How I Approach Social Change, 2/27

Those who can't, start think tanks

Noah Smith interviews the founders of a progress-oriented think tank. Caleb Watney says,

you need just the right amount of attention in the right places. You want staffers and elected officials to understand that your policy idea is 1) within the Overton Window of their political party 2) worth prioritizing over other issues and 3) aligned with their preexisting goals. This process often plays out on Twitter, where influential writers and thinkers can endorse policy ideas and signal to their in-group that they are worth pursuing. A lot of Hill and agency staffers lurk on Twitter, and they see all your good (and bad) tweets. If you can sell your idea there (and again in-person over drinks) but keep it off Tucker Carlson or Rachel Maddow, then you’ve hit a sweet spot of salience.

I have been an “affiliated scholar,” meaning that I have neither a salary nor an office, with two think tanks—the Mercatus Center and the Cato Institute. It is tempting to think that working at a think tank is a way to encourage possible social change. After all, government is a major factor in society, and think tanks try to influence government. But I am very cynical about the value of think tanks.

Perhaps the best way to explain where I am coming from is to offer the social-change aspects of my autobiography. I have had a lot of creative ideas that I thought at the time were important. This autobiography will describe the goal, the idea, and the result. The TL;DR is that the working in business is the activity most likely to achieve positive social change.

In 1980, I completed my Ph.D dissertation. My goal was to solve a theoretical problem in Keynesian economics, hoping to steer the profession away from the “rational expectations neoclassical” macroeconomics that was all the rage within top economics departments. The idea was to explain price stickiness as based on information problems.

Trying to solve an important theoretical problem is a terrible strategy for a dissertation. Instead, you should figure out what the departments that are hiring are looking for. What they were looking for at the time were dissertations that were based on the rational expectations approach.

For me, the result was particularly bad. A professor who interviewed me on the job market stole my idea and published it before I did, without acknowledgment. The idea had no impact on the profession. In fact, it has been periodically rediscovered since (with no credit either to me or to the man who stole it), but again with no impact. This experience has left me less than excited about the academic research process in economics as a way to generate social change through ideas.

My first job out of grad school was at the Fed. I do not recall coming up with any significant ideas when I was there.

In 1986, I started working at Freddie Mac, the mortgage giant. At the time, I thought of it as a profit-seeking enterprise, albeit with some peculiar features. One feature was that it was supposed to “serve” upward mobility in the housing market.

Another feature was that with its government guarantee, its debt costs were low, and this gave it an advantage in undertaking certain forms of financial arbitrage. I was appalled when an economist told me excitedly about an anomaly in the Eurodollar market that he thought that Freddie could and should exploit. To me, that seemed like an abuse of Freddie’s Congressional charter.

In the 2000s, long after I had left, Freddie and Fannie took on riskier borrowers in a (misguided) attempt to serve upward mobility in the housing market. They also engaged in more of what I thought of as abuses of their low-cost debt status. I ended up happy to see them shut down during the financial crisis of 2008. I think that profit-seeking and a tight relationship with the government were ultimately a bad combination.

But working at Freddie gave me the best opportunity I have ever had to produce social change. My most significant idea there was to change the underwriting process to reduce judgment and rely more on data. Instead of trying to use AI to imitate human underwriters, I pushed for using credit scores. I also promoted using the Case-Shiller method for estimating home prices in an attempt to reduce the reliance on appraisals. The goal was to reduce the cost of obtaining a mortgage loan, to reject fewer good loans, and to accept fewer bad loans.

Freddie Mac adopted the credit scoring approach in 1994. For me personally, this was more bitter than sweet. Just as the idea I was pushing for was adopted, I was treated to a humiliating demotion, and I soon left the company.

As far as social change is concerned, the move toward credit scoring dramatically changed the mortgage industry. Yes, underwriting costs fell, and decisions became more accurate, with fewer good borrowers turned down and fewer bad borrowers accepted. But it also allowed new players to enter the mortgage lending market. Some of these players developed the so-called subprime mortgage market, with mortgage securities provided by Wall Street. These new players were central actors in the financial crisis of 2008.

I am not saying that I personally brought down mortgage lending costs, or that I personally caused the financial crisis. My guess is that the move toward credit scoring was going to happen at some point, anyway, and so that my efforts accelerated the process by at most a few years. And the financial crisis had many causal elements, mostly involving the political economy of mortgage lending in the U.S. I still think that introducing credit scoring into mortgage lending was socially beneficial, at least directly. But in a complicated world, the indirect effects of actions are difficult to assess.

When I left Freddie Mac in April of 1994, I created The Homebuyer’s Fair, one of the first commercial sites on the World Wide Web. The goal was to use the Internet to disintermediate in real estate and mortgage lending. For me personally, it worked out well. But apart from any role that the site played in stimulating interest in the Web (we got tons of press in 1994-1997), I would not say that it came anywhere close to achieving any major social goals. As any number of people who have tried to get rid of the excessive costs in real estate can tell you, institutional resistance to change is strong. In principle, the Internet could have eliminated real estate commissions 25 years ago. In practice, not so much. In principle, a digital property database could eliminate the title insurance industry. In practice, the title industry’s grip on Congress is too strong.

After our web site was sold in 1999 (to a subsidiary of the National Association of Realtors(tm), ironically enough), I “retired” to a career of teaching and writing. I taught on a volunteer basis for 15 years at a local high school. I mostly taught AP economics and AP statistics. I hated the AP econ curriculum, because my experience in business had led me to believe that a lot of mainstream economics is poorly conceived. I really liked the statistics curriculum, although one of my students, an autodidact who was a follower of the rationalist community, chided me for teaching a “frequentist” rather than a Bayesian approach.

My goal was to have some long-run influence with at least a few students. I think I was somewhat successful, although by 2015 or so I was struggling with what appeared to me to be the reduced maturity of the students.

I also taught for a few years at George Mason University, at the ridiculous adjunct salary of about $1500 for a class of 100 students. And, as my wife is fond of pointing out, GMU even made me pay for parking. Again, I would have been happy to reach a small number of students with a long-term impact, but I don’t think that I did.

Although I have mostly bitter memories of GMU teaching, I did have one student, Adam Gurri, who I am happy to have encountered.

In the early 2000’s, I looked into health care policy, which seemed to be very much “in play” at the time. At some point, Cato’s Brink Lindsay and Michael Cannon signed me up to write a book, which became Crisis of Abundance. One goal was to explain the consequences of health “insurance” that insulates people from paying for medical procedures, even when procedures are affordable. This model of “insurance” stimulates the use of procedures with high costs and low benefits. The analysis in the book is correct, and I encourage people interested in the topic to read it. But its message got no traction, as far as I can tell.

I had one more entrepreneurial idea, which was to try to develop a platform for teachers to develop apps for classes. I was trying to move the world a little closer to the “primer” of Neal Stephenson’s The Diamond Age. I tried to execute this some time around 2013 or 2014. Attempting the project helped me to catch up on the many changes in the software world that had taken place over the preceding 15 years. But the project did not go well, and I quickly shut it down.

Each of my books was an attempt to have a social impact. My first book, Under the Radar, was a salute to bootstrapping an Internet business. At the time, I was very idealistic about the way that the Internet would allow “the little guy” to start a business without venture capital. Reality turned out to be pretty much the opposite, with VC-backed companies coming to dominate the Internet economy.

With Nick Schulz, I wrote From Poverty to Prosperity (later retitled Hidden Wealth), with a goal of emphasizing the importance of intangible factors in economics. The institutional and cultural setting of an economy is the “operating system.” And the entrepreneurial ideas are the invisible sources of growth. I would call the book an intellectual success, but it did not have much impact.

Another book that came out at the same time was Unchecked and Unbalanced. The goal was to articulate the immense scale of government relative to any corporation, and to propose institutional changes to reduce the monopoly power of government. I always refer to this book as “widely unread,” although George Gilder cited it as an influence on his book Knowledge and Power.

My most commercially successful book was The Three Languages of Politics. I wrote it for a fixed fee and get no royalties. My goal in writing it was to reduce the level of anger in political discourse. Since then, political discourse anger has given way to outrage.

I also wrote Specialization and Trade. My goal was to re-orient economics away from the textbooks that follow the Samuelson tradition or other neoclassical works. I would describe it as a treatise that is deep yet understandable. But not many people have read it, and few appreciate how deep it is.

Recently, I proposed a model for the regulatory state that I call the COO/CA model, meaning a chief operating officer and a chief auditor. I am not going to go on a crusade to get this model adopted. But I think that the essay is worth reading, and I will periodically bring it up to point out the ways that the existing regulatory state is poorly structured.

Last year, I attempted the Fantasy Intellectual Teams project. I recommend this recent podcast with Robert Wright on the goals of the project and what made it difficult to execute. There are people who get the idea and really like it, but I don’t think that the potential appeal is broad enough to make it succeed.

Recently, I offered my thoughts on an alt-academy. I think that this idea has a lot of potential to produce positive social change. It could give bright, motivated young Americans a way out of the college indoctrination complex. And it could be attractive to people from overseas who cannot afford the existing college model.

The best business model for the alt-academy would be to have the professors come from the business world. They would use teaching in the alt-academy to help build a “pipeline” into their business. There are many people in business who have deep knowledge, on non-business topics as well as business topics, that they could use for courses.

Anyway, I intend to discuss this idea more going forward. I think that achieving social change by nudging young people in better directions is what appeals to me most right now.

Of all of my attempts to produce positive social change, the attempt to improve underwriting at Freddie Mac was the most satisfying. This is in spite of the possibility that in the context of the peculiarities of America’s policies that skewed the financial system toward mortgage lending there were adverse unintended consequences.

Hence, my experience is consistent with the view that working in profit-seeking enterprises has the highest probability of producing positive social change and the lowest probability of causing social harm. Profits and losses provide valuable feedback on how well you are serving customers with the resources that you employ.

If you work at a think tank, the ultimate customers are the donors. A major accomplishment at a think tank is getting a policy proposal enacted. One can hope that this in turn produces positive social change. I think that this mechanism is less reliable than seeking to produce change in the private sector.

As a specific example, suppose that you would like to reduce the harms caused by Facebook and Twitter. I think that this is much less likely to come from government regulation than from new firms.