The question posed in the title of this post is what makes the task of interpreting economic data (indeed, any scientific data) very difficult. It is also why it is impossible to just “follow the data.” Theory, properly rationalized, must be present in order to make sense of data and evaluate whether it makes sense or not.
Allow me to start with a silly, but 100% true, example:
Every time I have watched a New England Patriots game from start-to-finish on TV for the past 16 seasons, the Patriots have won. That’s quite a lot of data points over a decently long period of time. It’s a correlation of 1 (perfect correlation). If we were to just “follow the data,” one could claim that I somehow cause the Patriots victories. After all, the correlation is there. Of course, such a conclusion would be erroneous. By simply pointing to the theory that it is good coaching and good players that affect the outcome of the game, not someone watching a game in Massachusetts/New Hampshire/Virginia, one could easily refute the claim. Further, if the claimant replied “you’re just being ideological!” in defense of his argument, he would be rightfully ridiculed.
While rightfully derided in sports situations, this kind of analysis is distressingly common in economic situations.* A perfect example is this NELP study and response by Nick Hanauer. Both do exactly the kind of poor analysis I mentioned above; by only looking at minimum wage’s correlation with employment growth (and no other factors), it leads to the incorrect conclusion that the Law of Demand is a “scam.” It’s pretty strong words to call a scientific law a scam based off of one flawed study’s conclusions. Going back to my example above, this would be akin to claiming “Bill Belichick and Tom Brady are scammers because this study proves Patriots victories are related to Jon Murphy watching the game!”
Unfortunately, these same kind of poor analytical outcomes are legion. To quote a list provided by Don Boudreaux:
Forcing wages up by legislative diktat helps workers exclusively at the expense of business owners or rich consumers – or maybe even at the expense of no one at all? Check! Allowing people to buy especially low-priced imports harms the domestic economy? Check! Trade deficits are both a signal and a source of domestic economic decline? Check! The destruction by natural disasters of buildings, inventories, and infrastructure is really an economic blessing? Check! Markets unregulated by politicians and bureaucrats poison consumers with foul foods, kill homeowners with shoddy construction, maim workers with perilous workplace conditions, and cheat savers with fraudulent investment services? Check! The simple key to a booming economy is maximum spending, especially by government? Check! Government debt held by that government’s citizens is no burden upon that economy because we owe it to ourselves? Check! Growing income and wealth inequality means stagnant economic fortunes for the middle class and increasing poverty for the poor? Check! In markets lightly regulated and lightly taxed by governments, the rich get richer and the poor get poorer? Check! In markets women, blacks, and other minorities are underpaid because of discrimination – a problem that can be solved only by government regulation? Check! Economic growth devours precious resources, making these resources ever-more scarce? Check! The more you “buy local” the more you enrich your community and protect the natural environment? Check! Free lunches abound? Check!!
Each one of these conclusions, which fly in the face of theory, are very popular (for various reasons), and yet they all suffer from the same problem I mentioned earlier.
Economic analysis is difficult. One shouldn’t be so willing to discard theory lightly. The important question to ask, regardless of the field, is this: “Does this make sense? What else could be influencing this?”
*I suppose other sciences it is as well, but I do not know enough about them to say anything definitive.