Atul Gawande compares controlling health care costs to controlling food costs at the start of the 20th century. The parallel is, by nature, inexact but this paragraph leapt out at me: 

Much like farming, medicine involves hundreds of thousands of local entities across the countryhospitals, clinics, pharmacies, home-health agencies, drug and device suppliers. They provide complex services for the thousands of diseases, conditions, and injuries that afflict us. They want to provide good care, but they also measure their success by the amount of revenue they take in, and, as each pursues its individual interests, the net result has been disastrous.

Our fee-for-service system, doling out separate payments for everything and everyone involved in a patient’s care, has all the wrong incentives: it rewards doing more over doing right, it increases paperwork and the duplication of efforts, and it discourages clinicians from working together for the best possible results. Knowledge diffuses too slowly. Our information systems are primitive. The malpractice system is wasteful and counterproductive. And the best way to fix all this iswell, plenty of people have plenty of ideas. It’s just that nobody knows for sure.

The history of American agriculture suggests that you can have transformation without a master plan, without knowing all the answers up front. Government has a crucial role to play herenot running the system but guiding it, by looking for the best strategies and practices and finding ways to get them adopted, county by county.

(Hat tip: 3QD)

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.