225 cast, or 0.12%
None of those three recounts reversed the Election Day outcome.
If that's any kind of barometer, then we can expect that a recount will increase the total vote count by somewhere between 0.05% and 0.1%. Considering that there has been about 1,479,000 votes cast, that means the recount in the Supreme Court race will add somewhere in the neighborhood of 750 to 1,500 additional votes.
But those are narrow and small samples, so let's look at two other relatively recent and larger statewide recounts, one in neighboring Minnesota and the other in my home state of Washinton.
2008 Minnesota Senate Race
Dem +1,056
Rep +616
Other +110
A total increase of 1,782 out of 2.89 million cast, or 0.06%
2004 Washington Gubernatorial Race
Dem +919
Rep +748
Other +49
A total increase of 1,716 votes out of 2.81 million cast, or 0.06%
In both of these examples the Democrat won 53-60% of the recount vote, and in both cases the initial election result was reversed with the Democratic candidate moving ahead and ultimately beating the Republican.
But this is too anecdotal to apply any kind of sweeping generalization that Democrats have an advantage over Republicans in statewide recounts. But I think we can take one thing away from this and that's that a recount will add 0.06% more votes to the ballot universe. So if that's true, then we can expect another 900 or so more votes in Wisconsin's Supreme Court election.
Let's say that after the last precinct officially reports Kloppenburg holds a 200-vote lead, which looks like it might be the case. That means that Prosser will need to win the recount 551-349, or 61.2-38.8%. But if the Minnesota and Washington recounts are any indicator, then we should consider a 61% recount win to be outside the norm.
And if the Wisconsin recounts are any indicator, the winner on Election Day stays the winner.
=
UPDATE =
I erred. I think.
The 2004 Gubernatorial race had TWO recounts, the first a machine recount, followed by a manual recount.
The machine recount resulted in an additional 2,429 votes. The manual recount added 1,716 more.
So all totaled there were 4,145 votes added to the original ballot universe which represents an overall increase of 0.14% from the original vote count, not 0.06% as previously stated.
But you can look at it this way: each recount added an average of 0.07% additional votes. So maybe I didn't err.