Earlier this week, I wrote a piece about how Bernie Sanders’ economic policies have come under criticism by a number of progressive economists for their unrealistic math, While it remains true that Sanders’ policies are mostly indefensible can’t possibly work, David Dayen at the New Republic, in his response to the criticisms, actually makes a very good point about economic forecasting. He writes “What I do want to challenge is the idea that there’s one serious, evidence-based way to perform economic forecasting. The truth is that most economic forecasts that look several years into the future are flawed, almost by definition.”
The modern obsession with “data driven, evidence based” everything generally overlooks the very real problems you encounter when trying to apply the model of natural science to human behavior. Since all people are different, and behave according to their wills instead of to mere physical laws, you cannot predict with any certainty how complex economic phenomena are going to play out. We can acknowledge that people respond to incentives and examine what the incentives of a particular policy are, but numerical precision is simply outside the scope of most social sciences, as anyone who has ever read an economic forecast must surely know. This is a basic tenet of the Austrian School of economic thought.
Bearing this in mind, I want to talk about a new study from the Economic Policy Institute, which claims that raising the federal minimum wage to $12 an hour by 2020 would result in a $17 billion savings to taxpayers, due to fewer workers needing public assistance. Of course, progressives are making much hay of these findings, claiming that “science” has proven the wisdom and expediency of their preferred policies.
Studies like this are dangerous and misleading, because the average reader, and the average journalist for that matter, is not equipped to spot the serious flaws contained in the research. The author achieves his numbers by doing statistical analysis of the wages of people on government assistance, and combining the results with the number of people currently earning for less than $12 an hour. It’s basically a simple arithmetic problem, which itself should give you a clue to the utter inadequacy of this approach to analyzing the incredibly complex American economy.
The key assumption underlying the conclusion is that employers will in no way react to mandated higher wages. There is no mention of reduced hiring, the inability to meet payrolls, reduced hours, or higher consumer prices to compensate for the higher wages. The paper takes it as given that you can tweak wages for millions of people without anyone noticing or reacting in any way.
It should be obvious to anyone who has spent time in human society that people do not work like this. They adapt to changing circumstances; they try to arrange things to maximize their own benefits and minimize their costs. Employers, when compelled to pay their workers more, will employ fewer workers, reduce hours, and raise prices to make up the difference. Some companies, sneaking by with razor-thin profits already, will be forced to close. The newly unemployed and underemployed people will consume more in government aid, and the higher prices will result in a real reduction of pay for everyone. The Economic Policy Institute pretends these incentive effects don’t exist, and journalists swallow it because it fits into their worldview and because they’ve been trained to think uncritically of anything labeled “evidence-based” or “science,” even when the conclusions fly in the face of basic logic.
The same mindset that led prohibitionists to think that banning alcohol would end drinking now leads foolish economists to think they can dictate wages without negative consequences. The minimum wage doesn’t help the poor or relieve taxpayers; it restricts opportunity and creates dependency for the most vulnerable workers.