Look at science today and you’ll find a venture full of holes—experiments that can’t be duplicated, results based on the flimsiest of evidence. (Andrew Ferguson discussed many of the problems with science today in this cover story from the Weekly Standard.)

I haven’t heard of a foundation whose goal was to make science sounder, until I read this article from Wired about the activities of the Laura and John Arnold Foundation in the field.

The piece, by Sam Apple, who teaches science writing at the University of Pennsylvania, is very fair and quite a surprising piece for Wired to run, given that this magazine has tilted sharply to the left under the editorship of Scott Dadich. From what I see, the Arnold Foundation is either conservative or one that likes to fund causes conservatives like to support.

Until now, the cause the Arnold Foundation was most involved in was reforming abuses in government pensions, with particular emphasis on pensions in their home state of Texas. Philanthropy wrote about the foundation’s efforts in this area in 2014.

Among the recipients of Arnold Foundation pension reform grants are Brookings and the Pew Charitable Trusts.[1] The Wall Street Journal reported that as a consequence of this,  union-backed groups yelled at Brookings and Pew about accepting Arnold Foundation money.

This leads me to suspect that the Arnold Foundation does what they do without loudly announcing their politics. According to Stein, the Arnolds say they are Democrats and in 2013 the foundation donated $10 million to keep Head Start running during a government shutdown. But perhaps they aren’t predictable—which makes them interesting.

John Arnold made his money as an energy trader. His rise began at Enron, where he was so good at trading natural gas contracts that he was given an $8 million bonus to stay on shortly before Enron went bankrupt in 2001. Arnold, who had nothing to do with the criminal part of Enron, then started Centaurus Energy, which traded futures contracts on natural gas. By 2007 he was a billionaire and by 2012 he could retire at age 38 and devote himself to philanthropy.

Enron may have been a “mine are bigger than yours” corporate culture, but Arnold was the quiet guy who was excellent at his job when the swaggering guys were loudly plunging Enron into bankruptcy. He apparently doesn’t talk much today, and Laura Arnold gives most of the interviews. Stein shows that Arnold does two things more donors should do: he reads widely and lets his reading inspire his giving.

University of Virginia psychologist Brian Nosek was convinced that many psychological experiments could not be duplicated, including many that came up with controversial conclusions. He created the Reproducibility Project to see if this was true, and got a network of volunteers to begin studying experiments. But for two years his efforts at getting grants led to closed doors.

Then in 2012 Nosek got an email from the Arnold Foundation, in response to an article in the Chronicle of Higher Education about Nosek’s work. The foundation has since given Nosek’s organization, the Center for Open Science $15 million to encourage psychologists to make their experiments as open and transparent as possible so that they can be easily duplicated.

A second interest of the Arnold Foundation is in nutrition research. In 2011 John Arnold heard this episode of the excellent podcast EconTalk, in which author Gary Taubes argued that there is little evidence that eating saturated fat makes people fat. He decided to give multi-million dollar grants to the Nutrition Science Initiative, founded by Taubes and a team of medical doctors who are determined to discover the true causes of obesity. The Arnold Foundation has also supported the work of Nina Teicholz, whose book The Big Fat Surprise also argues that saturated fat is not the primary cause of obesity.

Finally, the Arnold Foundation supports the work of British journalist Dr. Ben Goldacre who came to fame as the “bad science” columnist for The Guardian. Dr. Goldacre is using his Arnold Foundation grants to “build an open, searchable database that will link all publicly available information on every clinical trial in the world.” Dr. Goldacre believes that drug companies foist worthless drugs on the public via clinical trials that aren’t publicly available, and throwing sunshine on the process of drug development will make consumers better informed about the drugs they need (or don’t need).

Why this emphasis on science? Remember, the Arnolds are in their early forties. They had originally thought about giving based on traditional, double blind methods of evaluation. But their extensive reading convinced them that these methods might not be effective because the foundations of science were in themselves flawed. So they thought fixing science might be a worthier goal for their philanthropy.

Stein interviewed Stuart Buck, the Arnold Foundation’s vice-president for research integrity. “In everything they do,” Buck said, the Arnolds “want to be evidence driven.” But if the nature of scientific evidence is called into question, Buck said, “you start to think: what is evidence? What do we actually know?” Answering these questions is the subject of the Arnold Foundation’s research program.

One of the many problems foundations have is groupthink, choosing projects that are just like everyone else’s. But I don’t know of any other foundation with the mission of the Arnold Foundation.

The Arnold Foundation deserves credit for taking a path no other foundation is taking. We should pay careful attention to what they are doing.

[1] Remember, Pew is a nonprofit and not a foundation so they can accept grants.