America's Bitter Pill

by Steven Brill

  Through the summer of 2013, the press became captivated not by the struggle to build the website that all of those panicked internal emails and consultant reports reflected, but by what everyone thought was the real drama—the Obama 2012 campaign reunion story. The campaign whiz kids were back, and they were gearing up to use big data, new media, and all of their other tricks to win one more campaign for Obama, a campaign to get everyone to sign up for health insurance.

  Put simply, the political debate and punditry around the launch was mostly about whether the old Hope and Change team could get people to come to the website exchanges, not what would happen once they got there.

  Simas, who now had the title of White House deputy senior adviser for communications, made the rounds of major news outlets, providing colorful details of how, under his direction, big data was being fed by the campaign whiz kids at Civis to the field troops being organized by former campaign colleague Anne Filipic at Enroll America. They were going to target specific precincts, say, in Miami or Houston, to identify the uninsured and convince them to enroll.

  “We want multiple touches,” Simas told me when I interviewed him in his tiny White House basement office, where he sketched out those “touches” on a whiteboard: social messaging, door-to-door canvassing, community rallies, targeted radio ads. “The idea is to take the abstraction of Obamacare, and make it into an actionable reality.”

  On July 17, The Washington Post’s Ezra Klein and Sarah Kliff published a long feature on the progress of the rollout. Unlike most of their competitors, they devoted a lot of space to detailing the technical and regulatory challenges involved in launching a complicated program still bitterly opposed by Republicans and barely understood by the public.

  Nonetheless, their headline was “Obama’s Last Campaign: Inside the White House Plan to Sell Obamacare,” under which there was a large picture of Simas gazing out from in front of the White House.

  Klein and Kliff set up the coming drama this way:

  Deep inside the White House, in a bare room that the chief of staff uses for meetings, David Simas is still thinking about turnout.

  Turnout has been Simas’s job for years now. As director of public-opinion research and polling for President Obama’s reelection campaign, Simas was at the center of the effort to find and persuade young and minority voters to go to the polls like they did in 2008.…

  Now Simas, a sad-eyed Massachusetts native with a facility for PowerPoints, needs to reach those same groups again—with a much harder ask. This time, he doesn’t just need them to vote.

  He needs them to buy health insurance, and, in some cases, spend hundreds of dollars a month for it. If they don’t, the new insurance marketplaces—the absolute core of Obamacare—will be filled with older, sicker people, and premiums will skyrocket. And if that happens, the law will fail.…

  Simas is focusing his formidable analytical resources on understanding this group. He begins clicking through a PowerPoint that holds reams of data on these young adults.…

  A couple more clicks and Simas is showing which television channels they like to watch (Spike TV, among others), which social-media platforms they use (Twitter, Facebook) and who they listen to (“No surprise. It’s mom.”). “We can figure out the message that works best for this group,” Simas says.

  The focus on young, minority voters. The heavy reliance on microtargeting. The enthusiasm about nontraditional communications channels. The analytics-rich modeling. It sounds like the Obama campaign. And administration officials don’t shy away from the comparison.

  Even Jeanne Lambrew—who was supposed to be the White House official overseeing the website build—agreed that Simas’s work was what the launch was all about: “When I hear the conventional wisdom about Obamacare,” she told the Post, “this is the difference between the Karl Roves who put their fingers to the wind and the Nate Silvers of the world who looked at the numbers.”

  Implementing President Obama’s most important domestic policy seemed to be all about campaigning, not about governing.

  The officials at CMS and HHS whom I interviewed during the summer of 2013 were similarly focused. I was regaled with stories about the demographic targeting being done. In fact, the CMS staff claimed that its data, more than that supplied by Civis, was fueling the campaign.

  My questions about the technology were brushed aside so briskly and confidently that my initial impression that not having one person in charge spelled disaster was beginning to feel like an overreaction.*16

  When I talked to Simas, he, too, assured me that “everything has been tested and is working perfectly.… Our challenge is getting the right people to show up.”

  THE BLANK CALENDAR

  What got me suspicious again had to do with what became the much-publicized “official White House” Obamacare calendar.

  A July 23, 2013, report in The New York Times—another one focusing on Simas, his spreadsheets, and his “war rooms”—described it this way: “Royal blue and emblazoned with the White House seal, the ‘Official Affordable Care Act Enrollment Countdown’ is a paper calendar that keeps track of the time before uninsured Americans can begin signing up for coverage under the president’s signature health care law. On Tuesday, the top page said simply: ‘70 Days Left.’ ”

  Ah, I figured, when I read that story. That must be the shorthand version of their detailed battle plan. On the back of each numbered day must be the list of things that have to have been completed on such and such date for everything to come together. In management circles, the long version of this project playbook is sometimes called a Gantt chart.

  That had to be why, I assumed, the Times had noted that “when Denis R. McDonough, the White House chief of staff, first saw the document in a meeting in the [White House] Roosevelt Room, he immediately dispatched his aides to the West Wing basement to get copies for all of President Obama’s top advisers.” McDonough—who was telling reporters he was personally spending an hour or two a day on Obamacare—obviously wanted everyone focused on getting everything done that needed to be done, task by task.

  So, I asked the White House press office to send me a copy of the “Official Affordable Care Act Enrollment Countdown.” That way I could see their real launch plan and, by reviewing all the checked-off milestones, understand why they were all so confident that everything was going to be ready.

  Seven weeks later, what I got was twenty pages, each with a descending number. There was nothing else on either side of the pages. It was all blank.

  IN LATE JANUARY 2013, deputy chief of staff Nancy-Ann DeParle had gone back into the private sector to help lead a private equity firm. DeParle had told Obama in 2010 she wanted to do that, but she had stayed on when he promoted her to deputy chief of staff. Her decision to leave before the launch—when all the key decisions still had to be made and implemented—surprised many of her colleagues, and had left Lambrew as the key point person supervising CMS and other agencies building the website.

  Throughout the summer, Obama had regular meetings about the rollout with Lambrew, chief of staff McDonough, Valerie Jarrett, David Simas, and other members of his communications staff. Except for Lambrew routinely reporting that the build was on track, the meetings were about the political effort. Which celebrity had Jarrett or Simas lured into making pitches for people to sign up? Should the president make some calls of his own to recruit them? Which states looked especially promising? What presidential speeches or other events to boost enrollment were on the calendar? Could retailers such as Walgreens be recruited to help by distributing literature?

  According to McDonough, Obama always ended every meeting by saying something to the effect of, “This is all good, but remember it won’t matter if we don’t get the technology right.” The president had the right instinct, but was talking to the wrong people. No one in the room had any idea how or if the technology worked. Along with mock-ups of computer screens showing that everything was looking good, they had only gotten reports tha
t there might be some of the normal glitches affecting every start-up, but that as a general matter everything was a go.

  “You’ve got to have a culture in place where people aren’t afraid to admit there’s a problem,” was all Obama would volunteer when I asked him about those meetings and whether his simple admonition about the technology at the end of each session might not have been enough.

  “YOUR PROBLEMS ARE ALL HYPOTHETICAL”

  On July 29, 2013, a stock analyst writing on the Seeking Alpha website recommended insurance company stocks, which he said were undervalued given the likely positive impact of the launch of Obamacare. Indeed, the stocks of the major insurers had all jumped dramatically since President Obama had signed the law in March 2010, outpacing the overall post-recession Wall Street recovery.

  Karen Ignagni, who ran the insurance lobby, wasn’t feeling as good about Obamacare. Her companies had frontline exposure to the problems with the build of the e-commerce site, because it was their products that had to be linked into the website for the e-commerce to happen. She had rounds of meetings with the CMS people in Baltimore, telling them that every report she got was that the exchange was not going to be ready. “Why not let October be a month where people can browse the various offerings, but not buy?” she suggested. “Or perhaps you could just delay everything a month? What are your contingency plans,” she kept asking. There were no contingency plans, she was told. “Your problems are all hypothetical,” the CMS officials told her. “We’re not seeing anything like what you are describing.”

  STILL A “BILL,” NOT A LAW

  All of this behind-closed-doors angst was playing out through a summer in which the politics of Obamacare seemed to have become even more bitter than they had been during the Tea Party summer four years before. Despite the president’s reelection, and despite the Supreme Court’s ruling, the Republicans acted as if Obamacare were still a pending bill to be debated, rather than the law of the land to be implemented.

  The closest historical precedent might have been resistance to the Supreme Court’s Brown v. Board of Education decision desegregating public schools. Yet that at least could be attacked as the decision of nine men in robes, not the duly elected House, Senate, and president.

  When Senate minority leader Mitch McConnell got wind of an effort by the Obama administration to enlist the National Football League and other professional sports organizations to get their players to do public service commercials to encourage Americans to sign up for coverage (much the way Mitt Romney had enlisted the Boston Red Sox to promote Romneycare), he and Senator John Cornyn of Texas successfully urged the league to steer clear. “Given the divisiveness and persistent unpopularity of the health care law, it is difficult to understand why an organization like yours would risk damaging its inclusive and apolitical brand by lending its name to its promotion,” McConnell and Cornyn wrote to the NFL.

  At the same time, a group of conservative twenty-somethings, called Generation Opportunity, sprang up to urge the young invincibles to ignore the law’s mandate. They should not be lured into paying high premiums for insurance they don’t need, Generation Opportunity argued, just to support older people, who would pay lower premiums if the invincibles bought into the scheme and, thereby, lowered the overall risk profile of those in the insurance pool. Paying the penalty for ignoring the mandate was a better deal—and the right way to protest the new law and help torpedo it.

  Generation Opportunity, which was funded largely by conservative billionaires David and Charles Koch, soon started showing up at some of Anne Filipic’s Enroll America rallies and even staging rallies of their own on college campuses perceived to have more conservative students. The group had a creative—or, their critics would say, perverse—sense of humor that yielded a soon-to-be-viral series of Web videos. The most popular showed Uncle Sam crawling out from between the legs of a woman being examined in a doctor’s office. The message was classic Frank Luntz: Reject government interference in the doctor-patient relationship.

  THE “DEFUND” CAUCUS

  In August 2013, a cadre of the most conservative Republicans, led by freshman Texas senator Ted Cruz, began staging a series of “defund” rallies, aimed at encouraging conservatives to demand that Congress stop funding Obamacare. They had the leverage to do it, Cruz argued. Republicans could simply threaten to block any new budget, due October 1, or block any increase in the debt ceiling due soon thereafter, if Obamacare wasn’t defunded as part of a new budget and debt ceiling deal.

  “We have, I believe … the last good opportunity we will have to defund Obamacare,” Cruz declared at a July 30 meeting of the Heritage Foundation, whose political action committee was organizing the rallies. “If we do not pursue this strategy, we are saying we surrender. Obamacare will be a permanent feature of the American economy.”

  KENTUCKY GOES ALL IN

  But in one of the reddest of red states, Obamacare was picking up steam.

  Through the spring and summer, Kentucky Health and Family Services cabinet secretary Audrey Haynes, her kynect executive director, Carrie Banahan, and Deloitte’s Mohan Kumar had pushed their team on all fronts. Attractive shopping bags and literature had been designed with the “kynect” logo and the pastel colors and rising sun that the focus group researchers had suggested. A plan was put together for how the material would be distributed to the “kynectors” or “navigators,” that Sebelius had promised Senator Baucus were going to be hired, and to community groups that Haynes, a seasoned political operative, was helping to recruit. A calendar was created full of fall events where the marketing materials and troops would be deployed—such as the annual Bourbon Festival in Bardstown and the Apple Festival in Paintsville, near the poorest, reddest part of the state.

  In Kentucky, unlike Washington, the technology part of the work was being given equal if not more attention at the top. Banahan and Deloitte’s Kumar regularly briefed Secretary Haynes and occasionally Governor Beshear, not only about their progress but also about problems they were encountering along the way.

  Banahan and Kumar rode themselves and the contractors mercilessly, often working through the night and weekends so they would not miss any deadline on the elaborate milestone chart that was their bible.

  They had farmed out discrete portions of the work to multiple subcontractors. Even CGI, which was having so much trouble as the main federal contractor, worked on a piece of the Kentucky project.

  By April 2013, Kumar—who would take no vacation during the fifty-one weeks preceding the launch and worked through multiple weekends—had finished a first version of the site for Carrie Banahan to take on the road to test with focus groups and in community meetings. She had come back with design and flow changes that Kumar’s team incorporated into new versions.

  By June, the Deloitte people began “pressure testing” their site, giving it the kind of end-to-end run-throughs that the people at CMS would be worried still had not been done on the federal website in August. They found multiple problems and sometimes had to tell Haynes or even the governor that one or more of the issues might force a delay. But they fixed them, one by one.

  Now, on August 22, the feds from CMS arrived to see if Kentucky had its act together. In this case, the feds were from CCIIO, not the CMS Office of Information Services that was building the federal exchange. CCIIO had its own technology team, working completely apart from the CMS Office of Information Services, because it was CCIIO that was in charge of vetting and regulating the fifteen state-run exchanges.

  Chris Clark, the soft-spoken engineer Banahan had pulled out of retirement to supervise Deloitte’s engineers, was happy to have the feds kick his tires. He had thrown himself into the project, thrilled that even his six-year-old son understood that he was, as Clark later told me, “working on something important.”

  “Did you do good work for kynect today, Daddy?” his son would ask on the nights Clark got home before his son went to bed. “He’d seen the kynect commercials and had the
[shopping] bag,” Clark later told me. “He was excited about it.”

  Clark had already hosted another group of federal inspectors who had tested his site for security vulnerabilities and found everything in order. Now it was time for the big end-to-end test. Clark wasn’t worried about it, because he and the Deloitte team had done all that themselves first.

  The inspectors made Clark put his exchange through six hundred different consumer scenarios—a single person who qualified for Medicaid, a family who had a twenty-five-year-old who already had employer-paid insurance, or someone who wanted a gold plan but would receive a subsidy to help pay for it based on the price of the second lowest priced silver plan (because that was the way the law worked).

  The testing lasted three days.

  “We were the first state to say we were ready for testing,” Clark recalled. “And those CCIIO guys put us through the ringer. But it was all fine.”

  IN WASHINGTON, 38 PERCENT STILL UNDONE

  The CMS feds from CCIIO were pleased when Clark showed data demonstrating that Kentucky’s site could handle at least ten thousand users at the same time. Yet the CMS feds from Chao’s Office of Information Services were hoping their federal exchange site covering thirty-six states would be able to handle only fifty thousand users (though it was actually being built to handle only ten thousand, if that).

  Kentucky has about 4.3 million people. The federal exchange site’s thirty-six states have a population of more than 200 million. In other words, at headquarters CMS was planning to handle fifty times Kentucky’s population with, at best, five times the computing capacity its people had seen in Kentucky.

  The same day, August 22, 2013, that one arm of CMS was pressure testing a fully built version of the Kentucky exchange, the Henry Chao arm of CMS was meeting in Baltimore with their CGI contractor team going over fifteen pages of spreadsheets spelling out items for the federal exchange that still had not been completed. As The Washington Post would report three months later, for each item CGI had listed its level of confidence—high, medium, or low—that it could be completed by the October 1 deadline. Most were labeled high or medium, though some were rated low. But even many of those not rated medium or even high were deemed by CGI to be dependent on the performance of other contractors.