Innovative State

by Aneesh Chopra

  “A partnership,” she said.

  We had a brisk breeze at our back. That July, the California Public Utilities Commission, the state’s utility regulator, had ordered that the utilities at least agree on a process for the design of a standard record format, making this stage merely about solidification and deployment. And, since the three utilities represented different parts of the state, they weren’t competing with each other for customers but, rather, free to compete together, against the clock, toward the achievement of similar goals.

  “The key was just to keep it simple,” Austin said. “A lot of the standards had already been thought through but never deployed.”

  To Austin, the idea of not getting something done, even in the compressed time frame, “didn’t cross my mind. The thought of creating separate formats, I don’t know, it wouldn’t have made sense to proceed that way. It was going to happen. I was comfortable that we were going to get there.”

  Before they left the room, they agreed upon a sketch of the key fields that a customer would need, as well as a standard record format for the data that would fill those fields. Then they assigned someone from every utility to deliver a common user interface. This was a relatively easy technical exercise, since the group adopted an existing standard that simplified the “what” and “how” data would be published and transmitted between parties.25

  Within 90 days, I visited California once again, to celebrate a number of Green Button commitments.26 Austin’s PG&E and her counterpart at San Diego Gas & Electric announced they were live for all of their customers; the technical body responsible for the data format, North American Energy Standards Board, announced a free starter toolkit that included detailed technical documents necessary for developers to build products and services on top of Green Button data. A few startups demonstrated early prototypes of apps that personalize energy savings tips through the use of a customer’s Green Button file.27 All of this was inspiring, but incomplete. Following the presentation, an audience member asked if the utilities would be publishing their rate schedules electronically as well, so customers could make energy decisions based on specific costs at a particular moment; in California, peak time pricing is double or triple the normal rate. Austin, who recognized the value of that data, related the reality: improving access to that information might take more than a year.

  In the interim, however, an entrepreneur named Jason Riley was working to prove that there were better ways of linking energy pricing and usage data. Prior to the Green Button launch, back in 2010, Riley had founded a startup called Genability on the premise that, to deliver meaningful savings, an energy consumer would need both pricing and usage data. The pricing side hadn’t been a problem; with the help of contractors, he had begun manually entering utility tariff information for hundreds of rate plans into an online database, offering access to developers for a fee. On usage data, however, Riley had been stuck, since that information had been inaccessible to the public. At least it was, prior to Green Button.

  I met Riley a week after the Green Button launch in California, while I was serving as a guest judge at a Cleanweb Hackathon in New York, where industry and government officials outlined specific challenges related to energy, such as the exorbitant cost of solar panel installation, and—in a bit of foreshadowing—New York City’s efforts to help residents better prepare for storm surges through the visualization of climate change. The organizer’s primary purpose, however, was to invite volunteer developers to spend 72 hours building prototype applications on top of a growing number of energy information services, while promising modest prizes.

  With access to Genability’s Electricity Pricing database, Green Button data for energy usage, and other open energy data sources, 15 developer teams completed prototypes by Sunday evening. One stood out, because, in using both Genability and Green Button data, it educated users about the most economical rate plans for them. The Watt Quiz was a simple, engaging customer survey that, after the customer uploads his or her Green Button file from the utility, helps answer a simple question, “Which tariff rate plan saves you the most money?” It showed how one family could save 44 percent simply by changing the rate plan.

  The CleanWeb Hackathon scaled into an international movement, not just because developers like to tinker and gather, but because of the promise of a new business model—making money while helping energy consumers save it. Take, for instance, Simple Energy, based out of Boulder, Colorado, which partnered with San Diego Gas & Electric in its launch of the Green Button service. After using Simple Energy’s Customer Engagement platform to see her family’s energy usage online, Heidi Bates deputized her six-year-old son Thaddeus as the “Light Police,” to run around the house unplugging unnecessary luminescence. “He really digs it,” she said. The enthusiasm spread throughout age ranges; a grandmother, Josephine Gonzales, saved over 20 percent on her electric bills using the Facebook-­connected platform.

  Meanwhile, in Northern California, in the wake of the Green Button rollout, Austin continued coming across customers who benefited from the access to information, and were eager to share experiences. That included one couple—she refers to them as Kelly and Jim—from San Luis Obispo that had downsized to a residence half the size, expecting the energy bill to decline in kind. “When it didn’t budge at all, Kelly got onto the Green Button, and took that data and then really went around her home and looked at her usage,” Austin said. “She was able to cut her bill from $160 per month to about $50 per month, which was an annual savings of $1,300.”

  Some customers even revealed that they had taken it upon themselves to spread the Green Button word around their neighborhoods, to improve the area’s overall efficiency. Understanding the power not only of the data but also the importance of presenting it in easily comprehended charts and graphs, Austin asked the board at Pacific Gas & Electric for resources to sponsor an Apps for Energy contest in partnership with the White House.28 Nearly 50 developers applied in the spring of 2012, some small companies, some single individuals. They produced a diverse range of concepts, with first-place honors going to Leafully—a visually appealing software tool accessed through social media that breaks a person’s energy footprint and environmental impact down into a corresponding number of trees saved. The second prize went to Melon.com, which combined Green Button and the EPA’s Energy Star portfolio manager to give more than one million commercial buildings a “simple and affordable” benchmarking analysis, allowing managers to compare their buildings with others around the United States, comply with the law, and save money. The third-prize winner, VELObill, was a colorful, intuitive application for consumers to view utility usage, compare it to peers, find ways to save, and find local contractors who can get them closer to their energy goals.

  Still, Austin was anxious to iterate further in collaboration with the White House, simplifying access for third parties and, thus, speeding the rate of innovation. My successor as Chief Technology Officer, Todd Park, would announce the release of Green Button Connect My Data, a program that would eliminate the need for customers to take possession of their own data via download in order to use third party applications. Through the Connect My Data service, that upload would happen automatically once consumers enrolled.

  Austin helped winners of the Apps for Energy contest go into production in her market. One was the previously mentioned Leafully. Another was PEV4me, which, by accounting for a user’s driving habits, calculates how much the user would save on gas by switching to an electric car. Another was UnPlugStuff, which assists in determining the cost of phantom usage, such as leaving your toaster plugged into the wall. In the first three months, after setting up all three with sufficient security protections, roughly 10,000 people signed up to use their services, and that was even with limited advertising during the election season.

  “So far, so good,” Austin said.

  Austin viewed this early progress to be a product of
“keeping it simple versus trying to boil the whole ocean, and getting everyone on the same page.”

  In this scenario, “keeping it simple” meant implementing the Green Button standard rather than attempting to account for every possible request for how to access someone’s individual energy data. It also meant moving incrementally, first providing the data in downloadable form, then automating that connection.

  And “everyone” meant representatives of the government and the private sector, not at odds with each other but in collaboration, sharing ideas and shaking hands, before handing off to entrepreneurs to innovate.

  Utility executives haven’t always embraced government involvement in their affairs—after all, as Austin noted, government regulation can sometimes slow the pace of innovation.

  “In this case,” Austin said, “government was a positive enabler for sure.”

  In 2013, President Obama would use the State of the Union to reiterate his commitment to the smart grid, by calling for a Race to the Top proposal which would provide financial incentives to states who adopt energy efficiency policies. As of press time, that had yet to pass, but players in this space had moved ahead anyway, with 35 utilities and energy providers voluntarily committing within the first year to provide 36 million homes and businesses with their own energy usage information in the consensus, industry-standard Green Button format.

  For all the inroads we were making in the energy sector, we knew there was a need to pour our energy, in terms of standards creation and deployment, into other industries. In the Obama administration’s calculus, no national priority ranked higher than health care. Thirteen months prior to the signing of the landmark Affordable Care Act into law, its technological underpinnings had been codified in the Health Information Technology for Economic and Clinical Health Act (HITECH Act), a provision of the Recovery Act that we mentioned earlier. Only weeks into his first term, the President pushed for a $26 billion incentive program that would encourage doctors and hospitals to adopt and use electronic health records. And on this rare issue, he didn’t encounter much pushback from either party, since many members on both sides of the aisle—from Hillary Clinton to Newt Gingrich—agreed that, no matter how we financed the care delivery system, it needed to be modernized through the application of information technology, to improve quality and reduce costs. Congress authorized the Centers for Medicare and Medicaid Services (CMS) to pay doctors up to $44,000 and hospitals millions based on their size—but only if they “meaningfully used” certified technology in their daily work treating patients.

  Why the broad consensus for intervention? Because, as the rest of the economy was surging forward, experiencing sizable productivity gains powered by the workplace application of information technology, health care was stuck on a treadmill. It was underinvested in technology, relative to comparable service industries. Even when the industry did invest or innovate, its priorities tended to be misdirected, focused on the wrong set of problems. That was partly a product of the predominance of the fee-for-­service structure in the $2 trillion health care system, a structure that financially rewarded quantity over quality. The more patients a physician sees, the more tests that physician orders, the more the physician can bill. There is little incentive in the system for investing and engaging preventive and chronic disease management, care coordination, medication management, and telemedicine services (e-mail, online chats, and videoconferencing). In fact, there is actually a perverse disincentive, as healthier patients need fewer services, thus reducing provider compensation. It should come as no surprise that health care professionals have sought IT-enabled products that improved efficiency in billing and scheduling for services that will be reimbursed, but largely ignored those that might improve individual care, let alone the health of the greater population.

  Consider the problem of premature births, which we described in a Virginia context in Chapter 3. The health IT industry has the capability to better predict if a mother is likely to deliver early, which often leads to admissions into neonatal intensive care units. But, when those admissions generate revenues of more than $40,000 per week, what is the incentive for a health care system to invest in that technology? Altruism alone? And, worse, even if you identified those vulnerable mothers, you would bear the costs of any preventative services that you offered, since most come with no reimbursement.

  Even while pushing the health IT movement forward, proponents understood that the payoff would be limited until the payment systems aligned incentives. This was even noted early in a 2005 report by the RAND Corporation, the nonprofit global policy think tank, which estimated the potential of more than $81 billion in annual savings through the widespread adoption of electronic medical rec­ords systems, with that figure potentially doubling through additional health IT-enabled prevention and management of chronic disease.29 The report declared that the full benefits of health IT were “unlikely to be realized without related changes to the health care system.”

  Lawmakers of all stripes referenced the sanguine predictions of the RAND study in advocating for incentives to encourage provider adoption of such systems, and that advocacy played a role in the passage of the aforementioned HITECH Act. They did not, however, provide the accompanying payment reform, at least in the short term. And when RAND reassessed its study in 2013, its findings prematurely rang alarm bells for many.30 It found that the $81 billion savings had not materialized, and health costs had actually risen, ironically, in small part due to the improved billing and documentation procedures that the health IT systems had made possible. So, rather than a refuge for consensus, health IT became more of a partisan talking point for those opposed to the President’s overall health care agenda and unwilling to consider the contributing factor of fee for service.

  There remained a need for a more virtuous cycle in health care, one outlined by the Center for American Progress.31 That think tank saw potential for health IT adoption, care delivery innovation, and provider payment reform to interact and flourish, with each helping to make the others work, instead of the absence of one or more of those elements stalling overall improvement. In its reassessment, RAND didn’t give up hope: “We believe that the original promise of health IT can be met if the systems are redesigned to address these flaws by creating more-standardized systems that are easier to use, are truly interoperable, and afford patients more access to and control over their health data. Providers must do their part by re-engineering care processes to take full advantage of efficiencies offered by health IT, in the context of redesigned payment models that favor value over volume.”

  Progress in those areas was already under way.32

  On payment models, the Affordable Care Act had included many payment reform provisions, chief among them the creation of a new Center for Medicare & Medicaid Innovation, seeded with $10 billion to run experiments on new payment models that have the potential to improve quality and lower costs. To complement the center, Congress granted a new regulatory authority. If the CMS’ nonpartisan actuary certified that a payment model achieved quality improvement and cost reduction, then the HHS Secretary could make it an option for every provider in the country.

  On interoperability, the HITECH Act had tied its $26 billion in incentives for health care providers to their adoption of technologies that incorporated standards, which would be defined in three stages over the next five years. And for the standards work, we had called upon an experienced hand—my old friend Dr. John Halamka, who had engaged in the standards process across two Presidential administrations.

  Under President Bush, who had pledged in 2004 that every American would have access to a personal health record within a decade, HHS would seed pilot investments for technical work to that end. As Chairman of the Healthcare Information Technology Standards Panel (HITSP), Halamka was exposed to well-intentioned actors, but also to a process primarily driven by a few senior government people and the vendor community, rathe
r than by the doctors, insurance companies, and patients. It was also a process that lacked any economic incentives for the parties to push themselves toward the best possible performance. In the current market environment, vendors and their hospital customers simply didn’t have much of a stake in making it easier to share information. The reimbursement system encourages health organizations to grow through increasing market share, an objective that would be undermined if it was easier, not harder, for patients to seek care elsewhere.

  The result of such a vendor-driven approach, according to Halamka, was “basically codifying the status quo.” The standards that they did create “were so cumbersome and so heavy that vendors could only successfully implement them by charging vast sums of money.”

  In his new role, as vice chair of the Health IT Standards Committee (HITSC) in the Obama administration, Halamka had a more formal legal foundation than during the Bush administration. Authorized by the HITECH Act, the HITSC would benefit from greater engagement and urgency among all parties, a clearer business case to scale what works, and a more constructive governance model. It was a collaborative model that owed plenty to the insights that Mitch Kapor, the Lotus Development Corporation founder, outlined and endorsed in his aforementioned speech calling for a Health Internet. Kapor called for a less complex, more open, “light federal approach” that would encourage an early critical mass of users to participate—something like a dozen or more products and services built on the standards—thus reducing costs and time to market while fueling innovation. That’s how the standards committee set about its work.

  “Let’s do it in an open, transparent, multistakeholder fashion that will be bottom-up rather than top-down and will be fueled by innovation, agility and low costs,” Halamka said. “We’ll make it a ‘do-ocracy.’ That is, you will be rewarded for actually achieving results. What ended up happening is that suddenly the implementation guides, instead of thousands of pages of complicated technology specs, became ten pages of simple technology specs. You saw existing Internet standards being leveraged for health care. You saw open source. You saw intellectual property freedom. And you suddenly actually got the vendors a little bit on the run, because now they were having to open their systems and enable platforms and they could no longer charge obscene amounts of money for simple tasks.”