The Elusive Quest for Best Practice

People love to find a best practice. Who can blame them? After all, we are confronted with so many problems, and they give us a headache. Wouldn’t it be great if we can follow some best practice and solve all these problems, or at least mostly solve the problems?


In public policy, we see the problem of corruption and bad policy. Wouldn’t it be nice if there is a system that could guarantee the public policy serves the people?

In banking, we see reckless moral hazard. Wouldn’t it be nice if we bind all fund managers to some best practice and manage money soundly?

In research, we see the problem of careless or even fraudulent statistical analysis, and the ensuing non-credible results. Wouldn’t it be great if we can whip researchers all into some best methodology, and take the ‘con’ out of everything?

Fake Solutions

In public policy, we were sold ideologies like democracy and Washington Census. We point to successful cases and ignore failed attempts, and say democracy can cure every country except those who cannot be cured. (borrowing from doctors before the start of Randomized control trials) Or we say, ahh, those are not real democracies, these and that. But face it, democracy is not a panacea. The Dupont story and Flint Water Scandal should give us pause: in a well-run society, discovering and exposing such deeds should not be so difficult. Or consider this, Ohio regulators rubber-stamped the merger between Aetna and Humana (two of the biggest US health insurance company) and kept the decision secret for more than a month.

In banking, we are sold all kinds of ideas: capital requirements, Bessel II, Bessel II, triple A….The 2007 financial crisis speaks to their limits (or absurdity)

In research, we were sold ideas like instrumental variable, natural experiment, propensity score matching. These made us feel like we are in the land of Asymptopia (see the brilliant criticism in Leamer’s paper), where we churn out more non-sense, just in more complicated forms.

Problems with Best Practice

I used to think that looking for best practice is a fool’s errand. Now, it feel it results more from laziness than stupidity. Intellectual laziness. It is the kind of wishful thinking that one can rest sound and tight, and never need to do anything, if one can find the best practice. But you never will.

Sometimes the best practice does not exist. Though people or countries may have similar symptoms, the underlying disease differ. What jump-started western economies (free market) could leave late-comers in a resource curse, while the opposite (industrial policy) actually enable Japan, South Korea, and others to leap ahead. Just as Dani Rodrik pointed out: Diagnostics before prescription.

More fundamentally, the problem with best practice is that it is a static concept: there is something set in stone. It might work well for handling physical objects, like atoms, or chemicals, but it won’t work for people. People find work-arounds and adapt to new circumstances (this idea is at least as old as Lucas critique, but deplorably, economists (including Lucas himself) address this critique with a best practice: rational expectation framework).

Starting with some best practice is great. The problem is most people stop there. It is only a starting point, and is no substitute for real work, like constantly monitoring and continuous problem solving. It is not enough to let people vote, it is paramount to stay vigilant against any forces that stealthily swap the democratic spirit for democratic form.

“Yes, there are all sorts of problems with best practice, but, what is the alternative?” people ask, “Doesn’t is prevent the worse from happening?” Maybe, but at a cost. It gives us a false sense of security, complacency or even over-confidence; it encourages us to join a wild party celebrating “the Great Moderation” without realizing the boat is sinking (yes, some economists were congratulating themselves for having tamed the financial crisis before the 2007 crisis). We might improve a little bit, but also swept the remaining mess under the rug of best practice.

It ain’t what you don’t know that gets you into trouble. It’s what you know for sure that just ain’t so.

The lesser evil paves the way for the greater evil.

Running for Life

When I see all those researchers thoughtlessly implementing Inverse Probability Weighting (IPW) or carrying out a structural estimate, and present their results as credible. I cannot help thinking to myself it is another triple-A rated Credit Default Swap being sold. I should scramble for life.

Like what you read? Give Sun Wei a round of applause.

From a quick cheer to a standing ovation, clap to show how much you enjoyed this story.