Jim Manzi responds to my post. It seems I came off a bit harsher than I intended. Other posts lead me to think that some believe I’m rejecting Manzi’s argument against overconfidence in models. Quite the contrary I am suggesting that academics don’t actually have the level of confidence Manzi and Brooks ascribe.
This post is going to be a bit long but I it seems our miscommunication is deep and it will take a bit to get us back on track.
The basic insight – echoed by Manzi and Brooks – that there are unintended consequences and that the economy is cannot be meaningfully guided by government is core to everything we do. The further insight that we cannot even build a model that roughly estimates all the particular decisions that private actors make was hammered further home by the failure of the infamous 500 equation models.
Some of these models are still up and running but they serve more of a war gamming function than an actual policy calculators. If you look at how the Fed policy makers talk about choices its in very general terms. If you look at how macro theoretic models are written they are in a very general form.
Manzi and Brooks seem to believe that policy advice comes plugging and chugging on the big models but instead it comes from an intuition honed by working with both the simple theoretic models and war gamming the big models.
I shouldn’t speak for other academics but when I advise that stimulus be undertaken or that money be loosened it is not because I think that I have a model which accurately maps how said stimulus will impact the various sectors of the economy. I would laugh at such an assumption and I am an explicit model builder. I am indeed, currently working on a model to allow war gamming on long term state budgets.
The basic tact both in the models and in my thinking is that for the most part markets work and that they will direct resources to whatever their best use is.
Our task is to deal with particular market failures. In the case of recession the market failure is that we do not have perfectly flexible wages and prices. If we did there would be no recessions as we know them.
I sense that I need to write more on this because many smart commentators still see recessions as either a calamity akin to a crop failure or as punishment for excesses. If that was what a recession was then the result should be our working harder to overcome the calamity or make up for our excesses.
However, the key feature of an actual recession is that Aggregate Hours worked FALLS. That is, people work less, produce less, create less. This is not a solution to a calamity or overconsumption.
In a world of perfectly clearing markets a fall in Aggregate Hours would represent a Great Vacation. That doesn’t seem like the proper response to a calamity or previous irresponsibility. Thus our search for a market failure.
When we work through our best guess at the source of this failure the answer tells me that: lower taxes, higher government spending and looser money would all serve to lean against this particular market failure.
I tend to advocate looser money for a variety of reasons. However, not least is that its “cleaner” than the other approaches. Hopefully it will become clear why.
Several times I have accused Manzi of slipping into industrial policy.
What I mean here is that Manzi is thinking of policies not in terms how then lean against the problem of inflexible prices, but in how they might set the stage for long term economic growth. This is a question that a lot of people are interested in but it is a distinctly different question than “how do we smooth the business cycle”
Nonetheless, once we open the door to government spending people are going to naturally want to engage in industrial policy. To paraphrase Keynes , burying hundred dollar bills in abandoned mine shafts is perfectly good stimulus. Indeed, it would likely have a higher multiplier than other forms we have tried.
The reason people turn to building roads, or funding research is because they hope not only to get stimulus but that the newly employed labor will serve a useful service for society going forward. The question then becomes: what would be a useful service for society going forward. This general becomes a question of industrial policy.
Manzi is right that industrial policy is hard. This is why I argue that we should not attempt to mix it with stimulus and instead fiscal stimulus should be performed by sending the needy or state governments cash.
The larger point, however, is that one shouldn’t confuse the argument for stimulus with the argument for industrial policy. Nor should we think that either the big war gamming models or the simple theoretical models have anything to say about industrial policy.
So one can rightfully criticize industrial policy without criticizing stimulus. Now before I move on I should say that you could also take a market failure approach to spending money. For example, you might argue that infrastructure is a public good and therefore we should spend more on infrastructure to cure two market failures at once. This is different than industrial policy because the argument is not that infrastructure will produce growth or an expansion of industry.
The argument is simply that left to its own devices the market will produce too little infrastructure. The benefit could be lower congestion, fewer fatalities or more trips to Atlantic Beach. We don’t know what people’s preferences are and we are not aiming for a particular result. We are simply saying that whatever the preferences are, the market will produce too few roads.
Now, to be clear Paul Krugman, for one, explicitly rejected this rationale for government spending in a recession, saying that he only preferred government spending because of its greater traction.
That is, cutting taxes and loosening money work assuming that there aren’t other major market failings besides the sticky price problem. Government spending should work in the presence of a host of market failures.
A second stimulus approach would be to cut taxes. This is more complex than government spending because to work against the particular market failure of sticky prices people must spend the money. Now spending doesn’t necessarily mean consumption. Investment is perfectly good spending.
What they cannot do is buy short term government bonds or place their money in a simple savings vehicle which will then in turn invest in short term government bonds. In times of uncertainty, however, people have a tendency to do just that.
There is still some tendency to spend more because of tax cuts. Measuring it is difficult for reasons I’ll have to save till later.
Taxes, however, often slide into some of the same messy issues as government spending. Not only is their fighting over who will get the tax cut but there are attempt to create tax cuts which some believe will have long term effects. These effects are sometimes referred to as supply side effects because they are attempt to boost Aggregate Supply not Aggregate Demand. Broadly speaking recessions are a problem of Aggregate Demand while growth is an issue of Aggregate Supply.
Thus once again concerns about long run growth are mixed with concerns about stimulus. One may or may not think supply side tax cuts are a good idea but this is separate from whether one thinks stimulus is a good idea. Just as we should not confuse government stimulus with industrial policy, we should not confuse tax cut based stimulus with supply side growth based cuts.
Like the two above monetary policy should induce stimulus. Indeed, I tend to think that the other two only work because they implicitly act through monetary policy but that is a debate for another time. The important thing is that monetary policy works to lean against the specific market failure inflexible prices.
Fortunately from my perspective pure monetary policy involves swapping one government liability for another. The Fed buys Treasury bonds and issues Federal Reserve Notes, also known as money.
There is in this action no necessary presumption about how the private sector should allocate resources. The Fed is simply discouraging private actors from attempting to give their money to the government by lowering the supply of bonds and discouraging them from holding excess cash by increasing the rate of inflation.
If that means that private agents invest in new technologies, then so be it. If it means that they buy a bunch of X-box games, then so be it. So long as they refrain from holding government assets as a store of value then it is entirely up to them what they choose to hold or whether they choose to spend.
In addition, once wages and prices adjust the long term affect of monetary policy should disappear. We will have a different price level but the relative prices of all goods should be essentially the same.
It is true as some people point out that if people engage in investment that they would have otherwise forgone then this will raise the level of growth. This is true but unless that investment is in the form of a non-decreasing returns to scale technology it should not affect the long run.
More investment today lowers the return on investment tomorrow. This is why developing countries can grow so much faster than developed countries once they have the proper institutions. They have had less investment in the past and so each unit of investment is more productive.
In the long run though everyone is stuck in the vicinity of the technological frontier. We could shoot ahead but we would simply be slowed down later.
Thus monetary policy should have no large predictable effects on the economy. There are always butterfly effect type stories we could tell – a crucial company getting funding at just the right time, etc – but from our perspective this is white noise. We could just as easily crush the butterfly as set him free.