In the summer of 2009, the state of Minnesota conducted a recount that had the nation holding its breath. Would Al Franken unseat incumbent Norm Coleman, and allow Democrats a filibuster-proof majority in the US Senate? After seven months, that’s exactly what happened. Townhall Magazine commissioned me to write an investigative cover piece on how Democrats flipped an Election Night deficit into a recount victory. (At the time, Townhall and Salem had not yet bought Hot Air.)

Rather than skullduggery, I found that Republicans had been outfought after ignoring lessons from a 2004 defeat in Washington. The painful loss provides lessons for Republicans — but have they learned them? Now that recounts will take place in Wisconsin and perhaps other states, it’s time to recall exactly what happened and what Republicans must do to win recounts decisively. Democrats will come prepared to push to the limits of the law, and Donald Trump had better make sure his campaign is prepared to play at the same level.  This will be the first time that the article has appeared since August 2009, and the first time ever on line.

Note: Big thanks to my friend and former colleague Chris Fields, former Townhall Magazine editor and now senior editor at The Blaze, for finding this in his files. (We think he might have some Hillary Clinton e-mails in there, too.)


As the nation celebrated the end of a tortuously long presidential campaign Nov. 4, 2008, and the decisive result that would eliminate any hint of uncertainty of the winner and loser, Minnesota looked into the abyss of a dead heat for its most important statewide race. Thanks to the long coattails of Barack Obama and the determination of Minnesota Democrats to make sure all of their party’s voters’ ballots were counted, Democratic-Farmer-Labor (DFL) candidate, comedian and political pundit Al Franken had battled incumbent Republican Sen. Norm Coleman essentially to a draw. Only a few hundred votes separated the two men that night, and only a few hundred votes separated them in the end—but those few votes would give Democrats a supermajority of 60 votes in the Senate with Franken’s win.

This process would challenge Minnesota’s pride in its clean politics and test its confidence in its electoral system. More, it would eventually show that the upstart politician prepared better than the experienced, seasoned statesman and that Republicans failed to learn lessons from Washington state in 2004 about how to fight recounts. In the end, the difference would come down to how hard each campaign was willing to fight—within the rules—in order to prevail.

In the days immediately after the election, the Coleman campaign was dismayed to see the vote totals change repeatedly—and always in favor of Franken. On Election Night, Coleman had held an unofficial 726-vote lead. The next day, that dropped to an official 477. Within a week, his lead dropped to 215, thanks to errors like the one committed by Pine County’s Partridge Township, where election officials simply left a digit out of the hundreds place for Al Franken’s votes in counting ballots by hand, adding another 100 after the initial count.

However, going into the recount, Coleman held that narrow lead and had every expectation that the recount would not change the results—as history shows, recounts rarely do, although in an election where almost 3 million ballots were cast, the margin of error for a 215-point lead seemed very thin indeed.

In large part, though, that confidence came from the knowledge that most Minnesota voters cast optical-scan ballots, which work in several ways to eliminate error. Voters fill in bubbles next to the candidates’ names and feed the ballot into the counting machines located in the precincts. If the ballot cannot be read, the ballot gets spit back into the hands of the voter, who can request a new ballot to correctly complete. Optical-scan ballots can quickly be counted and recounted mechanically and, unlike punch-card ballots, do not degrade from repeated handling. Counting machines can rapidly complete their work and transmit results back to the counties.

Minnesotans had every expectation that they could avoid a Florida 2000-type controversy and not just have a clean, reliable election, but a clean, reliable recount, if necessary. In that, Minnesota largely succeeded, albeit painfully and not without criticism of its processes. However, the real historical parallel was not the hotly disputed presidential election between George Bush and Al Gore in Florida but a 2004 gubernatorial race in Washington in which Democrats put to use the lessons of Florida—lessons Minnesota Republicans failed to heed.

As previously noted, most recounts do not change Election Day results, but the norm did not prevail in the election between Dino Rossi and Christine Gregoire in 2004. The two candidates for governor in Washington came down to a 261-vote lead for Rossi when the initial election results were certified Nov. 17, 2004, and only 42 votes separated the two when the first machine recount was completed Nov. 30. Gregoire then demanded a manual recount, which took a month to complete. When it was over, Gregoire had captured the lead, and Rossi challenged the results in court.

The Washington election crisis did have key differences from the Minnesota Senate election four years later. While the conduct of the election itself in Minnesota was mostly considered fair, allegations of improprieties had filled the Rossi-Gregoire battle. Much of the focus in Washington fell on the issue of illegal felon voting, which both sides claimed had skewed the results in favor of their opponents.

Both sides took those challenges to court, but Rossi’s team made a critical error. They assumed that the very notion and the alleged scope of illegal felon voting would be enough to convince the judge to throw out the election or, failing that, to proportionally deduct votes from each candidate to correct the size of the alleged distortion. Judge John Bridges did neither. Instead, he stuck to the statute and considered only cases where each side could substantiate specific illegal voting. Rossi’s team had not presented any, where Gregoire’s team found a handful of illegal felon votes for Rossi and presented testimony to support the challenges. As the Seattle Post-Intelligencer put it afterward, Rossi spent millions of dollars and five months of court time to lose four votes and the election.

Democrats grasped a valuable lesson: learn the law, do the groundwork and overwhelm the opposition through massive amounts of vote-specific arguments.

Four years later, the Coleman team faced off against the Franken campaign, but as recount volunteers tell Townhall, they represented two different approaches. The Coleman recount team was staffed by Minnesota veterans used to the collegiality of the state’s political establishment. Franken, ironically, had built a team of seasoned national professionals, people who understood how Gregoire won the election challenge in Washington and had every intention of applying those lessons.

Scott Johnson of the Power Line blog introduced the Coleman cast in an April National Review column: “[T]he attorneys who publicly led Coleman’s team through the recount (local lawyers Fritz Knaak and Tony Trimble) appeared like Pop Warner players going up against an NFL team. When Coleman filed the election contest challenging the result of the recount, he brought in a new lawyer—the respected local criminal-defense attorney Joe Friedberg—to represent him in the election-contest trial. This has not proved a winning formula.”

In contrast, Franken had his team prepared for a recount even before Election Day and had assembled, well, an NFL team to conduct it. In fact, as Doug Grow reported for the left-leaning, Franken’s team began strategizing for a recount when the race became razor-thin two months earlier:

“Many not only give [campaign manager Stephanie] Schriock credit for the success Franken had in closing the gap during the campaign but in being so well positioned to succeed in the recount process. The first recount plans were being drawn up as soon as September. The day after the election, Schriock and the entire Franken staff had put the plan in place.

“[Marc] Elias, the top legal adviser for Team Franken on the recount, was involved before the polls were even closed. Schriock also had set in motion her plan that, at its peak, had 2,000 people working as either volunteers or in paid positions, to be in the field to observe recounting at polling places across the state. All were trained.

“There was never—not for a second—a letdown between the campaign and the recount.”

The Franken team had a clear mission. Since they were behind in votes, they needed to either find new Franken votes or discredit Coleman votes in order to close the gap. Learning from the Gregoire campaign, they trained their volunteers to understand the limits of Minnesota law and to aggressively challenge ballots.

Coleman’s team, on the other hand, had a much more delicate mission, as they explained to their volunteers. Each of the precinct workers interviewed for this story had the same description of the instructions given by the campaign: Do not get overly aggressive in challenging ballots. They did not want to be seen as the campaign that “disenfranchised” Minnesota voters, as successful ballot challenges do by removing ballots from the count.

One Coleman volunteer explained the instruction as an explicit message from Team Coleman that “we don’t expect to be in the business of suppressing Franken votes, and we’re not trying to fi nd new Coleman votes. … Don’t go out of your way to make what we think will be frivolous challenges.”

This instruction came specifically about overvotes—instances where a voter filled in two or more bubbles on the same race, which would have led the counting machines to reject the ballot for that race. Franken’s team latched onto the overvotes and tried to argue on as many as possible that the intent of the voter was to support Franken. Coleman’s team knew from the beginning that Franken’s volunteers would use this strategy, another Coleman recount worker said, but believed that “we should not engage them like that.”

But according to Minnesota law, as the state discovered during the process, the question of voter intent on overvotes is a legitimate area of challenge in a recount. In this case, it appears that both campaigns understood the parameters of action, but Coleman’s team simply didn’t want to avail itself of the entire range of action allowed by Minnesota statutes. They trained their recount volunteers to engage only on the most obvious cases and to refrain especially from giving the appearance that the Republicans wanted to invalidate ballots on a massive scale. Predictably, this led to many missed opportunities for Coleman challenges.

Because of the training received, the GOP volunteers assumed that many of the Franken team’s challenges in the precincts were invalid and would be tossed out by the Canvassing Board, the bipartisan panel that ruled on each challenged ballot. They were shocked to see the types of challenges later upheld by the panel, and they lamented the passive manner of the Coleman team’s recount effort, especially in the opening days. One volunteer estimated that he could have produced between 10-20 ballot challenges himself that the Canvassing Board would have upheld, based on their later rulings.

Franken’s team didn’t rest on its organizational edge during the recount, either. They gathered information from all the precinct recounts, even using tally sheets to note trends on questionable ballots, and apparently analyzed them overnight. Coleman volunteers recall seeing coordinated efforts to focus on new issues almost every day from their counterparts during the process.

“These guys seemed to be professional recounters,” one Coleman worker said, and while that wasn’t literally the case, Franken got some seasoned precinct veterans—from Washington. “A lot of them [lawyers and other volunteers] had been involved in the Gregoire recount,” especially in the Minneapolis precincts, said the Coleman worker. Because Minneapolis was a heavily DFL city, the Franken campaign expected to reap more votes there, a hope that largely went unfulfilled, according to another volunteer.

Team Coleman had recount experience—in Minnesota. Rep. David Minge, DFL, lost a close election to Republican Mark Kennedy in 2000, which triggered a recount. Unlike the Franken-Coleman contest, it did not attract national support staff, and that recount was mostly a collegial affair, with both sides trying to exemplify Minnesotan values of fairness. The Coleman team had a number of Minge-Kennedy veterans but did not have the kind of hardball experience that Franken got from the Gregoire team.

The hand recount lasted 14 days, from Nov. 20 to Dec. 4. During that time, the volunteers never got an explicit message to get more assertive in their approach, but a sense of urgency eventually made it into the precincts. Some of that urgency came from a natural inclination to give a tit-for-tat reaction to the more aggressive Franken volunteers, “typical game strategy,” as one put it.

Eventually, Coleman managed to catch up to the count of challenged ballots, but as the volunteers point out, the quality of the late challenges were likely not as good as the opportunities missed earlier in the process. That would put Coleman at a serious disadvantage with the Canvassing Board. By that time, however, the die had been cast.

At the end of the hand count, the Canvassing Board took center stage. Minnesota Secretary of State Mark Ritchie, DFL, a man known for his partisan bent, appointed a well-balanced judicial panel to hear the recount challenges from both campaigns. Two of the judges had been appointed to the bench by Republican Gov. Tim Pawlenty, one by independent former Gov. Jesse Ventura and another who had won her seat on the bench by election. For the most part, the panel took its job seriously enough that only a handful of the hundreds of decisions made on ballots were not unanimous. Even Ritchie appeared to sense that partisan actions at this juncture could create chaos in the process and put the entire effort in jeopardy of getting rejected entirely in the expected legal contest to come.

The first action from the board was a demand to both campaigns to reduce the number of challenges. Initially, more than 6,000 ballots had been challenged by both teams, and eventually they brought it down to about 1,500.

Immediately, Coleman ran into problems with the Canvassing Board. The team of aggressive, well-trained Franken volunteers had its match in the panel that intended to try to discern voter intent, if at all possible, rather than adopt the more “common-sensical” approach urged by Coleman trainers to their volunteers. While the Canvassing Board kept an eye towards consistency, they also felt a mandate to determine voter intent, especially on overvotes, a role given them by Minnesota statute despite the rather clear and unambiguous nature of the optical-scan system. That mandate fit perfectly with the Franken approach, which used the limits of the law to argue for overvotes to count for Franken rather than get tossed or counted for Coleman—an effort not matched by their opponents.

By the time the Canvassing Board completed their work, Franken had taken the lead by 49 votes—and the question of absentee ballots had arisen. Franken alleged early in the recount process that thousands of absentee ballots had been erroneously rejected and wanted the recount process to address that. Coleman objected, stating that the question should come up in the contest process, which comes after the recount in Minnesota law.

Inexplicably, the Minnesota Supreme Court on Dec. 18 ordered the recount process to address the issue and declared that the campaigns should try to agree on a standard for accepting erroneously rejected absentee ballots rather than have the state enforce the standards themselves. As Scott Johnson notes, the Coleman campaign should have stood its ground at this point:

“The Minnesota Supreme Court held that absentee ballots identified by local officials during the recount as wrongly rejected should be included in the recount subject to agreement of the parties (and also subject to the possibility of sanctions on the parties’ lawyers for withholding agreement in bad faith).

“The Coleman campaign had argued (correctly, in my view) that previously excluded absentee ballots should not be included in the recount. The Coleman campaign was caught flatfooted by the Minnesota Supreme Court’s December decision. Coleman should not have agreed to the inclusion of a single one of these ballots until he secured some agreement on the uniform treatment of absentee ballots. Instead, Coleman’s team agreed to the inclusion of 933 of 1,346 previously rejected absentee ballots identified by local officials during the recount as having been improperly rejected.”

Johnson says that the court had practically handed the Coleman team the election at this point: They could have used that ruling to block further consideration of absentee ballots until the contest phase. Just a day later, however, the Canvassing Board calculated that Franken had jumped into the lead based on ballot challenges. Nevertheless, the inclusion of a portion of the absentee ballots put Franken’s lead from a tenuous 49 to a relatively stronger 225—where it would remain for more than three months.

On Jan. 5, 2009—two months after the election—the Canvassing Board certified the election results, showing that Al Franken had unseated incumbent Norm Coleman by that 225-vote margin. A day later, Coleman fi led his election contest, the next legal step in fighting the results, and promised a tough campaign to “count every vote.”

Coleman’s court case rested on several premises, but primarily that Minnesota’s counties used varying standards in accepting absentee ballots as valid. As a remedy, Coleman wanted more of the rejected absentee ballots opened and counted, which he insisted would provide him with the extra votes he needed to win.

Unfortunately for Coleman, his attorneys took much the same tack as Dino Rossi’s 2004 legal team. Instead of finding individual ballots and making individual cases for inclusion, the Coleman lawyers attempted to focus on “categories” of ballots to gain acceptance of large blocs of previously rejected ballots, many of which had been demonstrated to fail against the statutory requirements for inclusion. Those statutory requirements include matching addresses between registration information and the absentee ballot, a proper and valid signature on the external envelope of the ballot, validation of the voter registration for that precinct and the establishment that the voter had not voted in person or with another absentee ballot.

Eric Black, who covered the trial for, explained that Coleman’s team argued for a consistent low standard of acceptance for absentee ballots in order to get the broadest number of new ballots to count. “At one point, they were arguing that if you were alive on Election Day and at a Minnesota address, your absentee ballot should count, and that that was the only way you could get it to a standard that could be applied uniformly,” Black recalled. “It would require you to ignore a great many things that the statutes require.”

Coleman’s team argued that his team deserved that remedy based on the lack of uniformity of treatment of absentee ballots between the counties. Because some invalid ballots got counted in Franken-leaning counties, they said, the only remedy was to count all of the ballots that matched the lowest denominator of acceptance on Election Night. In essence, they argued that the judges should deliberately violate the law in order to remedy earlier violations of it, whether intentional or accidental.

At the same time, Franken’s team worked to find more examples of absentee ballots they could request for opening during the contest, but they went after cases on an individual basis. Again, this parallels Rossi/Gregoire, where the Democrats understood that they needed to remain on offense even in the contest phase. Instead of trying to convince the judges to allow ballots outside of the law, Franken’s lawyers found more ballots from the rejected pile to argue for inclusion within the law.

The three-judge panel hearing the contest unanimously found that Coleman had failed to prove his case that extraordinary relief would substantially change the outcome of the election. They bluntly wrote that Coleman’s legal team failed to prove its case at all, saying, “There is no evidence of a systemic problem of disenfranchisement in the state’s election system, including in its absentee balloting procedures.” To add insult to injury, they had added 87 more votes to Franken’s lead nine days earlier, widening the gap to 312 votes, after opening up 351 absentee ballots, which is all that the panel said had been proven necessary to count by either side. The judges had, in the end, focused on individual ballots rather than categories and generalities, just as Judge Bridges had done in Washington four years earlier.

Had the contest succeeded, would Coleman have won enough votes to upend Al Franken? The Coleman team asked the chair of economics at St. Cloud State University, King Banaian, to testify in support of their statistical analysis. Banaian, contacted only after the Coleman team filed the contest, had just a week to prepare and presented evidence that the distribution of rejected absentee ballots showed something other than chance at work, which would have bolstered Coleman’s insistence on extraordinary remedies. Franken’s team successfully challenged Banaian’s status as an expert on statistics and moved to exclude his testimony.

That might have been just as well, because Banaian’s study of rejected absentee ballots indicated that Coleman had a very slim chance of succeeding even had they opened the nearly 5,000 ballots Coleman first requested. Initially, Banaian calculated that Coleman could argue for the inclusion of almost 2,000 absentee ballots of the 12,000 rejected on Election Day, based on higher rejection rates in some counties than the state average for that election. Assuming that the vote distribution matched that of the general vote in each of the counties, Coleman could gain only 94 votes—far short of what he needed.

Banaian says, “That is why, in my opinion, the Coleman team requested opening 5,000 ballots instead.” But that would have generated only a net gain for Coleman of 166 votes, assuming a matching distribution once again— still short of the final 312-vote gap. Banaian puts the likelihood of winning enough votes out of the 5,000 ballots to overcome even the intermediate 225-vote gap at “only 0.8 percent or about 125-to-1 against.” Coleman’s team had pinned those hopes on a motion to reverse the inclusion of what they alleged were 133 double-counted ballots. Had the contest panel agreed with them and had the count not moved to 312 votes, Banaian still put the likelihood of reversing the election at “about 3-to-1 against,” not a convincing case for any judge to throw out the statutory requirements for absentee ballots and allow correctly disqualified votes into the count.

Norm Coleman appealed his case to the Minnesota Supreme Court to no avail and dropped his legal case June 30. Al Franken took his seat in the Senate a week later.

The truth about the Minnesota election is that it was statistically a dead heat, which magnified the problems inherent to all elections. Most elections do not have close enough outcomes for the irregularities to matter. Even the best systems in the best states have their issues, and while Minnesota will work to improve the shortcomings exposed in the recount, it can never be made so perfect that a close election will have no problems in counts, absentee ballots and other issues.

But was the election really “stolen”? Republicans around the country repeatedly alleged that Franken had cheated in the recount, but the Minnesota Republicans actually involved tell a very different story. “They [Franken’s team] looked at it in the way extraordinarily good attorneys look at a case,” one remarked. “Our clients have interests. It’s our goal to get those interests pushed to the fore. Our job is not to be objective or fair.”

The legal analogy is a good one, another volunteer confirms. In the Minneapolis precinct in which she worked, she said that at times the Franken campaign seemed to have “10 lawyers for every volunteer working.” In fact, the precinct officials had to put tape on the floor to keep the lawyers from interfering with the recount process. Another volunteer estimated that in his precinct Franken had eight to 10 lawyers for every Coleman attorney.

This disparity put the two different missions in stark relief. The Franken campaign saw this as an adversarial process not unlike a court case, while the Coleman team saw it as something else entirely. Rather than steal the election, the lawyers were there to make their case on each and every ballot. The message from the Coleman team was that they were there to ensure overall fairness, acting more like judges or referees than the counterpoint adversaries to the Democrats.

Franken did not “steal” this election in the recount process. The Republicans lost this election because they failed to learn from their mistakes in Washington four years earlier and because they failed to realize how serious Democrats are about winning recounts for significant political offices. The Democrats have made it a major adversarial business, working much like defense attorneys or litigators do to find every legal advantage available to them in order to prevail.

As long as Republicans continue to handle election recounts as an afterthought, act like referees rather than stakeholders and fail to match resources with their opponents, they will lose these close elections. And if Republicans insist on believing that fraud rather than their own mistakes led to the recount loss, they will fail to learn this lesson—and lose the next one as well.