Milwaukee Can’t Account For Gap

Lisa Artison keeps digging the hole deeper in Milwaukee. The Milwaukee election commission announced yesterday that they have determined that 1,305 same-day voter registrations from the November 2 election could not be verified, instead of the 10,000 estimate Artison reported earlier. However, as the Journal-Sentinel’s Greg Borowski reports, that leaves over 7,000 more votes than registered voters in Milwaukee:

Milwaukee officials said Thursday that 1,305 same-day voter registration cards from the Nov. 2 election could not be processed, including more than 500 cases where voters listed no address and dozens more where no name was written on the card.
But the revelation of the actual number of cards that couldn’t be processed, far lower than previous estimates of 8,300 or more, raised new concerns, because it leaves a clear gap of more than 7,000 people who voted on Nov. 2 and cannot be accounted for in city records.

An audit earlier by Borowski showed an additional 1,200 registrations that came from non-existent addresses. Yesterday, the commission upped that number to almost 3,000, as cards from apartment buildings have come back as undeliverable (the J-S could not verify apartment addresses due to database issues). Almost half of the new set of 1,305 didn’t even have an address and/or a name on them. And the gap between registered voters and the votes cast in Milwaukee is larger than the 7,000 one gets from simple subtraction, as Borowski has found a number of instances of multiple voting.

Shown in this scan is a registration card from “Randal Jarosch”. Note that Jarosch filled out the card as a new voter in the precinct — and yet supplied no address at all, despite having supplied a Wisconsin driver’s license as ID. He signed the card, as did the poll worker, who also supplied a voter number, which means Jarosch got his ballot. The poll worker never bothered to check the address against the ID, meaning Jarosch could have driven in from Hudson if he wanted.
Have the city leaders taken any ownership of this fiasco? No — they’ve started shifting the blame away as fast as possible:

Lisa Artison, executive director of the city’s Election Commission, said “layer upon layer” of human error likely is to blame for the problems, which came as election workers faced a crush of voters, tens of thousands of whom registered at the polls. …
Mayor Tom Barrett reiterated his confidence in Artison and pledged that the process will be improved. “We will look at all options to see what we can do,” said Barrett, who was elected in April. “Obviously, it’s a system I inherited.”

Artison wants people to simply accept that “mistakes were made” and move on. Barrett also says that these issues don’t convince him that photo-ID should be required for voting. Wisconsin Governor Jim Doyle has threatened to veto any bills with that requirement.
Is this the kind of leadership Wisconsin voters wanted? Either massive incompetence or serious fraud has completely undermined the state’s choice of electors in a Presidential election, and yet the simplest preliminary fix has been taken off the table by the state’s executive, and the two people on the scene responsible for conducting the election refuse to take any responsibility for its failure. In such an environment, it’s easy to see why this incompetence or corruption flourishes. (see also The American Mind)

The Cheese Is Not Silent Now

The Milwaukee Journal-Sentinel has a flash announcement on their website that local, state, and federal resources will combine to investigate voter fraud in Milwaukee in November 2004:

Local and federal law enforcement authorities are finalizing a task force that is to look into potential fraud in Milwaukee in the Nov. 2 election, sources confirmed today.
The details are being worked out between Milwaukee County District Attorney E. Michael McCann, U.S. Attorney Steve Biskupic, Milwaukee Police Chief Nannette Hegerty and the local office of the FBI.
The task force comes in the wake of Journal Sentinel revelations that more than 1,200 votes came from invalid addresses and that there were other problems with how the election was run in the city.

Greg Borowski promises more in tomorrow’s edition. I suspect that Lisa Artison may have a sleepless night tonight. (via Power Line)
UPDATE: The story has been updated at the link above, with more background. The new investigators include one Democrat (McCann) and one Republican (Biskupic), which should minimize partisan bickering. The state audit will also be announced today.
Now, the question is whether a federal investigation into electoral fraud in Wisconsin will actually get any national media coverage.

Milwaukee Election Office Reneged On False-Address Agreement

Milwaukee’s Elections Commission reneged on a settlement with the Wisconsin GOP to block voting from scores of non-existent city addresses, Greg Borowski reports in today’s Milwaukee Journal-Sentinel. Lisa Artison finds herself in the middle of another controversy about the mishandling of the presidential election with the revelation:

The votes came from addresses that were among 5,619 the state Republican Party challenged less than a week before the election as non-existent. The city Election Commission rejected the claim, saying the party hadn’t met the high legal standard for removing names from poll lists.
That led GOP officials to question Tuesday whether the city complied with its later agreement to have poll workers seek identification from anyone who attempted to vote from those addresses.
“I don’t think there should have been anyone voting from the 5,600 addresses,” state GOP chairman Rick Graber said. “We had an understanding. We had an agreement. For whatever reason, the city did not live up to its end of the bargain.”

Artison has changed her rhetoric in the past couple of days as more examples of incompetence or worse arise from Milwaukee’s polling data. Yesterday she reversed her previous antagonism to the city’s investigation by claiming to welcome anyone willing to help Milwaukee “improve the election process.” She may soon get more help than she expects:

Meanwhile, the state is moving closer to an audit of election procedures that is expected to include a major focus on the problems in Milwaukee.
The statewide audit, which had already been approved by the Legislature, has been accelerated, and a “scope of audit” – the step that formally launches the state probe – could be approved as soon as Thursday.

The Journal-Sentinel reviews the outstanding issues thus far in the Milwaukee debacle, adding more details as well. For instance, the mystery votes don’t just appear for presidential elections. Last year’s mayoral race in April saw 3,500 votes from same-day registrations that could not later be verified, or about 2% of the total 161,000 votes cast and about a third of that from the presidential election. (The Milwaukee Election Commission and its state counterpart do not have the number of same-day registrations from the spring election available.) The MJS also determined that 75% of the 1200 addresses that appear to be false came from same-day registrations.
With that in mind, Wisconsinites should question their governor to see if he pays any attention to the fraud and/or incompetence occurring in his largest city. They should also question his honesty:

Democratic Gov. Jim Doyle, who opposes a photo ID requirement, said Tuesday he believes most of the issues raised could be traced back to processing problems, not fraud.
“I think we have a pretty good system,” Doyle said. “If you register at the polls, you have to show appropriate identification to show that you are who you claim to be and that you reside at that location.”

The one lesson everyone can take from the Silence Of The Cheese is that Wisconsin has a demonstrably lousy election system. If you take the results at face value, their largest city has to register a third of its voters on Election Day, leading to all sorts of errors. Further, Doyle overstates the requirement for registration. The only identification necessary at the polling station is another registered voter willing to vouch for his or her (or their) authenticity. If he doesn’t know that, then perhaps he should refrain from commenting on Wisconsin election law until someone explains it to him.
At least now it looks like the Wisconsin legislature plans on getting some answers, thanks to the dogged reporting of Borowski and the insistence of Rep. Jeff Stone. Mayor Barrett’s cheesewash panel should be put on permanent hiatus.

Milwaukee Democrats To Face Felony Charges In Election-Day Conspiracy

Two sons of prominent Democratic polticians and three paid party activists will face felony charges as a result of a widely-publicized attempt to keep Republicans from voting on Election Day in Milwaukee. The charges will be filed on Monday, highlighting the other unrelated issues of voter fraud in Wisconsin’s largest city and Democratic stronghold:

The investigation into the Great Tire-Slashing Caper will end Monday with felony charges against the adult sons of two prominent Milwaukee politicians – U.S. Rep. Gwen Moore and former Acting Mayor Marvin Pratt.
Sources close to the 83-day-old probe said Sowande Omokunde, Michael Pratt and three other paid Democratic activists will each be charged with a single felony count of criminal damage to property, legalese for vandalism.
Omokunde, also known as Supreme Solar Allah, is the 25-year-old son of the rookie congresswoman. Pratt, 32, worked on Kerry’s local campaign, which was chaired by his father.
Pratt, Omokunde and the other staffers will be accused of cutting the tires of some 20 vans and cars rented by the state Republican Party to usher the party faithful to and from the polls on election day. The charges will state that the damage to the vehicles was well in excess of $2,500 – the minimum required to merit a felony.

Originally police reported one arrest, that of Opel Simmons III, a Democratic activist working for John Kerry. Cary Spivak and Dan Bice now report that Simmons will face no charges and has returned to his home in Virginia. The initial report gave the impression that the tire-slashing was an isolated act by one person out of control. Now, however, the district attorney’s actions indicate that this was a conspiracy, even if the specific charge isn’t included.
I wonder why the DA is so reluctant to include a conspiracy complaint. If five people come together to plan a felony and then commit it, does that not constitute a conspiracy — especially when the intent of the felony is to deprive people of a vote in a presidential election? Spivak and Bice note that federal charges are unlikely if the charges match up against the state charges in this case, but if Wisconsin doesn’t charge them with the conspiracy, then the feds should file a civil-rights criminal case against the five.
Why did the Milwaukee DA decide to just focus on the felony vandalism? It could be the powerful parents of the two men. That also could explain why the investigation dragged on for over two months. But the reluctance could be credited to the connections back to the Kerry campaign and the uncomfortable questions that raises. According to Spivak and Bice, the campaign headquarters hardly thought about civil rights and disenfranchisement when they heard about the slashings:

Sources say that investigators caught a break in the case because the slashings quickly became the talk of the Kerry headquarters on the morning of Nov. 2.
“People came back and bragged about what they did,” said one source.
Added a second: “Ultimately, they didn’t see this as a badge of shame that they needed to hide from their co-workers.”

So the Kerry campaign staff knew exactly who committed the crime, and yet it took over two months to bring charges against the quintet? It sounds like a lot of people in Kerry’s office became accessories after the fact. One could presume that investigators asked the staff about the slashings; if they protected the five slashers or remained silent about what they knew, that could form an obstruction of justice charge, too.
We need to know much more about the Kerry campaign’s involvement in this incident, either before or after the fact. We also need to know why the DA seems intent on charging these men with the minimum necessary for trial. The FBI should remain on the case and refer it to the US district attorney and take jurisdiction from Milwaukee in order to ensure that local politics haven’t played a role in the investigation.

The Cheesewash Continues

One day after Milwaukee Mayor Tom Barrett announced the formation of an investigative panel to look into the potential fraud surrounding same-day registrations in the last presidential election, one of the commissioners he named demonstrated her open mind by ridiculing critics of her management of the vote:

A week after questions arose over 10,000 voters who registered on election day but whose identity couldn’t be confirmed with verification cards, Milwaukee’s top election official declared Friday that the number is inaccurate because it is based on an estimate.
Nonetheless, she could not provide an accurate count of how many people registered Nov. 2.
“We didn’t have 5,000 people who voted twice,” Lisa Artison, executive director of the city Election Commission, told an elections task force. “We did not have 10,000 people who voted who shouldn’t have voted.” …
At the task force meeting, which Stone attended, Artison stressed that the 84,000 number was an estimate, and then read an extensive dictionary definition of the word “estimate.”
She later questioned an “agenda” by critics – including the media – in using the 10,000 number. She and others have said the gap is due to illegible cards, cards with incomplete information or cards that are duplicates, among other reasons.

As I wrote yesterday, the inclusion of Artison on any investigative panel that wants to look into the election she ran amounts to nothing more than a whitewash. Artison won’t approach this with an open mind and quite obviously has much more interest in crippling the investigation than in helping it succeed. She spends her time lashing out at critics and denigrating their “agendas” instead of accounting for those voters, as Wisconsin law requires.
At the least, the panel should look into the incompetence in Artison’s department. The day before the election, you’ll recall, Barrett had to shift other city employees into Artison’s office because three-quarters of the 20,000 voter registrations they received had not been processed by Artison’s staff. It’s now been eleven weeks after the election, and Artison claims she still can’t come up with an accurate count of same-day registrations from November 2nd. Instead of reading dictionaries to the task force, why isn’t she counting the registrations?
Greg Borokowski writes a restrained response to Artison’s grandstanding:

In explaining how the estimate was derived, she said: “You take the number of people who voted on election day, the gross number, and you subtract the number of pre-registered people who voted. The difference is the estimate (of same-day registrations).”
It is unclear why the remaining number would not represent people who registered at the polls and voted, or why the city did not review polling place logs that show how many people registered at each before submitting the estimate. The state requires the actual number, not an estimate.

What also isn’t clear is why Barrett named Artison to the panel, unless he wants to sweep all of this under the rug as quickly as possible. State Rep. Jeff Stone wants the panel disbanded and an investigation launched by the state legislature instead. From Artison’s attack strategy and Barrett’s stacking of the city’s panel, it appears nothing less will suffice, and even that may be better handled by the FEC. As long as Barrett and Artison remain in charge of the investigation, the Silence of the Cheese will remain deafening.

Get Ready For The Cheesewash

Milwaukee city officials, under pressure for their handling of the flood of same-day voter registrations in the last two presidential elections, have now formed a panel to investigate the issue. However, Greg Borokowski reports that the independence of the city panel leaves a lot to be desired:

Amid new questions about the Nov. 2 election in Milwaukee, a task force appointed by Mayor Tom Barrett to review problems and procedures will launch its efforts today.
Members will dig into an election that featured heavy turnout, huge demand for early voting, a GOP challenge to thousands of addresses and, based on a Journal Sentinel review of election-day “incident logs,” a general frenzy of activity across the city.
But the committee – consisting entirely of city officials – faces critics who question whether it will be able to conduct an impartial review.

As well they should; until the last moment, the panel consisted of seven members, all of which worked for the city and under the authority of Mayor Barrett, a Democrat whose administration has the responsibility for following Wisconsin law. Not only did he stock the panel with people who owe him their jobs, one of his selections for the panel will also provide key testimony. Lisa Artison, the executive director of the Election Commission, will find herself in the almost-unique position of judging her own testimony to the panel. (CQ readers will recall the issue of Jamie Gorelick serving on the 9/11 Commission while having large responsibility for constructing the “wall” that kept law-enforcement and intelligence agents from sharing information.)
Issues with Artison go back to before the election. In a November 1st article, the Journal-Sentinel noted that the Election Commission failed to process three-quarters of the city’s registration entries before Election Day, leaving 15,000 votes in limbo:

As the executive director of the Milwaukee Election Commission, it’s Artison’s job to ensure everything runs smoothly in today’s election.
Yet in recent weeks, the commission has been at the center of controversy, first over how many ballots it needed, then over whether newly registered voters were listing nonexistent addresses.
Mayor Tom Barrett has accused Republicans of stirring up those controversies to disrupt voting in the largely Democratic city.
On Monday, however, Barrett admitted Artison’s staff had not processed 15,000 to 20,000 voter registration cards from newly registered voters. Barrett didn’t learn about the backlog until Monday, when he ordered other city staffers into the election office in a last-minute push to get the job done before voters show up.

Artison had been a controversial nominee last summer for the position:

In July, the council confirmed Artison on a 9-6 vote, the closest margin of any Barrett nominee, with former Acting Mayor Marvin Pratt lobbying against her. Besides questions about her qualifications, observers wondered whether her job was a payoff for work by her and her husband, radio talk-show host Eric Von, on Barrett’s campaign against Pratt.

Artison has hardly acted like an agnostic on the presidential vote, either. She has asserted that nothing went wrong and that the critics have “an agenda at work” in their complaints. Any independent investigation of the Milwaukee vote should focus their attention on the competence and/or the potential malfeasance of Artison and her staff. Having Artison help investigate the matter makes clear the panel’s purpose — to mollify critics by slapping together a whitewash that will eventually tire out those who keep watch on the situation.
The last-moment addition of two people from the comptroller’s office makes little difference to the makeup of the nine-person panel if Artison remains on board — indeed, if the supermajority of the panel all report to Barrett, the die has been already cast. This is a last-ditch effort to head off a state investigation into voter fraud in Milwaukee. Let’s hope it doesn’t succeed.

Racine Election Officials Broke The Law: State Rep

While most of the focus of election fraud in Wisconsin has been on Milwaukee and its 30% same-day registration in two succeeding presidential elections, the smaller city of Racine also had its own problems. State representative Robin Vos accused Racine officials of ignoring electoral law by failing to even attempt the verification of same-day registrants as required by Wisconsin law:

State Rep. Robin Vos accused the city of Racine on Wednesday of violating state law by failing to send out voter verification cards to people who registered to vote on the same day as the Nov. 2 election.
“Racine County residents deserve fair and proper elections,” Vos, R-Caledonia, said. “I’m disappointed to see that the city of Racine isn’t doing everything possible to ensure this happens, especially when Wisconsin law requires it.”
Racine City Clerk Carolyn Moskonas, who was recently appointed to the position, confirmed Wednesday that the city doesn’t send out postcards to same-day voter registrants. She added that there is no record of the city ever sending out postcards to verify the registrations.
“My predecessor didn’t do it, and there’s no money in my budget to do it,” Moskonas said.
At issue are the more than 3,000 people who signed up to vote in the city of Racine on election day. In the weeks leading up to the election, Moskonas and her three-person staff called or sent letters to people who registered to vote.
But since Nov. 2, Moskonas said her staff has focused on processing the election, which saw about 38,000 city residents vote. Moskonas said her office won’t be done with the election until the end of January – long after anything could be done to impact the election.

Note the level of same-day registrations in Racine. While 84,000 of Milwaukee’s 277,000 votes came from same-day registrants, the same process only 3,000 of Racine’s 38,000 votes, or 7.9%. That comes much closer to the percentage we saw in St. Paul with similar laws (9.4%). It’s too many, but a far cry from the 30% that voted in Milwaukee.
Racine obviously ignores state law, and Racine needs to be held accountable. However, the examples of both cities should propel the Wisconsin legislature to eliminate their same-day registration system in favor of something more secure. The ability of anyone off the street to find a legitimate elector (a registered voter in the precinct) to simply vouch for them in order to cast a ballot leaves the state open for mischief and fraud, with no practical way to either stop it or investigate it later. At the minimum, the state should restrict same-day registrations to those with state-issued IDs that list the address from which the voter makes his claim of legitimacy and then treat the ballot as provisional until the name can be checked against the voter rolls.
Until the state legislature acts, this will happen in every Wisconsin election. And until the legislature receives enough humiliation, they won’t lift a finger to change anything.

Wisconsin Doesn’t Follow Its Own Electoral Law

Greg Borokowski continues his dogged pursuit of the Wisconsin voting irregularities in 2004 for the Milwaukee Journal-Sentinel, even though no one in the national media appears to notice, outside of the Washington Times. In today’s paper, Borokowski reveals that despite having one of the most liberal and fraud-susceptible voter-registration systems in the country, the elections boards rarely refer invalid registrations to the district attorney’s office:

In the wake of Milwaukee’s inability to send confirmation cards to some 10,000 newly registered voters, a Journal Sentinel review suggests that a little-discussed – but key – safeguard in election law is not routinely followed.
The provision requires that any confirmation cards that the U.S. Postal Service cannot deliver be sent to the local district attorney’s office for investigation for possible fraud.
District attorneys from around the state said Wednesday, however, that they receive few such referrals – and some did not know it was a requirement. And some cities, including Racine, don’t send out confirmation cards at all. Meanwhile, some clerks have complained to state election officials that the returned cards they do submit are not investigated by authorities.

So we have in Wisconsin a system in which anyone can walk up to any precinct and register on Election Day, as long as they have someone who will vouch for their eligibility. Now we know from Borokowski’s extensive report — he quotes a number of DAs for their reaction — that the chances of a criminal referral from a fraudulent registration approach absolute zero. Most of the Wisconsin DAs appear unfamiliar with the requirement, and all of them have yet to pursue any criminal action.
How many people do you think are aware of the Wisconsin free pass? I would guess that quite a few figured it out in time for the 2000 election, and when nothing happened from that, tried it again for 2004. Milwaukee had 84,000, or over 30% of its entire voting base, register on Election Day. Meanwhile, in Illinois, where the GOP stupidly ran the carpetbagging extremist Alan Keyes against the popular Barack Obama for the Senate and John Kerry had a comfortable lead over George Bush, the Democrats knew they could afford to relocate a substantial number of voters — and Milwaukee is only 100 minutes away from Chicago.
The Wisconsin legislature needs to make some changes in its election law, and the FEC should look into what happened in Wisconsin.

A Mighty Big Coincidence

The Milwaukee Journal-Sentinel continues to cover the questions swirling around the Wisconsin presidential election results, even if the national media yawns at the prospect. Greg Borokowski reports that the 84,000 election-day registrants in the city of Milwaukee just about matches the same number as 2000 (hat tip: CQ reader JB):

Lisa Artison, executive director of the city Election Commission, said the number of cards that could not be sent out this time was comparable to the number after the 2000 presidential election. …
At issue is a gap between the city’s estimate of 84,000 election-day registrants and 73,079 verification cards that were sent, as required by law. …
If the 84,000 estimate of election-day registrants is accurate, 13% of the cards could not be processed. The 84,000 number, about 30% of the 277,535 people who voted in the November election, includes regular voters who may have moved, as well as new voters.
The 10,000 votes questioned represents 3.6% of all voters.

Part of the confusion about this story results from confusing the city and county numbers from Milwaukee, a problem I had myself earlier today. In this case, however, the MSJ specifies that the 84,000 same-day registrations came just from the city — and that they account for 30.2% of all votes cast in the city. How many other cities had almost a third of their registered voters show up for the first time on Election Day?
As it turns out, Milwaukee did the exact same thing in the prior presidential election:

After the 2000 presidential election, she said, the initial city estimate to the state was that 81,000 people registered and voted. That year, the cards that were ultimately able to be processed numbered 73,847.
In 2000, 245,670 people voted in Milwaukee. The 7,153 same-day registrations that couldn’t be processed then represented 2.9% of the total.

That means in just one four-year span, the city of Milwaukee processed 165,000 voter registrations in just two days, both of them Election Day — in a city that turned out 245,000 and 277,000 votes on each day. Either Milwaukee has a 30% turnover in residency over a four-year period, or a substantial fraud occurred in both years.
For this election in my state of Minnesota, which allows the same kind of same-day registration, the number of successful same-day registrants (440,263) only accounts for 15.6% of all votes, and only 12% of the estimated total of registered voters in Minnesota. Even those numbers seemed high to us here, but next to Milwaukee, they’re almost insignificant. In Hennepin County (Minneapolis), the most analogous to Milwaukee County in Wisconsin, the same-day registrations only account for 15.4% of the votes, while in neigboring Ramsey (St. Paul), they comprise an anemic 9.4% of the total number of votes cast.
Do a third of Milwaukee residents change their residency every four years and fail to re-register to vote? Or could it be that the overwhelming number of provisional registrations show a concerted effort to skew Wisconsin presidential elections in one particular direction? I find it hard to believe that two neighboring states could have such a wide variance in residency. Instead of focusing their ire on Bush’s wide margin of victory in Ohio, the Senate should call for a federal investigation into the razor-thin results in Wisconsin for both 2000 and 2004.

More Milwaukee County Demographics

Part of the continuing look at voter fraud in Wisconsin and the lack of media attention, which I called The Silence Of The Cheese
For a bit more analysis on Milwaukee County’s presidential election results, let’s take a look at the population dynamics over the past 13 years. As these articles make clear, Milwaukee County has seen a continuing flight of residents; the county decreased by 19,000 people between the 1990 and 2000 census, and the US Census Bureau estimates that the drop has steepened since. They now estimate that 32,000 fewer people live in Milwaukee County, including 29,000 voting-age adults.
From the Milwaukee Journal-Sentinel of March 8, 2001:

Meanwhile, Milwaukee County Executive F. Thomas Ament was relieved to hear that his county didn’t lose more people. Milwaukee County dropped 2%, from 959,275 to 940,164.
“Obviously, I’m not pleased with losing population,” he said, but this drop is “not as large as many have expected.”
Ament said Milwaukee County’s population decline reflected a trend of “less people per household,” particularly in long-developed areas.

And it dropped further, according to Census Bureau estimates, since the last election cycle. In 2000, the Census found 692,339 adults living in Milwaukee County. In 2003, the Census Bureau estimates 663,517 adults, a drop of almost 30,000 potential voters, and a much steeper decline than in the previous decade.
Now, with that information, one would expect a decline in voter registrations and turnout, at least in real numbers. However, the election turnout has shown a remarkable and unnerving result that belies the residency figures. Ballots cast have increased, and not by a statistically insignificant margin. Here are the numbers for the last three presidential election cycles:
1996: 365,387
2000: 433,537
2004: 482,236
The increase in votes for the 2000 election was 18.7% over 1996 in a county that had had a net decrease in population of 2% over the past decade. In 2004, despite a population decrease of 3.2% in the past three years, the voter response increased in Milwaukee County by 11.2% over the already-ballooned turnout of 2000. While Milwaukee County lost 5% of its overall population over the past thirteen years and accelerating in the past three, votes cast increased an unbelievable 32% in the past eight years.
I’d like to see the statistical analysis which supports that kind of voter turnout improvements, especially with only a 6% increase overall in national turnout between 1996 and 2004. The results point to some other force at play than a monumental increase in civic response among Milwaukee County adults.
UPDATE: This certainly could be one explanation, an e-mail from a Wisconsin poll worker to Dirty Harry at Stranded On Blue Islands:

We have a large low income apartment complex located in the district covered at the polling place I worked. We had more same day registrants for that address than could live there.
We were unable to contest any of those registrations because of a little trick called “Voter verification” It goes something like this: A voter has an out of state driver’s license and wants to register.They are required to present a utility bill or some other form of information that would verify they live where they say. If they don’t have that, a “legitimate elector” can verify their residence.
What is a “legitimate elector” you may ask? Well let me give you an example. Groups from, Vote Now, ACT, and local GOTV would drop off a group of “voters”. One of the drivers would act as the “Legitimate Elector” and say this person was located where he claimed. Presto – that person is a voter.

Do that enough times and the 32% increase in votes from 1996 to 2004 becomes a lot more understandable.