It looks like I will be contributing to a set of essays on this topic. Here is what I am thinking of in terms of an outline.
1. What is Keynesian economics?
I think that there are two important versions. There is the folk Keynesianism of policy makers and journalists, and there is the academic Keynesianism of graduate school and peer-reviewed papers. When pressed to give a narrative interpretation and discuss policy, economists tend to fall back on folk Keynesianism. When other economist object to some basic flaws in folk Keynesian economics, macroeconomists fall back on academic economics. In Greg Mankiw’s terminology, the policy engineer does not make use of the scientific paper, but he can pull a copy out of his back pocket if somebody asks to see it.
What I call Folk Keynesianism is what is used in narrative interpretation and policy discussions. Spending creates jobs and jobs create spending. The Folk Keynesian economy is one in which the price mechanism for adjustment and clearing markets has disappeared. Instead, quantities determine other quantities.
The other version is what I call academic Keynesianism. Think of it as an attempt to introduce re-introduce prices into an otherwise Keynesian economy. You start with Hicks sneaking the interest rate back in as a determinant of investment. Then you have the Phelps volume sneaking the real wage back in as a determinant of employment. Then you have the Lucas critique, which is answered by bringing in intertemporal optimization. You end up with Blanchard’s “state of macro” paper.
2. What is the alternative?
PSST, of course, which interprets macroeconomic phenomena as the problem of coordinating in a world of very complex trading patterns. Compared with folk Keynesianism, PSST fails to offer as compelling a basis for a narrative interpretation and policy discussions. If you want a satisfying narrative that probably is false, then stick with Keynesianism.
3. What can we learn from the data?
There are at least two problems with macroeconomic data. One is the fact that it is non-experimental. We cannot test important hypotheses against one another. It is easy to impose competing narratives on the same data. I might also mention Leamer’s attempts to document patterns in the data.
The other problem is that the classification and aggregation of data is questionable. According to economic theory, education is an investment. But in the national income accounts, what consumers spend on education is called consumption. What government spends on education is government consumption. And even what businesses spend on training is not typically included in investment. If I buy a cell phone for my household, it is consumption. If my company buys me a cell phone, it counts as investment. Much of what is counted as labor income in the national income accounts is actually a return on investment. The concept of “the real wage rate” is ridiculous. Neither the numerator nor the denominator is the same across individuals. Aggregate productivity statistics are also ridiculous.
4. Where did Keynesian economics go wrong? I think it focused too much on (the lack of) price adjustment, and not enough on trial-and-error adjustment. And, more fundamentally, I think it errs by thinking in crude aggregate terms. The economy is not one giant GDP factory. It is a pattern of specialization and trade.
I think of Keynesianism as stop-the-bleeding economics, and Hayekian/Schumpeterian as growth-support econ. Former needed in deflation spiral to prevent govt overthrow, latter needed in normal times to grow broadly shared prosperity. Both have a place, but as usual, true believers in each political camp strive for one-size-fits-all policy monoply.
Interesting. If Keynes provides economics of crisis, then it reinforces itself when misapplied.
Keynesian economic policy sounds an awful lot like folk Keynesianism. Can we argue they are one and the same? I think of it as “bizarro economics.” A world where saving is bad, any spending is good, purposeless Government spending has a magical multiplier effect, and we should wish to discourage innovation because it reduces the need for work, and work is good.
So Keynesians pull what Scott Alexander calls a Motte and Bailey: http://slatestarcodex.com/2014/07/07/social-justice-and-words-words-words/
My crazy idea is that the whole savings = investment thing is wrong. Savings in the real economy is inventory. There is consumption with the purpose of improving future production and there is normal consumption. That is why you can have a cell phone be consumption or investment depending on the use of the phone. Since money creation is now endogenous, loans just merely determine who gets first use of consumption and for what purpose based on bankers decisions.
Several suggestions:
– The single most important piece of evidence that destroys Folk Keynesianism is the empirical evidence that the multiplier is less than 1.
-A fundamental problem of interpreting aggregate macro data is that you are implicitly measuring the average when one of the top 5 most important econ 101 principles is that rational decision-making and price setting happens at the margin.
-Both folk and academic Keynesianism have difficulty dealing with the policy implications of the reality that price stickiness is a function of time. More time –> more price flexibility. The practical problem with a lot of Keynesian interventions is that the process of identifying a case of “market failure” arising from price stickiness, coming up with a solution, and then implementing the solution takes a long time. Even a theoretically valid (valid in keynesworld that is) intervention will in practice come into effect long after the market has adjusted, rendering the intervention a shock pushing the market away from rather than towards an equilibrium. For a specific example, think of how the NBER usually doesn’t even declare a recession as existing until it has been ongoing for 5-6 months.
– Both Folk and Academic Keynesians have difficulty coming to grips with the importance of expectations, and really have difficulty with the limited ability of policymakers to affect expectations.
– Folk Keynesians confuse income and wealth
– Folk Keynesians confuse income with cash flow (or to put it another way, they confuse value creation with velocity of money)
Ignore the problem so no solution is necessary. Aggregates are important though, the most important concern of people in fact, so dismissing them just dismisses any utility.