Will Macroeconomists Ever Agree?

Kevin Drum wonders if macroeconomists will ever be able to agree:

The part I can't figure out is why there's so much contention even within the field. In physics and climate science, the cranks are almost all nonspecialists with an axe to grind. Actual practitioners agree pretty broadly on at least the basics. But in macroeconomics you don't have that. There are still polar disagreements among top names on some of the most basic questions. Even given the complexity of the field, that's a bit of a mystery. It's understandable that economics is a more politicized field than physics, but in practice it seems to be almost 100 percent politicized, with the battles fought out by streams of Greek letters demonstrating, as Matt says, just about anything. I wonder if this is ever likely to change? Or will changes in the real world always outpace our ability to build consensus on how the economy actually works?

I took a shot at answering this in April 2011:

... Why can’t economists tell us what happens when government spending goes up or down, taxes change, or the Fed changes monetary policy? The stumbling block is that economics is fundamentally a non-experimental science, particularly in the realm of macroeconomics. Unlike disciplines such as physics, we can't go into the laboratory and rerun the economy again and again under different conditions to measure, say, the average effect of monetary and fiscal policy. We only have one realization of the macroeconomy to use to answer important policy questions, and that limits the precision of the answers we can give. In addition, because the data are historical rather than experimental, we cannot look at the relationships among a set of variables in isolation while holding all the other variables constant as you might do in a lab and this also reduces the precision of our estimates.

Because we only have a single realization of history rather than laboratory data to investigate economic issues, macroeconomic theorists have full knowledge of past data as they build their models. It would be a waste of time to build a model that doesn't fit this one realization of the macroeconomy, and fit it well, and that is precisely what has been done. Unfortunately, there are two models that fit the data, and the two models have vastly different implications for monetary and fiscal policy. ... [This leads to passionate debates about which model is best.]

But even if we had perfect models and perfect data, there would still be uncertainties and disagreements over the proper course of policy. Economists are hindered by the fact that people and institutions change over time in a way that the laws of physics do not. Thus, even if we had the ability to do controlled and careful experiments, there is no guarantee that what we learn would remain valid in the future.

Suppose that we somehow overcome every one of these problems. Even then, disagreements about economic policy would persist in the political arena. Even with full knowledge about how, say, a change in government spending financed by a tax increase will affect the economy now and in the future, ideological differences across individuals will lead to different views on the net social value of these policies. Those on the left tend to value the benefits higher, and place less weight on the costs than those on the right and this leads to fundamental, insoluble differences over the course of economic policy. ...

Progress in economics may someday narrow the partisan divide over economic policy, but even perfect knowledge about the economy won’t eliminate the ideological differences that are the source of so much passion in our political discourse.

A follow-up post in February empahsizes the point that it is not at all clear that the strong divides in economics can be settled with data, but it's not completely hopeless:

...the ability to choose one model over the other is not quite as hopeless as I’ve implied. New data and recent events like the Great Recession push these models into unchartered territory and provide a way to assess which model provides better predictions. However, because of our reliance on historical data this is a slow process – we have to wait for data to accumulate – and there’s no guarantee that once we are finally able to pit one model against the other we will be able to crown a winner. Both models could fail...

I think the Great recession has, for example, provided evidence that the NK model provides a better explanation of events than its competitors, but it is far from a satisfactory construction and it would be hard to call its forecasting and explanatory abilities a success.

Here's another post from the past (Sept. 2009) on this topic:

... There is no grand, unifying theoretical structure in economics. We do not have one model that rules them all. Instead, what we have are models that are good at answering some questions - the ones they were built to answer - and not so good at answering others.

If I want to think about inflation in the very long run, the classical model and the quantity theory is a very good guide. But the model is not very good at looking at the short-run. For questions about how output and other variables move over the business cycle and for advice on what to do about it, I find the Keynesian model in its modern form (i.e. the New Keynesian model) to be much more informative than other models that are presently available (as to how far this kind of "eclecticism" will get you in academia, I'll just note that this is exactly the advice Mishkin gives in his textbook on monetary theory and policy).

But the New Keynesian model has its limits. It was built to capture "ordinary" business cycles driven by price rigidities of the sort that can be captured by the Calvo model model of price rigidity. The standard versions of this model do not explain how financial collapse of the type we just witnessed come about, hence they have little to say about what to do about them (which makes me suspicious of the results touted by people using multipliers derived from DSGE models based upon ordinary price rigidities). For these types of disturbances, we need some other type of model, but it is not clear what model is needed. There is no generally accepted model of financial catastrophe that captures the variety of financial market failures we have seen in the past.

But what model do we use? Do we go back to old Keynes, to the 1978 model that Robert Gordon likes, do we take some of the variations of the New Keynesian model that include effects such as financial accelerators and try to enhance those, is that the right direction to proceed? Are the Austrians right? Do we focus on Minsky? Or do we need a model that we haven't discovered yet?

We don't know, and until we do, I will continue to use the model I think gives the best answer to the question being asked. The reason that many of us looked backward for a model to help us understand the present crisis is that none of the current models were capable of explaining what we were going through. The models were largely constructed to analyze policy is the context of a Great Moderation, i.e. within a fairly stable environment. They had little to say about financial meltdown. My first reaction was to ask if the New Keynesian model had any derivative forms that would allow us to gain insight into the crisis and what to do about it and, while there were some attempts in that direction, the work was somewhat isolated and had not gone through the thorough analysis that is needed to develop robust policy prescriptions. There was something to learn from these models, but they really weren't up to the task of delivering specific answers. That may come, but we aren't there yet.

So, if nothing in the present is adequate, you begin to look to the past. The Keynesian model was constructed to look at exactly the kinds of questions we needed to answer, and as long as you are aware of the limitations of this framework - the ones that modern theory has discovered - it does provide you with a means of thinking about how economies operate when they are running at less than full employment. This model had already worried about fiscal policy at the zero interest rate bound, it had already thought about Says law, the paradox of thrift, monetary versus fiscal policy, changing interest and investment elasticities in a  crisis, etc., etc., etc. We were in the middle of a crisis and didn't have time to wait for new theory to be developed, we needed answers, answers that the elegant models that had been constructed over the last few decades simply could not provide. The Keyneisan model did provide answers. We knew the answers had limitations - we were aware of the theoretical developments in modern macro and what they implied about the old Keynesian model - but it also provided guidance at a time when guidance was needed, and it did so within a theoretical structure that was built to be useful at times like we were facing. I wish we had better answers, but we didn't, so we did the best we could, and the best we could involved at least asking what the Keynesian model would tell us, and then asking if that advice has any relevance today. Sometimes if didn't, but that was no reason to ignore the answers when it did.

Part of the disagreement is over the ability of this approach -- using an older model guided by newer insights (e.g. that expectations of future output matter for the "IS curve") -- to deliver reliable answers and policy prescriptions.

More on this from another past post (March 2009):

Models are built to answer questions, and the models economists have been using do, in fact, help us find answers to some important questions. But the models were not very good (at all) at answering the questions that are important right now. They have been largely stripped of their usefulness for actual policy in a world where markets simply break down.

The reason is that in order to get to mathematical forms that can be solved, the models had to be simplified. And when they are simplified, something must be sacrificed. So what do you sacrifice? Hopefully, it is the ability to answer questions that are the least important, so the modeling choices that are made reveal what the modelers though was most and least important.

The models we built were very useful for asking whether the federal funds rate should go up or down a quarter point when the economy was hovering in the neighborhood of full employment ,or when we found ourselves in mild, "normal" recessions. The models could tell us what type of monetary policy rule is best for stabilizing the economy. But the models had almost nothing to say about a world where markets melt down, where prices depart from fundamentals, or when markets are incomplete. When this crisis hit, I looked into our tool bag of models and policy recommendations and came up empty for the most part. It was disappointing. There was really no choice but to go back to older Keynesian style models for insight.

The reason the Keynesian model is finding new life is that it specifically built to answer the questions that are important at the moment. The theorists who built modern macro models, those largely in control of where the profession has spent its effort in recent decades,; did not even envision that this could happen, let alone build it into their models. Markets work, they don't break down, so why waste time thinking about those possibilities.

So it's not the math, the modeling choices that were made and the inevitable sacrifices to reality that entails reflected the importance those making the choices gave to various questions. We weren't forced to this end by the mathematics, we asked the wrong questions and built the wrong models.

New Keynesians have been trying to answer: Can we, using equilibrium models with rational agents and complete markets, add frictions to the model - e.g. sluggish wage and price adjustment - you'll see this called "Calvo pricing" - in a way that allows us to approximate the actual movements in key macroeconomic variables of the last 40 or 50 years.

Real Business Cycle theorists also use equilibrium models with rational agents and complete markets, and they look at whether supply-side shocks such as shocks to productivity or labor supply can, by themselves, explain movements in the economy. They largely reject demand-side explanations for movements in macro variables.

The fight - and main question in academics - has been about what drives macroeconomic variables in normal times, demand-side shocks (monetary policy, fiscal policy, investment, net exports) or supply-side shocks (productivity, labor supply). And it's been a fairly brutal fight at times - you've seen some of that come out during the current policy debate. That debate within the profession has dictated the research agenda.

What happens in non-normal times, i.e. when markets break down, or when markets are not complete, agents are not rational, etc., was far down the agenda of important questions, partly because those in control of the journals, those who largely dictated the direction of research, did not think those questions were very important (some don't even believe that policy can help the economy, so why put effort into studying it?).

I think that the current crisis has dealt a bigger blow to macroeconomic theory and modeling than many of us realize.

Here's yet another past post (August 2009) on the general topic of the usefulness of macroeconomic models, though I'm not quite as bullish on the ability of existing models to provide guidance as I was when I wrote this. The point is that although many people use forecasting ability as a metric to measure the usefulness of models (because where the economy is headed is the most improtant question to them), that's not the only use of these models:

Are Macroeconomic Models Useful?: There has been no shortage of effort devoted to predicting earthquakes, yet we still can't see them coming far enough in advance to move people to safety. When a big earthquake hits, it is a surprise. We may be able to look at the data after the fact and see that certain stresses were building, so it looks like we should have known an earthquake was going to occur at any moment, but these sorts of retrospective analyses have not allowed us to predict the next one. The exact timing and location is always a surprise.

Does that mean that science has failed? Should we criticize the models as useless?

No. There are two uses of models. One is to understand how the world works, another is to make predictions about the future. We may never be able to predict earthquakes far enough in advance and with enough specificity to allow us time to move to safety before they occur, but that doesn't prevent us from understanding the science underlying earthquakes. Perhaps as our understanding increases prediction will be possible, and for that reason scientists shouldn't give up trying to improve their models, but for now we simply cannot predict the arrival of earthquakes.

However, even though earthquakes cannot be predicted, at least not yet, it would be wrong to conclude that science has nothing to offer. First, understanding how earthquakes occur can help us design buildings and make other changes to limit the damage even if we don't know exactly when an earthquake will occur. Second, if an earthquake happens and, despite our best efforts to insulate against it there are still substantial consequences, science can help us to offset and limit the damage. To name just one example, the science surrounding disease transmission helps use to avoid contaminated water supplies after a disaster, something that often compounds tragedy when this science is not available. But there are lots of other things we can do as well, including using the models to determine where help is most needed.

So even if we cannot predict earthquakes, and we can't, the models are still useful for understanding how earthquakes happen. This understanding is valuable because it helps us to prepare for disasters in advance, and to determine policies that will minimize their impact after they happen.

All of this can be applied to macroeconomics. Whether or not we should have predicted the financial earthquake is a question that has been debated extensively, so I am going to set that aside. One side says financial market price changes, like earthquakes, are inherently unpredictable -- we will never predict them no matter how good our models get (the efficient markets types). The other side says the stresses that were building were obvious. Like the stresses that build when tectonic plates moving in opposite directions rub against each other, it was only a question of when, not if. (But even when increasing stress between two plates is observable, scientists cannot tell you for sure if a series of small earthquakes will relieve the stress and do little harm, or if there will be one big adjustment that relieves the stress all at once. With respect to the financial crisis, economists expected lots of little, small harm causing adjustments, instead we got the "big one," and the "buildings and other structures" we thought could withstand the shock all came crumbling down. ...

Whether the financial crisis should have been predicted or not, the fact that it wasn't predicted does not mean that macroeconomic models are useless any more than the failure to predict earthquakes implies that earthquake science is useless. As with earthquakes, even when prediction is not possible (or missed), the models can still help us to understand how these shocks occur. That understanding is useful for getting ready for the next shock, or even preventing it, and for minimizing the consequences of shocks that do occur. 

But we have done much better at dealing with the consequences of unexpected shocks ex-post than we have at getting ready for these a priori. Our equivalent of getting buildings ready for an earthquake before it happens is to use changes in institutions and regulations to insulate the financial sector and the larger economy from the negative consequences of financial and other shocks. Here I think economists made mistakes - our "buildings" were not strong enough to withstand the earthquake that hit. We could argue that the shock was so big that no amount of reasonable advance preparation would have stopped the "building" from collapsing, but I think it's more the case that enough time has passed since the last big financial earthquake that we forgot what we needed to do. We allowed new buildings to be constructed without the proper safeguards.

However, that doesn't mean the models themselves were useless. The models were there and could have provided guidance, but the implied "building codes" were ignored. Greenspan and others assumed no private builder would ever construct a building that couldn't withstand an earthquake, the market would force them to take this into consideration. But they were wrong about that, and even Greenspan now admits that government building codes are necessary. It wasn't the models, it was how they were used (or rather not used) that prevented us from putting safeguards into place. ...

I'd argue that our most successful use of models has been in cleaning up after shocks rather than predicting, preventing, or insulating against them through pre-crisis preparation. When despite our best effort to prevent it or to minimize its impact a priori, we get a recession anyway, we can use our models as a guide to monetary, fiscal, and other policies that help to reduce the consequences of the shock (this is the equivalent of, after a disaster hits, making sure that the water is safe to drink, people have food to eat, there is a plan for rebuilding quickly and efficiently, etc.). As noted above, we haven't done a very good job at predicting big crises, and we could have done a much better job at implementing regulatory and institutional changes that prevent or limit the impact of shocks. But we do a pretty good job of stepping in with policy actions that minimize the impact of shocks after they occur. This recession was bad, but it wasn't another Great Depression like it might have been without policy intervention.

Whether or not we will ever be able to predict recessions reliably, it's important to recognize that our models still provide considerable guidance for actions we can take before and after large shocks that minimize their impact and maybe even prevent them altogether (though we will have to do a better job of listening to what the models have to say). Prediction is important, but it's not the only use of models.