BrianGarst.com

Data is important in public policy. In many situation, the right piece of information can settle a dispute, though even then it probably won’t. But sometimes data can mislead, especially if the interpreter lacks any theoretical basis upon which to judge the information.

Let me give an example. Better yet (for me), let me borrow one from Don Boudreaux, though I’ll present the information in a slightly different order.

Suppose a new business opens in the area, let’s call it Mal Wart, and it employs mostly low skilled labor and pays low wages. Sometime after the business opens, a new survey finds that average wages in the area have declined. Has Mal Wart made the people around it worse off?

Looking just at the data I described, many would say yes. But they would most likely be wrong. Consider Boudreaux’s example:

Suppose that only just yesterday did some clever entrepreneur – seeking only to increase his own material wealth – figure out a way to profitably employ workers each of whom contributes (say) $10.00 or less per hour to any employers’ revenues.  Until this entrepreneur – let’s call him Wally Marsh – devised this new business plan (or technological breakthrough), no employer in the United States has ever before found it profitable to employ any workers whose productivities are so low.

Profit-grabbing capitalist that he is, Wally implements his innovation and hires scores of low-skilled workers, paying each of these workers wages commensurate with their productivity, with none of these newly hired workers being paid more than $10 per hour.

What happens to the average and median wage rate in the U.S. as a result of Wally’s innovation?  (Answer: it falls.  Before Wally’s innovation, only workers who produced $10.01 per hour or more were employed; no worker, therefore, was paid an hourly wage less than $10.01.)

The data in this case is not wrong, but realizing what it measures and what it does not is crucial to properly interpreting the information. A person with a theoretical foundation in economics is likely to properly see what the data measures and means than a person without such a foundation. This is why theory matters and not everything can easily be reduced to a simple empirical test.

Boudreaux’s post brought to mind another good example of frequently misinterpreted data involving income mobility. This video from LearnLiberty explains: