This seems to be the week for passive-aggressive yet self-congratulatory columns: David Brooks has one (with nice replies by Jonathan Chait and Ta-Nehisi Coates), and Stevenson and Wolfers have two, ostensibly about the debate surrounding Reinhart and Rogoff’s empirical results on economic growth and debt, but going to some effort to avoid making any statements about the relationship between growth and debt. The point of the first of their columns was to point out the “weakness of our public discourse over statistical analysis” and the second, entitled “Six Ways to Separate Lies From Statistics,” had,
So how can non-experts and policy makers separate the useful research from the dross? Allow us to offer six rules.
So they’ve written two columns for Bloomberg View about the R/R debate criticizing the analysis and portrayal of the errors, while making no effort to analyze the main conclusions and issues (i.e. does low growth cause higher debt or does higher debt cause low growth) themselves.
Anyway, some of their rules are, IMO, sensible sounding but ridiculous. Their #3, for example:
Be wary of scholars using high-powered statistical techniques as a bludgeon to silence critics who are not specialists. If the author can’t explain what they’re doing in terms you can understand, then you shouldn’t be convinced.
This seems smart, right? But let me propose my own rule, the “Iron law of Macroeconomics:”
Every theoretical and empirical finding has a simple, intuitive, commonsense, and horribly wrong explanation.
Is deficit spending intuitive? Shouldn’t the government “tighten its belt” just like a household? (no). Can we look at high unemployment and conclude that the stimulus was an expensive failure, or should we try to estimate the hypothetical path of the economy under a no-stimulus policy? (the second, which is going to need higher-powered statistics than just looking at the unemployment rate). So I think their third rule is unhelpful. The others are a mix of unhelpful and patronizing as well.
Let me propose a different guideline for their question, “how can non-experts and policy makers separate the useful research from the dross?”
If you want to have a serious, worthwhile opinion on anything, you have to put in the time to understand it. If you don’t put in the time, you shouldn’t invest much emotion in your opinion. There are way too many sensible, intuitive, and wrong answers out there, and too many confident, smooth, and persuasive people pushing those answers (some because they’re dishonest, some because they’re honest but mistaken), for any mental checklist to be much good at all.
Now for a non-threatening example: I like to run. I ran cross-country my senior year of high-school; I walked on the track team my freshman year of college (and walked off…); I’ve run a marathon (SF, once). I run sporadically when I have time to train now. So, I think I have more familiarity and experience with running and fitness than most people have with macroeconomics. But I don’t study this stuff; I don’t spend time researching different training programs. And so I shouldn’t have serious opinions about injury rehab and whether RGIII is or is not doping because he’s beating his rehab schedule. And I shouldn’t have thought that people cared what I thought about Floyd Llandis and Lance Armstrong and whether or not they were cheating. My opinions on these issues are pretty uninformed and objectively shouldn’t matter to other people.
So, if you want to learn how to “separate the research from the dross,” for, say, macroeconomic policy, you need to learn about macroeconomic policy. And it helps to learn in advance, so that you’re relatively dispassionate and unbiased when you’re learning. And, to learn about this sort of thing, it helps to:
- Read widely in that area. Read textbooks, magazine articles, blog posts, popular books, tweeted links, etc., and pay attention to the range of explanations you encounter. Then figure out how to reconcile contradictory views.
- Try to use what you learn (start right away) and pay attention to your failures. For example, what do different views of the economy say about future inflation? About the stock market? Then pay attention to what happens with inflation and the stock market. Which predictions were right? If no one was right, which ones were the least wrong? Try to understand why.
Note that these steps take a long time. Sorry. If I knew of shortcuts, I’d have taken them myself. Also realize that, even if you put in the time, you’ll still make mistakes. Lots. Keep that in mind too. The (historical) example I try to remember is that R.A. Fisher, the person who basically invented modern statistics, strongly believed that smoking did not cause cancer. Mistakes abound.
Edit/update: just to be clear, I think that people should have opinions on this R/R issue (and others); but that it takes work to get up to speed. Please, please, please put in that work! This stuff’s important and you can do it! I don’t think that there’s some 6 or 7 step checklist that will let you figure out whose opinion to adopt as your own, which seems to be the S/W suggestion.
Some interesting news