Tuesday, June 4, 2013

Krugman versus Reinhart & Rogoff

Cartoon by Jim Gibson


Imagine the bar fight when Danny accused Mike of not being a scientist. You aren’t a scientist – nah nah nah. Yea, but your mom wears combat boots. It sounds pretty stupid – and sounds even stupider when Economist Krugman dukes it out with colleagues Reinhart and Rogoff.  I am guessing they are not having lunch together at the faculty table. Is Macroeconomics a science? Are these guys scientists? And why does it matter? In brief I would say yes, yes, and because we think it matters even if it doesn’t.


Wikipedia says a science is “a systematic enterprise that builds and organizes knowledge in the form of testable hypotheses and predictions about the universe…refers to a body of knowledge itself, of the type that can be rationally explained and reliably applied.”

For those of you not sufficiently trained in globbygook, just about anything fits this definition of a science if it is trying to “figure-out” something. Your two-year old puts your best carving knife into the light socket and you explain to him that there is something called ‘lectricity and note that the knife will take the ‘lectricity out of the wall and into your body and it will hurt.  You might not know it but you are acting as a physicist and you are using a body of knowledge called physics. Physics explains why the electricity will go into your body and harm you. Hundreds of other kids have already tested the hypothesis and laid a solid empirical foundation to support the theory. Pretty cool – physics is obviously a real science.

But so is macroeconomics. I can’t tell you how many hours I spent theorizing about econ stuff and how many more hours I spent running little experiments that end up either supporting or not supporting hypotheses.  For my doctoral dissertation I had a theory that said that Nixon’s Wage and Price Controls of the early 1970s would reduce inflation only if they reduced inflationary expectations. I spent about a year drinking JD and formulating a model, gathering data, and then running a bunch of statistical tests that might reject or not reject my theory.  To be scientific you need to start with a problem that you want to explain, have an explanation, and then try to put your explanation to the test. If your explanation fails the test, then you move on to sales or painting. If your explanation does not fail the test, then you hold on to it for a while – a least until someone has a better explanation than yours. It is like being a sheriff in the cowboy days of the Old West. You were sheriff until someone came along who was faster at the draw. Truth, like sheriffs, evolves. 

If you are still awake, you might retort that while economists pretend to do all that stuff, they are not REAL scientists like physicists and geneticists and sexologists. And for that opinion, you would be wrong. Why? Because Wikipedia cannot be wrong.  Look again at the definition of a science and you will see that the only real criterion is that you are trying to test your hypotheses. There is no single criterion that says whether your test is strong enough.  There is no single test of a theory being “good enough.”  You are the sheriff until someone beats you to the draw.

You might be surprised but if you think about it, it isn’t so strange. Take meteorology.  There are some things these weather people know and get right every single time. If a cloud reaches a saturation point then it is very likely to rain soon. But if you ever lived in Florida during hurricane season you know that there are some things that meteorologists almost never get right. Where exactly will the full force of a hurricane hit Sanibel Island? When will it arrive? How long will it stay there? These are very critical and important questions. Meteorology has a lot of good theories and they have put these theories to the test many times – but they still don’t succeed much when it comes to the most important questions.  Geneticists know a cell deformity or marker when they see one. But they still do not know when or if YOU will get pancreatic cancer. Other medical sciences can treat the symptoms of your cold but they cannot stop millions of people catching colds every year. 

Every science is the same. There are no REAL sciences. You either use the scientific method to advance knowledge or don't. Every science has things it can easily explain and other things it cannot. Let’s face it we have contrasting theories with respect to every aspect of our life. Should we floss or gargle? Take vitamins? Use synthetic oil in your car? Should I have my prostate checked every year? Who was better the Beatles or the Stones?

I am not belittling science. There is much we know. I am just making the point that there is much we don’t know. The low-hanging fruit is gone and every science is in the process of discovering new and important things. In that discovery process there are always competing models, competing scientists, and competing opinions. This is true in macroeconomics and it is true in physics.

So what’s the rub with respect to macro? The answer is that there is room for difference of opinion in some very important areas. Think of what we are trying to accomplish with policy. It looks easy on a power point slide – but how do we REALLY get the US economy to grow faster? Note that whatever policy we deliver, it will involve the actions and reactions of firms and households, spenders and savers, investors, domestic citizens and foreigners, and so on. Psychology and expectations will always be at work. The ages of the population this year and next will impact how policies create changes in behavior. It is a wonder that we ever get this question right. What size ark should we build when the floods come? I am guessing we could get some very different answers to that question too.

What galls me is not that economists get things wrong and not that they disagree about policy. What galls me is when they do not even try to act like scientists. Not checking data that supports a very strong policy position is poor science. Shooting from the hip without any real attempt to utilize a large body of information and data is likewise unforgivable. Arguing in voices that sound more like competitors in divorce court does nothing but reduce the confidence that people have in macro science. Economists need to put on their big-boy pants and quit acting like children. There will always be politics and ideology. The best economists will try their hardest to be above that fray. 

12 comments:

  1. Philosopher Karl Popper defined something as scientific if there is a logical possibility of it being found false. Economists see economics as a science under Popper's definition.
    Economics has a set of theories that explains how economies work. Economists developed these theories after extensive observation, research and critical analysis. These are elements of the scientific method. Therefore,......

    Just because weather guessers get it wrong about 30% of the time doesn't make meteorology any less a science. Likewise, just because economists miss a great deal doesn't make economics of any type less of a science. Weather models can be impacted by unforeseen phenomena which make 90%+ accuracy very difficult. Economics is also influenced by outside forces which most times are unpredictable, for instance, the Fed, Congress, and human nature. Remove those factors and we'd probably see economics perform like Tiger Woods pre-9 iron incident.

    I remember you drinking JD and chasing models as an undergrad at GT. I see nothing changed much as a PhD candidate. BTW, Danny and I never had a fight.



    ReplyDelete
    Replies
    1. Fuzz, I said "imagine" such a fight. I can't imagine you two weenies fighting actually! Anyway thanks for the back-up with Popper -- apparently you got educated somewhere along the way! :-) Apparently I did better with the JD than with the models...

      Delete
  2. Nice piece...well done prof!

    ReplyDelete
  3. There are always variables and the trick is to identify the ones that demonstrate statistical reliability. Run what if models with them and then do an assessment once trend lines are established. Keep the politics out. If the scientist infiltrates feelings then those will influence his/her work and the interpretation of the outcome. Economics is a dismal science but also one without feelings.

    ReplyDelete
    Replies
    1. Right on. Wasn't there a song -- Feelings wo wo feelings...? I guess it wasn't for economists.

      Delete
  4. Dear LSD. One branch of science not mentioned—at least I don’t recall it being mentioned—might have been when I was asleep while reading your stuff—is medicine. Note that docs say they practice medicine, but stop just short of saying they know the cause of a malady or the specific “sure fire” cure. They prescribe antibiotics, salves, massages, shock treatments, acupuncture, surgery and give hope to the patient that he/she will heal, but if healing doesn’t occur they say take an aspirin and call in the morning. Sometimes the prescribed cure kills the patient, exacerbates the situation, or delays healing, and sometime does cure. Economists practicing econ are like docs practicing medicine.

    Often folks know the causes of maladies and those with enough sense avoid the known causes to avoid the predicted maladies. Too bad some economists don’t simply admit to known causes/effects and continue to practice econ anyway—is it possible to revoke an economist’s license? Those sorts sort of remind me of voodoo medicine.

    I didn’t read the debate between Krugman versus Reinhart & Rogoff but I assume it was about macro and how to conduct monetary/fiscal policy. I’ll bet none said simply that a govomit should live within its means, similar to a doc saying everything but in moderation—stay within known limits and use reason. Sort of like common sense isn’t all that common.

    ReplyDelete
  5. Dear Charles,

    The medicine/macro analogy is really good -- thanks for that perspective. R&R pretty much took the side you suggest -- the common sense notion that debt matters. Unfortunately they made some technical errors that weakened their results and Krugman and others have jumped all over them. Most economists can disagree for good reasons. Especially when there is not a global panic. To continue with medicine -- think about Dr. House. He and his staff only get involved with the toughest cases -- the cases where there is no simple solution. So they try lots of things. Sometimes what they try doesn't always make sense. Traditional medical science often offers no straight forward solution in those cases so there is room for a lot of difference of opinion. Not all opinions are correct but the situation is such that most are at least considered. In most years macro is not so complicated and controversial -- but after 2007 things got pretty murky and it opened the door to lots of voodoo. No Charles,licenses are not withdrawn.

    ReplyDelete
  6. Careful there, Chuck. My old grandpa always told me that excessive exacerbation can cause blindness. I think that's Krugman's problem.

    ReplyDelete
  7. Dear LSD. Licenses not withdrawn—REVOKED.

    Thanks, Fuzz Meister—good laugh early this a.m. My mother told me I’d be wearing white gloves if she found evidence of excessive exacerbation—I think Krugman should be required to put them on.

    ReplyDelete
    Replies
    1. Boys - this is not an r-rated macro site. Please watch your exacerbations.

      Delete
  8. Yeah, LSD. NSA might be watching.

    ReplyDelete