When thinking about user participation and the co-creation of value, it’s easy to focus on technology platforms that explicitly feature the creations of their users, like Wikipedia, YouTube, Twitter, Facebook, and blogs. Yet in many ways, the breakthroughs in Web 2.0 have often come from exploring a far wider range of possibilities for collaboration:
Open source technology platforms such as the TCP/IP protocol suite and utilities created as part of Berkeley Unix, as well as Linux, Apache, and MySQL, and open source programming languages such as Perl, Python, PHP, and Ruby, all built and maintained by collaborative communities, provided the fundamental building blocks of the Internet as we know it today.
The World Wide Web itself has an architecture of participation. Anyone can put up a website and can link to any other website without permission. Blogging platforms made it even easier for any individual to create a site. Later platforms like Facebook and Twitter are also enablers of this kind of explicit participation.
First-generation web giants like Yahoo! got their start by building catalogs of the content assembled by the participatory multitudes of the Net, catalogs that later grew into search engines. eBay aggregated millions of buyers and sellers into a global garage sale. Craigslist replaced newspaper classified advertising by turning it all into a self-service business, right down to the policing of inappropriate content, having users flag postings that they find offensive. Even Amazon.com, nominally an online retailer, gained competitive advantage by harnessing customers to provide reviews and ratings, as well as using their purchase patterns to make automated recommendations.
Google’s search engine dominance began with two brilliant insights into user participation. First, the PageRank algorithm that Larry Page and Sergey Brin created while still at Stanford was based on the realization that every link on the World Wide Web was a kind of vote on the value of the site being pointed to by that link. That is, every time any of us makes a link to another site on the Web, we’re contributing to Google. Second, Google realized that it could provide better advertising results not by selling advertisements to the highest bidder, but by measuring and predicting user click-through rates on ads. A $10 ad that is twice as likely to be clicked on is worth more than a $15 ad. Google could only deliver these results by understanding that every click on a Google search result is a kind of user contribution. Since then, Google has gone on to mine user participation in many other aspects of its core business as well as in new businesses, including speech recognition, location-based services, automated translation, and much more. Google is a master at extracting value from implicit participation. It makes use of data that its users provide simply in going about their lives on the Internet to provide them with results that quite literally could not exist without them.
Just as Google has become the bellwether company of the Internet era, it is actually systems for harnessing implicit participation that offer some of the greatest opportunities for Government 2.0.
There are great examples to be found in health care. As costs soar, we discover that costs and outcomes aren’t correlated. Atul Gawande’s New Yorker article[38] on this disconnect—outlining how McAllen, Texas, the city with the highest health care costs in the U.S., also had the worst health outcomes—led to what Health and Human Services CTO Todd Park referred to in a conversation with me as a “holy cow moment.” Todd is now working on what he calls a “holy cow machine,” a set of services that will allow every city to understand how its health care costs and outcomes compare to those of other cities.
We have all the data we need—generated by the interactions of our citizens with our health care system—to understand how to better align costs and outcomes. Taking this idea to its full potential, we need to get beyond transparency and, as Google did with AdWords, start building data-driven feedback loops right into the system. Google’s tools for estimating the effectiveness of keyword advertising are available to advertisers, but that’s wonky, back-office stuff. The real magic is that Google uses all its data expertise to directly benefit its users by automatically providing better search results and more relevant advertisements. The most amazing thing about Google is how dynamically the prices for its advertising are set. Every single Google search has its own automated ad auction. The price is set dynamically, matching supply and demand, seven or eight billion times a day. Only financial markets operate at this kind of speed and scale.
A Gov 2.0 analogue would not just be a “holy cow machine” for transparency; it might, for example, be a new, dynamic pricing system for Medicare. Currently, an outside advisory board makes recommendations to Congress on appropriate Medicare reimbursement rates. As David Leonhardt noted in the New York Times, “Congress generally ignores them, in deference to the various industry groups that oppose any cuts to their payments.”[39] Leonhardt’s solution: an independent body, akin to the Federal Reserve, empowered to set reimbursement rates in the same way the Fed sets interest rates.
But shouldn’t such a body go even further than periodic resets? Technology would allow us to actually manage reimbursements in much the same way as Google dynamically adjusts its algorithms to produce optimal search results and optimal ad placements. Google takes into account hundreds of factors; so too could a Medicare rate-setting algorithm. To take two examples from Leonhardt’s article:
Each year, about 100,000 people die from preventable infections they contract in a hospital. When 108 hospitals in Michigan instituted a simple process to prevent some of these infections, it nearly eliminated them. If Medicare reduced payments for the treatment of such infections, it would give hospitals a huge financial incentive to prevent them….
There are a handful of possible treatments for early-stage prostate cancer, and the fastest-growing are the most expensive. But no one knows which ones work best.
By measuring outcomes and linking reimbursements to those outcomes—rather than the current “fee for service” model, which encourages unnecessary procedures—Medicare could pave the way to a real revolution in health care.
Because of the political difficulty of such an intervention, it’s unlikely that Medicare would be allowed to unilaterally introduce such an algorithmic payment system. As a result, I do suspect that this kind of innovation will come first from the private sector, which will trounce its competition in the same way that Google trounced its competitors in the search advertising market. As a platform provider, though, it’s possible to see how government investment in the data infrastructure to measure and report on outcomes could jump-start and encourage private sector investment.
Real-time linkage of health costs and outcomes data will lead to wholesale changes in medical practice when an innovative health care provider uses them to improve its effectiveness and lower its costs. Such a breakthrough would sooner or later be copied by less effective providers. So rather than attempting to enforce better practices through detailed regulations, a Government 2.0 approach would use open government data to enable innovative private sector participants to improve their products and services. And to the extent that the government itself is a health care provider (as with the Veterans Administration) or medical insurer (as with Medicare), it can best move the ball forward by demonstrating in its own operations that it has been able to harness technology to get the job done better and more cost-effectively.