The Unprincipled Randomization Principle in Economics and Medicine

Over the past decade randomized field experiments have gained prominence in the toolbox of economics and policy making. Yet randomization enthusiasts have paid little attention to the ethical issues, economic costs, and theoretical difficulties caused by the so called randomization principle. Randomized trials give placebos or no treatment at all to vulnerable individuals, withholding best treatments from the control group. Randomization has been proven to be less precise and less efficient than “Student’s” balanced alternatives - particularly when effect sizes and confounding from unobserved systematic effects are large. From medicine to economics, randomized trials are rarely if ever repeated. A good thing, perhaps, given that randomizers are willing to sacrifice the well-being of study participants in order to “learn”. We consider here the ethics and logic of that sacrifice. We present new evidence from a 25 question survey of randomization, statistical significance, and validity we applied to all the full length articles using randomization techniques and published in the American Economic Review, 2000-2009 and New England Journal of Medicine, 2000-2002. A study of history shows that today’s “principle” of randomization was fabricated out of nothing by R.A. Fisher, in a little known battle he waged in the 1920s against W.S. Gosset aka “Student”. We conclude that the most reliable ethical character of economics, Adam Smith’s “impartial spectator,” would not approve of randomized trials as practiced in economics and medicine