David Brooks has a piece today in the NYT today asking whether markets work for health care. He looks at data about what economists have found about the impact of competition in health care services. I won’t weigh in on the final answer in the health economics literature–which is pretty vast. (I don’t think it’s right to say “proponents of market-based health care rely less on theory and more on data”–there are plenty of proponents of market-based health care who know nothing more about why they like the idea beyond theory. I also don’t think it’s right to say most progressives say markets don’t work–citing to Arrow’s seminal 1963 article on the economics of health care. Arrow–who was my PhD advisor–is deeply progressive but certainly thinks markets work–the story here is about how much, when, and with what kind of institutional structure markets can do better than public provision. ) But I do want to point out that we’re not even having this conversation about how well markets work for law. I would but I can’t weigh in on what the data show about the value of markets in law because we nobody is studying that systematically.
My pitch in Rules for a Flat World is about the need for a more vibrant, bottom-up, and market-based approach to producing legal infrastructure–ranging from legal services to legal/regulatory rules and procedures themselves. My argument is based mostly on theory: we can’t innovate the basic platform of rules we need to manage complex human interactions and drive progress in both economic and social terms if we rely, as we currently do, almost exclusively on top-down public provision and highly closed legal markets, and almost exclusively on complex text-based rules that require expensive interpretation and generate expensive and slow adjudication to resolve disputes. That production technology is simply not up to the task of responding to the speed, complexity and intangibility of a digitized, globalized economy. The reason I pitch for finding ways to get more of our rules and legal processes developed by non-government entities–both for-profit companies and non-profit community organizations–is to harness what markets offer: a distributed and decentralized mechanism for processing massive amounts of information and identifying when, where, and how doing things differently might work to produce value. Market incentives–which I understand broadly to mean not only profit motives but also the incentive to achieve private non-profit goals–pull resources into the problem of innovating better ways of doing things. We need more (properly regulated) markets involved in producing legal infrastructure because markets are a better technology for directing resources to complex problems.
I do my best in the book to provide empirical evidence to back up these theory-based claims. And the “show me the evidence” response is one I’ve heard a lot in the decade I’ve been working on these issues and to which I’m acutely sensitive. I would love to be able to review that literature. But the the data either don’t exist or haven’t been assembled and shared for analysis by researchers, the researchers aren’t being trained, there’s next to no funding available (as I mention in the book, the NSF budget for law is about $6 million–one-tenth of one percent of total NSF funds; medical research is a multi-billion dollar endeavor), and the studies aren’t being done. There is no “public health” or “epidemiology” equivalent in law. There are next to no economists analyzing the health of our legal markets.
The irony is that much of the debate in health care economics is about the relative impact of different rules governing the markets in health care–it’s about the legal infrastructure of health care. Although Brooks’ piece ostensibly compares a market-based to a non-market-based approach, in fact even single-payer systems are heavily market-based.Both pre- and post-Obamacare, there are lots of markets involved governed by lots of legal rules and regulations. Where’s the data on how well markets for producing and implementing those rules would work?