America’s current financial system is essentially broken for the many working Americans who live on the edge financially. Mismatched incentives are at the heart of the problem. One has only to look at the sad history of payday and car title loans, subprime mortgages, or bank overdraft “protection” schemes for examples of financial products and practices that are profitable for providers but damaging for vulnerable populations.

Unfortunately, regulation hasn’t done much to create win-win solutions. Deference to efficient market theory and “consumer choice” has created a regulatory system that largely places responsibility — absent the most egregious abuse — on the individual consumer. In a new report, we argue that what is needed is a different regulatory approach that ties financial-services providers’ profit to customers’ financial health. In short, banks should only do well financially when their customers do well financially.

Such a regime would be similar to experiments currently happening in health care that pay providers for improving patients health, rather than paying them simply for treating patients regardless of the outcome of the medical intervention. Of course, the markets for health care delivery and consumer finance differ considerably — in competitiveness, concentration, public and professional ethical, and regulatory regimes and incentive structures. But while the U.S. health care system is not perfect, there is a general consensus that quality of care and patient outcomes among the insured population have benefited significantly from investment in, and deployment of, patient data and associated analytics to measure and improve patient outcomes.

We desperately need similar innovations in the financial services industry. The good news is that many of the pieces are in place for reform. Recent decades have witnessed an enormous expansion in both the quantity and quality of consumer financial data and data science techniques that can be used to improve credit analysis, customer authentication, risk management, and marketing. At the same time, more sophisticated measures of consumer “financial health” have been developed. The best known of these have been created by Consumer Financial Protection Bureau and The Financial Health Network, and measure things like whether individuals spend less than they earn, pay their bills on time, plan ahead for expenses, and have sufficient liquid and long-term savings, a sustainable debt load, access to affordable credit, and appropriate insurance. Applying metrics like these to the trove of financial data held by providers using data science techniques makes it possible to watch an individual’s overall financial well-being evolve over time, as well as discern how their financial health may be affected by their use of specific financial products and providers.

Our regulatory proposal has three stages and would be implemented over time. The first stage would require large consumer financial services providers to periodically make available to regulators internal data that regulators can use to analyze and measure changes in customer financial health. At the same time, regulators and the industry will collaborate on the testing, refinement, and standardization of a set of consumer financial health measurements which can be used at both the product and provider level. These outcomes measures can be normalized to avoid favoring providers who serve affluent customers, to track differences in outcomes by income, age, sex, education, race, geography, etc., and to control for the impact of recessions and other macroeconomic factors. Once the initial set of consumer financial health metrics are ready for use, the regulatory data sciences team will use the compiled provider data to analyze and measure correlations between financial product usage, product characteristics, individual providers and provider practices, on the one hand, and outcomes, on the other hand.

Read the rest of Todd Baker and Corey Stone’s article at Harvard Business Review