Experimentation

The other day at Cafe Hayek, Don Boudreaux responded to an e-mailer who argued that minimum wage experiments should be run for social science (readers may recall I faced a similar question this past Fall).  Don raises many good points, but I want to talk on a more broad scale.

Those who make this “experiment for the sake of science” argument make one fatally wrong assumption: that the experiment can only be run by government.  That government is needed for “bold, persistent experimentation,” to borrow the words of FDR.

The reality of the matter, however, is that privately these experiments are constantly being done.  As I write this (and you read it), some people are experimenting paying their people a higher wage.  Some are experimenting paying them less.  Some are experimenting with discriminatory business practices.  Some are experimenting with inclusivity.  Some experiment with imports.  Some experiment with domestic-only.  These experiments are being run day-in, day-out by millions of people. There isn’t any need for government direction.

In fact, probably the worst thing that could happen is for government to conduct social experiments.  Aside from the fact that the times it has in the past haven’t exactly been positive (looking at you “progressive” reformers behind Jim Crow, eugenics, forced sterilization, etc), government tends to crowd out other kinds of experiments, the ones not approved of.

Another great thing about market experimentation is, when it fails, that’s the end of it.  The business loses money (or goes out of business).  Other participants see this as a signal and don’t engage.  Government, however, not so much.  Milton Friedman once said “There is nothing so permanent as a temporary government program.” Governments are, almost by their nature, very bad at admitting failure and cutting their losses.  In fact, since they are often isolated from the costs of their actions (unlike other economic actors), they don’t often know if an experimentation is succeeding or failing.  This makes them terribly unqualified for any kind of experimentation.

I am all for experiment.  Maintaining the status quo is stagnation.  However, I think those experiments should be left to those who bare the costs and who are flexible enough to end it should it become a failure.