MANY PEOPLE HAVE WONDERED WHEN the campaign really returned to the momentum it had throughout the spring after the setbacks in October. To my mind, it really began in November and was propelled by four critical events. The first was our entrance into the paid media market. The second was the CBS presidential debate. The third was the heavy-handed misplay by the DNC in December when the chairwoman tried to shut down the campaign. And the fourth, which was crucial, was Bernie’s ability to campaign full time.
* * *
Having ceded the Clinton campaign control of the airwaves for three months, we finally aired with our first ads in Iowa and New Hampshire. We finally could speak directly to voters without the media filter—to introduce Bernie as a person and to present to the nation his powerful message of building a more equitable society.
There is a lot of criticism that television ads (too often negative) are used as a substitute for a grassroots campaign that has a real message and an authentic messenger. I share that criticism, to the extent that campaigns devolve into air wars with little substantive debate or involvement by the people who actually are going to have to vote for one of the candidates. Bad candidates may have the appearance of strength because their monetary advantage convinces the political and media establishment that their ability to dominate the airwaves will assure victory.
But I can also say that television advertising—when used as a complement to, rather than a substitute for, a grassroots campaign—is extremely powerful in communicating with voters. In fact, empirically, it is the single most powerful tool. Some in the progressive movement have become so disgusted with the role of big money (a feeling I most certainly share) that they advocate abandoning television. It is hugely expensive. This cost is seen as one of the drivers of the constant quest for campaign dollars. That’s a great position, if you want to make a point. It’s a horrible position if you want to actually win an election.
As we were to learn, being wildly overmatched on television cannot be overcome with a ground game. You have to do it all. You have to build a grassroots effort and you have to be on television. And online and in the mail and a bunch of other places. You don’t need to have the greatest number of ads, but you have to have enough. What is enough depends upon the contest; and is not always clear before the voting takes place. In a presidential campaign, enough is a lot.
Of course, the real answer to this problem is reforming the campaign finance system by overturning Citizens United and implementing various public matching or public financing proposals. This will create a more level playing field in terms of candidates’ ability to speak to voters, because more candidates will have access to TV and a host of other media. Big money won’t monopolize the airwaves. Candidates who can mobilize a grassroots following will be able to compete.
* * *
The second big turning point was the second Democratic presidential primary debate, held at Drake University in Iowa, although it certainly didn’t seem like that was going to be the case the afternoon before the debate. Enter the White Rabbit.
Before every debate there is a large media contingent waiting at the debate location for the event to begin and to interview people in the spin room afterward. During this time the campaigns try to pre-spin the press. Tad describes it as releasing a white rabbit in the room—and then the media chase it around for a few hours before the debate. If they chase your white rabbit, you influence the pre-debate story line. If a campaign is unscrupulous, it can put out a half-baked story that will get pushed out by the media online and on social media before the debate, and there will never be a chance to correct it. Coverage of the debate itself will overtake the media’s interest.
On November 14 our opponents released their white rabbit, which the media proceeded to chase. Here’s how it went.
The format of each of the debates was negotiated extensively by the campaigns and the sponsoring media organizations beforehand. The media organizations have the ultimate say, but all the campaigns have input. In each of the negotiations, it was our view that the candidates should have an opening statement and a closing statement. The longer the better—it gives each candidate more time to speak directly to the people who matter: the voters.
The media hates the statements, preferring the give-and-take of the debate because it makes for “better TV.” They tend to push back on lengthy statements, particularly opening statements. From their perspective, there should be blood on the floor almost immediately.
In the case of the November 14 CBS debate, all the parties involved, including CBS, agreed that each of the three remaining Democratic candidates would have a ninety-second opening statement. Now, ninety seconds is not really that long, but getting the media to agree to statements of a minute and a half each was a huge success.
On November 13, ISIS terrorists launched a cowardly attack on the people of Paris. It dominated the news, and of course we knew that it would become a huge part of the discussion during the debate. In a conference call the afternoon before the debate between the campaigns and CBS, the network informed the campaigns that it was going to eliminate the ninety-second opening statements and replace them with a ninety-second open-ended question about the Paris attacks.
Mark Longabaugh, who was representing our campaign on the call, pushed back. Of course Paris would be a topic of discussion, but it didn’t have to happen at the expense of the opening statements. The moderators could just substitute a Paris-related question for one of the other questions they intended to ask. After all, the campaigns didn’t have the questions ahead of time (at least not at this debate; more on that later), so the networks could always change up what they were going to ask.
After the call, one of our opponents promptly picked up a small cage in their work area, walked to the media room, and opened the cage door. Out darted the fluffy, long-eared allegation that Mark was in fact trying to limit debate discussion of the Paris attack—because of course Bernie was wholly unprepared to talk about foreign policy. For the record, the O’Malley people would later blame the Clinton campaign. The Clinton people blamed the O’Malley campaign.
The media immediately went into a frenzy trying to grab the little creature. I had the misfortune of being in the spin room when it was let loose. Annie Karni of Politico asked me about it. I told her what I understood from Mark—that we had pushed back on the attempt to eliminate the opening statements and had been successful. However, what was being put out online was that I’d said we were successful in limiting discussion of the Paris attack. C’mon, people. You didn’t think we knew we would be talking about Paris the day after the attack?
Social media lit up, and not in a positive way. Our media team saw it immediately. Symone Sanders and Sarah Ford arrived on the scene to extricate me. I called Mark to confirm the details. Meanwhile it was spreading like wildfire. Once one person puts up the story everyone else in the media is compelled to follow suit.
Our defense came from an unlikely source: the DNC. Erik Smith had been given the unenviable job of being the DNC’s point person between the campaigns and the sponsoring news organizations over debate logistics issues. Mark called Erik, who had also been on the conference call, and asked him to help set the record straight, which he did.
Some of the media—for instance, CNN’s Brianna Keilar and Dan Merica—incorporated Erik’s comments in updates to their stories:
But a DNC debate coordinator tells CNN that the argument on the call earlier Saturday was not about whether the debate would focus on Paris, adding that it is unfair to say Sanders’ campaign did not want to debate foreign policy.
“No, no,” the coordinator said when asked about whether Sanders’ aides were worried about debating Paris.
“What happened was that we have had the format of the debate finalized for several weeks,” the coordinator said. “CBS wanted to get rid of the opening statements and start with a 30-second answer on Paris.… It was never a conversation of whether we talk about foreign policy or not. They just wanted a longer opening comment.”
Annie Karni, who was more responsible than any of the rest of the media for creating the scramble for the white rabbit that day, left her story uncorrected. Mission accomplished by our opponents.
The big lesson is that even reputable journalists may stand by stories pushed by unnamed partisans of a campaign well after those stories are debunked by a neutral third party. They might just turn it into a “one person said this and the other person said that” story. The allegation is still hanging out there. Journalists more focused on clickbait don’t even bother.
This is the environment in which Donald Trump thrives. He consistently lays out one untrue statement after another, and the media then goes about finding people to say it’s not true—turning a false claim into a “President Trump says this and these other folks say the opposite” story. The result is that those predisposed to believe Trump do, and those predisposed to disbelieve him do. But his original outlandish statement gets amplified over and over again as it is repeated on cable news.
It’s fair now to ask how, given these developments, the CBS debate gave our campaign new momentum. As the afternoon ended, the scoreboard read: Untrue and Discredited Attacks: 100; Truthful Information for Voters: 0. But social media giveth as well as taketh away. As the debate got under way, it became clear that we had only partially won in the conference call. The candidates got only sixty seconds to open. John Dickerson invited the candidates in that time to “share your thoughts about the attacks in Paris, and lay out your vision for America.” In the context of the earlier allegations, talking about anything other than the Paris attacks during the opening remarks (which Bernie did) invited all the talking heads to say that he was avoiding foreign policy. In the Clinton work room, no doubt, staff were busy adding that criticism to the talking points to be taken up in the spin room after the debate.
The debate at Drake was much more contentious than the earlier CNN debate in Las Vegas. The candidates sparred over foreign policy, the minimum wage (Hillary Clinton came out against raising the federal minimum to $15 an hour), and guns (with O’Malley going after both Clinton and Sanders). But the critical moments came when the debate turned to the issue of money that Secretary Clinton had received from Wall Street in the form of speaking fees and campaign dollars.
Bernie and Hillary started to debate the relative merits of their financial reform plans. Then Bernie made the point that large financial interests make contributions because they believe there’s a benefit:
Here’s the story. I mean, you know, let’s not be naive about it. Why do—why, over her political career has Wall Street been a major—the major campaign contributor to Hillary Clinton? You know, maybe they’re dumb and they don’t know what they’re going to get, but I don’t think so.
(All the quotes from the CBS debate are from the transcript published by the Washington Post.)
Bernie’s articulation of the fact that big corporate donors don’t give out of pure generosity threw Secretary Clinton completely off her game. Her response was unbelievable:
So, I represented New York, and I represented New York on 9/11 when we were attacked. Where were we attacked? We were attacked in downtown Manhattan where Wall Street is. I did spend a whole lot of time and effort helping them rebuild. That was good for New York. It was good for the economy and it was a way to rebuke the terrorists who had attacked our country.
The only real answer would have been to say that Wall Street can expect whatever they want, but they are not going to get it from me. (That’s essentially the line that President Obama’s people used.) It’s not a great response, but it’s the only one there is. Instead, Secretary Clinton used the “9/11 made me do it” defense.
What should have been a political ground rule double was turned into a grand slam when CBS’s Nancy Cordes later asked a follow-up to Secretary Clinton based on the following Twitter post from an Iowa voter, which was displayed on the screen: “I’ve never seen a candidate invoke 9/11 to justify millions of Wall Street donations until now.”
I looked over to Tad. “That was devastating,” I said. Hillary Clinton would never use the 9/11 defense again, but the questions surrounding her Wall Street ties, particularly the high speaking fees and her refusal to release the transcripts of her speeches, became a real and lingering liability. Later, leaked emails revealed all too clearly why the Clinton campaign preferred to take the political hit for not releasing the speech transcripts rather than just making them public.
As we headed to the spin room, we knew that there would be some media focus on the question of foreign policy chops, but we all knew that the Wall Street gaffe would be the one that voters remembered. It completely validated the central message of our campaign: Americans were hurting because of a rigged economy held up by a corrupt system of campaign finance.
* * *
December 16, 2015, started as any other crazy Wednesday in Bernie Sanders’ run for the Democratic nomination. That is, until the misdeeds of a few staffers thrust our campaign into what looked like a serious crisis. By the end of the week, the bumbling intervention by DNC chairwoman Debbie Wasserman Schultz against the campaign gave us perhaps the biggest positive bump to that date.
The crisis first came to my attention when Rich Pelletier, then national field director, called me to let me know that a data firewall failure had occurred in the DNC’s voter file, and that other campaigns’ information was made available to some of our staff. I instructed Rich to investigate and to inform all staff that they were not to make use of any other campaign’s data. At the time, I believed this to be a repeat of a data firewall failure that had occurred back on October 7.
The next morning, in a fit of anger, the chairwoman unilaterally, and in violation of our contract with the DNC, shut off our access to the voter file we had purchased and all the data that our campaign had put into it. As I described it two days later at a national press conference, Debbie Wasserman Schultz had imposed the death penalty on our campaign. Here’s why the voter file was so important.
The DNC maintains a central file of all voters in the country by state. The file contains voters’ contact information and a wealth of other data—including in which past election the voter has participated, issues that a voter is interested in, and, in many cases, whom the voter supported in various past elections. The file is a critical asset that allows candidates who pay to access it for the purpose of obtaining this information. It is particularly important in early primary and caucus states because it allows you to target voter communications being sent by staff and volunteers. And it is important for modeling—more on that below.
One of the important conditions of using the file is that the voter information you input will become incorporated into it and made available to future candidates once your election is over. This keeps the file fresh for future users. During the course of a campaign, when a campaign identifies that a voter is supporting one or another candidate, that information is loaded into the copy of the voter file that the campaign is using. However, these voter IDs are not shared with anyone else while the campaign is ongoing.
Voters are rated on a 1 to 5 scale, with 1 being a strong supporter, 2 being a leaner, 3 being undecided, 4 being a leaner toward another candidate, and 5 being a strong supporter of another candidate. In a two-person race between Sanders and Clinton, a Sanders 1 was a Clinton 5. This information is used to efficiently manage expensive communications to voters, such as mail and telephone calls. No matter how much money a campaign has, it cannot reach every voter with the same level of intensity. Voters who are a 1 do not need to receive communications designed to persuade voters but certainly would be targeted for Get Out the Vote (GOTV) efforts. On the flip side, a 5 would not receive either persuasion or GOTV communications. Those who are 3s would be obvious targets for persuasion communications—to move them to a 2 or a 1. And one person’s 3 is every candidate’s 3, so that person can expect to receive a lot of communications from all the campaigns.
A single outside vendor, NGP VAN, provides the software necessary to access and manipulate the Democratic voter file process, and almost all experienced campaign field staffers are familiar with using “the VAN.” This voter identification data is also at the heart of a campaign tool called modeling. Modeling is a process by which every single voter in a state’s voter file is assigned a score based on how likely it is that he or she will support a given candidate. The modeler starts with the actual results of voter identification. The modeler then combines this known information about some voters with polling results and commercially available data about all voters, but only information that the modeler believes is predictive of how someone will vote.
This is where the true judgment of the modeler becomes important. There is endless information available about all of us in commercial data banks, like what kind of car you drive, where you shop, what magazines you subscribe to, what you buy, and so forth. Some of that—for instance, the fact that you own a blue car—may be irrelevant to your voting patterns. But the fact that you own a minivan may be highly relevant. (For the record, I am not a modeler, so I have no idea whether owning a minivan is relevant or not. I’m just trying to illustrate a point.)
In a general election, whether you identify as a Democrat or as a Republican is highly predictive of how you will vote. Independents who voted in the Democratic nominating process (in those places where it is allowed) overwhelmingly supported Bernie.
The modeler then combines all the information from the various sources and assigns a predictive score for all the voters in the file, even Republicans. This again allows the campaign to target votes. The campaign will focus on contacting voters (for whom it has no actual information) with a high score, because the model predicts that they will be more open to supporting your candidate. As more actual voter ID data comes in over the course of the campaign, the model is updated.
The modeler can also produce predictive models of behavior other than voter preference. They can score everyone on such factors as how likely they are to talk with a volunteer caller on the phone, how likely they are to answer their door in the evening, or whether they will even show up to vote at all. This last determination is critically important. Voters who score high in their support for your candidate, but low in the likelihood of voting at all, need special attention to ensure they show up at the polls.
When the model is run against the voter file, it produces a list of tens of thousands or hundreds of thousands of voters and their respective scores. It is this resulting list and the attached scores that are the useful product of the modeling and allow campaigns to put them to use.
In many cases, campaigns will use the modeled results to turn out voters who have a high score, even if the campaign does not have actual information about a particular voter. In a perfect world, models are about 70 percent accurate. In a best-case scenario, 30 percent of the calls to turn out voters are going to supporters of one’s opponents.
Of course, if the modeler is not sufficiently skilled, he or she will include a lot of data in the model that is not predictive of voter behavior at all. And, as a result, the model will not produce more efficient use of campaign voter outreach assets like phone calling or door knocking. Worse even than that, the campaign’s GOTV calls based on modeling will reach greater percentages of an opponent’s supporters. Our modeling was done by Ken Strasma, Andrew Drechsler, and others at a firm called HaystaqDNA—the best in the business.
This background is not meant as an exhaustive discussion of the voter file and modeling. Rather, it is important to understanding the drama in mid-December. Following my conversation with Rich Pelletier about the data breach, I received an email from the DNC at 6:25 that evening. The email stated that the DNC was suspending the VAN access of five of our employees. The email was copied to the DNC’s legal counsel. I discussed the issue by telephone with Lindsey Reynolds, the DNC’s chief operating officer. We agreed that the Bernie 2016 campaign would assist the DNC with a complete review of the matter. But I wanted to have time to consult with our counsel, as the DNC was already involving its counsel. We agreed to communicate the next day. Our final conversation occurred after 8:00 p.m. These discussions were perfectly calm and businesslike, with no hint of distress on the part of the DNC. (It did raise a red flag, however, that the DNC was already bringing in their legal counsel—counsel who happened to be at the same law firm as the Clinton campaign’s.)
As the day ended, as far as I was concerned, the matter was under review. We had agreed to cooperate completely. The five staffers who were under suspicion had their VAN access temporarily suspended. And I knew that no Clinton data would be used by our side. It’s one of the advantages of a leaner, less bureaucratic organization. If this information—and at the time we had no idea what it was—was going to be used for any real purpose, I would know about it.
The next morning began like every other on a presidential campaign. Seemingly endless phone meetings, a huge pile of emails, and a ton of action items that had to be dealt with. The data breach was certainly on the list, but given that it was under active review internally, it didn’t require my personal attention that morning until more information was available. Apparently Debbie Wasserman Schultz felt otherwise.
Wasserman Schultz called Bernie personally. She also took it upon herself to suspend access to the VAN for the entire Bernie 2016 campaign—an action directly at odds with the express language of our contract with the DNC. After talking with Bernie, I called Wasserman Schultz myself. Amy Dacey, the top DNC staffer, was also on the phone.
Wasserman Schultz informed me that a new incident report was being created by NGP VAN about what they believed had happened. I asked to have that document to help with our own internal investigation. I contacted her by email at 2:58 p.m., asking again for the new report. At 3:25 I emailed again because it had not arrived. It finally arrived at 3:53. While the DNC staff were obviously eager to hear from me, they were in no hurry to communicate in the other direction. At 5:08, I emailed Amy Dacey a full report of the actions we had taken and the status of our investigation.
I spoke to Amy by telephone, and she said that it would be hard to get in contact with Wasserman Schultz to discuss the report, and that Amy had no idea when our access to the VAN would be restored.
The impact of the VAN shutdown was immediate. Volunteers were showing up all over Iowa and in other states to make phone calls to voters. But it was impossible for them to do their work without access to the VAN. Volunteers had to be sent home. Particularly in the complicated caucus environment in Iowa, direct voter contact is the lifeblood of the campaign. Shut it off and the campaign dies. The Iowa caucus was a month and a half away. Debbie Wasserman Schultz was giving the Clinton campaign extra days of voter contact at a critical time. If our VAN access was not restored, our Iowa efforts would die.
Meanwhile, information about the data issue was leaked to the media by either the DNC or the Clinton campaign. Needless to say, the Clinton spin operation kicked into high gear. My counterpart on the Clinton campaign, Robby Mook, engaged in some hyperbole, saying that Sanders’ staffers “had access to the fundamental keys of our campaign.” Some in the media who knew less than the reader of this section now does about modeling and the voter file were quick to pile on to the Clinton “aggrieved” narrative, often completely misstating the type of data at issue or its relevance. The story was just too complicated for the modern internet news cycle. Not the media’s finest hour.
What every review of the matter concluded, including our own internal review, the preliminary results of which were detailed in my December 17 email to Amy Dacey, was that no information had been downloaded.
What did happen was that the firewall that is supposed to exist between the various campaigns’ data came down briefly. During that time, some staffers were able to access Clinton models, but not the underlying voter ID data. The staffers saved this modeling data on the NGP VAN system itself but did not download it to their computers—except for an index, which listed the names of the models. A list of model names is valueless, because models produce lists, as we have seen, of tens of thousands of names when run against the voter file. It’s not the name of the model that has value, it’s the search results. Each of the searches in this case was done about a minute apart. It would have required superhuman ability to memorize a list of that size in a minute’s time. And the lists created were deleted by NGP VAN itself shortly thereafter and never exported.
Moreover, we would not have used information from the Clinton campaign’s modeling if we had had it. I was convinced then and am now that Ken Strasma, Andrew Drechsler, and all the folks at HaystaqDNA who had done the modeling for Barack Obama’s 2008 primary campaign were producing far more accurate models that the Clinton campaign. Why in the world would we have substituted the professional judgment of our modelers for an inferior product? My confidence in Haystaq and pollster Ben Tulchin would be confirmed during the Iowa caucus, when our projections proved right and the Clinton campaign missed the mark substantially.
That’s not to say that we took the matter lightly. We quickly fired the most senior staffer involved and put two others on administrative leave and forbade them from coming into the office or having any contact with the campaign. Even though no information had been downloaded, their snooping across the firewall was unacceptable.
Yet Debbie Wasserman Schultz refused to restore our campaign’s access to the VAN. This was not some broad-based conspiracy at the DNC. Later, Donna Brazile told us that DNC executive committee members had urged the chairwoman to restore our access. Wasserman Schultz refused.
We had done everything we could do. We fired people. We provided a complete report of our findings. We agreed to cooperate with any further reviews. But none of it mattered. Soon the weekend would be here. Weekends were important and busy times for our volunteer-based campaign. If we did not have our data back before the weekend, we would have lost almost 10 percent of the days remaining before the Iowa caucus.
It became clear to us that this was a deliberate attempt to torpedo our Iowa efforts. Our contract with the DNC required them to give us ten days’ notice to cure any breach. Certainly, they had not done that. The fact that either the DNC or the Clinton campaign leaked the incident was further evidence that what they wanted was political points and not a real resolution.
This was galling. Back in October 2015, our modelers had discovered that when they downloaded Sanders voter ID data into the models, the VAN also downloaded Clinton data. They had immediately segregated the tainted downloads into a password-protected file and notified the DNC about it. We started modeling in October. The Clinton campaign surely started earlier. Did they have access to our voter IDs throughout the summer and early fall? The DNC said no. I would say there’s better than even money they did.
But we may never know. Despite express promises to investigate this October incident, it has never been done, to my knowledge. The point here is that we never ran to the media in October. We tried to resolve what we believed to be a technical issue in good faith, without taking unfair advantage. Turns out we were the only party operating on that basis.
Staffers who had much more experience with the VAN alerted me that firewall problems were common occurrences. One staffer who had worked for a state party told me that often when NGP VAN did a software update, they received voter file data from other states. Another reported that on a statewide campaign, the opponent’s voter ID information had always been available.
I called Robby Mook personally during this flap. I wanted him to hear from me that we did not have any information and that we would not use it if we did. I was blunt. “We want to beat you, and beat you badly,” I told him. “But not like that.” I didn’t expect him to believe me or to say that he believed me even if he did. They thought they had an advantage and they were going to fully exploit it. I get that.
Beyond the tremendous impact the DNC’s improper withholding of our data was having on our field program, we viewed it as a form of theft. We had paid the DNC and state parties an enormous amount of money for access. Each of the first four states in the process sells access to its voter file individually. The Iowa voter file costs $110,000. The other three in the first four ranged in price from $65,000 to $95,000. The voter file for each of the rest of the states is purchased from the DNC at a discount if you buy them all. The cost in 2016 for the rest of the states was $250,000.
The Clinton campaign had an advantage over us in being easily able to afford the $250,000 DNC tab. The $250,000 can be satisfied by writing the DNC a check or by raising $250,000 for the DNC from other sources. This allows well-heeled supporters of a candidate to write the DNC a big check and have it credited against the amount the candidate would otherwise have to pay for the voter file—even if the supporters have already maxed out to the candidate under federal election law.
We had informed the DNC that we didn’t have large donors who would pony up the cash for us to have access to the voter file. They offered to set up a joint fund-raising committee with us (which we did) and to organize DNC fund-raisers. Bernie would only have to appear, and whatever the DNC raised would be credited against his bill. The DNC never set up a single event. In the end, we wrote them a $250,000 check straight from the campaign. As we would later learn, these joint fund-raising agreements were used to funnel large sums of money for the benefit of the Clinton campaign—money that was falsely solicited for state parties.
It was bad enough that we had to pay for the voter file directly when our opponent did not, but now to have the file withheld—including the data that we had entered as a result of our volunteers’ effort—was unacceptable.
Our backs were against the wall. We could not afford to lose a whole weekend of organizing. The Clinton campaign and the DNC were pounding us in the media—a media unable to digest the technical aspects of the story and too willing to accept the Clinton narrative. This was one of the cases in which the Clinton folks really tested the line in terms of providing false information to the media—and the press bought it hook, line, and sinker.
If we had learned anything from the drubbing we took over my “vice president” comment, it was that we had to get off the defensive and take the fight to them. Bernie, Tad, and I consulted with our attorney Brad Deutsch. I advocated going to court to win back our access to the data. If our case was successful, not only would we have access to the data but we would have the implicit validation of a federal court that the DNC had acted inappropriately. This was not a strategy without huge risks. If the court ruled against us, the other side would have the same validation. So we decided that we would hold a press conference and lay out our case, giving the DNC one last opportunity to restore our service.
There was a lot of hand-wringing by the younger, less experienced staffers on the campaign. They were pushing for a conciliatory tone and a formal apology—from us. That would have been a disaster, because we would have been apologizing for the “alternative facts” put out by the DNC and the Clinton campaign and validated by the media. The next morning, Friday, our press shop put out a very short advisory that there would be a 2:00 p.m. press conference in front of our Washington, DC, office, a row house on Massachusetts Avenue. Bernie, Tad, Briggs, Brad Deutsch, and I met at Bernie’s small DC house. Bernie, Tad, and I had each prepared draft remarks, and we went to work creating a single document. The big question that remained was who would deliver them.
There was certainly a strong case for Bernie delivering them. He was unquestionably the person in the room with the most credibility. It would also ensure that the statement would be extensively covered by the media. On the other hand, it would put him personally in the middle of the maelstrom. We had avoided that up to this point.
We finally decided that I would deliver the remarks. If the whole thing went south, I would be the public face of the disaster.
We walked over to the office and went in through the back. The front lawn was packed with reporters and cameras. The press corps spilled over into the front of the next building. Many of the reporters expected the campaign to come out and apologize. That was definitely not our game plan, and we took them by surprise.
I walked out with Brad Deutsch and Michael Briggs. I immediately gained a new appreciation for what Bernie does on a regular basis—step in front of a wall of cameras that capture every word and every movement. There are no do-overs. I delivered the prepared statement. Then I took questions. After about half a dozen, I gave the press conference “thank you” and stepped back inside. It seemed to have gone well. Now we waited to see if the DNC would blink.
They didn’t, and neither did we.
That afternoon we filed for immediate relief in federal court. The judge brought the parties together on a 7:00 p.m. conference call and asked if this could be worked out. The problem was that the DNC would still not agree to our central demand—give us back access to our data. Hours went by. Finally, the judge said that she would hold an 11:00 p.m. hearing.
The DNC capitulated. If we would agree to cooperate with an investigation (which we already had), and if we would provide a statement about the chronology of events (which I had done the day before), they would turn the data back on.
The impact of this confrontation was considerable, and it provided a valuable lesson going forward. Reports came in from our offices in Iowa that during the press conference our staff and volunteers were cheering. (In our Iowa offices, the whole incident became known as VANghazi—a play on the Republicans’ trumped-up attacks on Hillary Clinton around Benghazi.) Similar reports came in from other states. Our willingness to challenge Debbie Wasserman Schultz had energized our people. Even though we did not make a direct appeal for money, over $2 million was donated in short order.
Attendance at our events over the following days swelled. A reporter would later relate that a key Clinton staffer told her that they expected the crisis to hurt our campaign, but that we had beaten them to a draw. If they were admitting, even privately, that we had beaten them to a draw, it meant that they felt we had won decisively.
Back in October, we had responded defensively to the Clinton campaign’s attacks over comments I had made that she should be vice president. No more. To win, one had to be on offense.
Despite this victory, Debbie Wasserman Schultz didn’t back down in her efforts to hurt Bernie’s campaign. We learned that the DNC had leaked activity logs related to the data issue to the media. We had asked for this type of information to help us in our own internal investigation. It was not until late Saturday, after a DNC staff meeting convened to decide whether they would even give us the information that the media already had, that they decided to give it to us. When we complained to the media that we had not been given the same access to information that was being distributed to the Clinton campaign and to the press, Debbie Wasserman Schultz relied on a tried-and-true tactic in politics. She told the media that we had in fact been given the information when everyone else had received it. In other words, she lied.
Debbie Wasserman Schultz’s conduct during VANghazi confirmed what so many people believed already—that the party establishment was putting its fingers on the scale for Hillary Clinton. For those who still had doubts, the leaked emails from the DNC and Clinton chair John Podesta and subsequent revelations by Donna Brazile made it crystal-clear months later.