Sunday, December 2, 2018

RE: Milton Friedman's 1968 critique of "the new economics"

Excerpts from and commentary on “Has the New Economics Failed? An Interview with Milton Friedman.” (PDF, 9 pages)

The second question:
Q: Assuming that the New Economics is nothing very new, how would you evaluate its accomplishments between 1961 and today?

A: It is very hard to say that the New Economics has accomplished very much of anything, except getting a great deal of publicity. I have always been amused by the fact that the New Economists emphasize fine-tuning—or changing the direction of the economy within a short span of time. But if you look at what actually happened, it certainly did not occur over a short span of time...
My first hint that by fine-tuning we mean "changing the direction of the economy within a short span of time" occurs in the middle of a sentence about something else. It is a definition-by-accident, in a way. Does it make sense? On my first read I said Yeah, okay. I accepted Friedman's reduction of the meaning of the term "fine tuning" to "in a short span of time".

But it's ridiculous. Nothing in the economy happens "in a short span of time". Friedman uses a kind of straw man here, a definition that can easily be shot down -- if you haven't already accepted it. It is not really a definition of the term, at all.

Worse, he introduces this definition mid-sentence, as if by accident. (Oops, did I forget to define the term?) The reader is trying to understand what Friedman is saying, and is therefore led to accept Friedman's definition rather than to evaluate it. If Friedman had defined the term separately first, we would have been led to evaluate his definition, and accept it or not.

The language Friedman uses, burying the definition mid-sentence, suggests that he does not want us to spend much time evaluating the definition.


The third question:
Q: To get around such delays, would you then favor giving the President the right to raise and lower taxes within a narrow limit subject to a Congressional veto?

A: No, I am opposed to it, because I’m not in favor of fine-tuning in that sense. I don’t believe we know enough to be able to do it. I don’t object to the aim; it would be very desirable to have an instrument that would enable us to keep the economy on an even keel. But I don’t think there is any. We do not know enough about either fiscal or monetary policy to enable us to make delicate adjustments from time to time that are going to offset other forces and make for greater stability in the economy.
This question immediately follows Friedman's answer in which he offers his accidental definition of "fine tuning". Notice that the interviewer accepts Friedman's definition and builds on it by asking about getting around such delays. The interviewer doesn't pause and wonder Is this a good definition? He runs with it.

I can only imagine that Friedman was hoping for similar acceptance of the definition from his readers: acceptance without evaluation.

By contrast, Investopedia's definition of fine tuning has nothing to do with delays or short spans of time:
Fine tuning refers to the process of making small modifications to improve or optimize an outcome. Generally, fine tuning seeks to increase the effectiveness or efficiency of a process or function. Fine tuning can be accomplished through a variety of ways, the methodology of which depends on the process being optimized.
Surely, Milton Friedman would not say he is opposed to "making small modifications to improve or optimize an outcome". Or maybe he would, I don't know. But it must be easier for people to accept a sweeping rejection of delay than a sweeping rejection of improvement. Friedman's definition of "fine tuning" makes his argument against the new economics easier to accept.

It makes the argument easier to accept.

Presenting the definition as if by accident makes the definition easier to accept. And then, accepting the definition makes Friedman's argument easier to accept. But it's all bullshit.

More accurately, it may all be bullshit; it depends upon our evaluation of his accidental definition of "fine tuning".

//

But this is good:
We do not know enough about either fiscal or monetary policy to enable us to make delicate adjustments from time to time that are going to offset other forces and make for greater stability in the economy.

That's a beautiful statement. An admission that economists don't know enough to be able to optimize economic outcomes. You don't see an admission like that very often. Not very often.


Question 4:
Q: What about rough adjustments? While that famous tax cut was very late, it did seem to work.

A: Did it? What effect did it have? Lets go back and look at the record. There was a slowing off in the economy in late 1962 and early 1963, and there was a picking up of steam in the economy in late 1963 and early 1964— all of which happened prior to the tax cut. That steam kept on picking up at the same rate after the tax cut. Now some people will say that the economy moved ahead in late 1963 in anticipation of the tax cut. But then they also say that the actual occurrence of the tax cut fostered the growth afterwards. You can’t count it twice; you can’t have it over and over again.
"You can’t count it twice", Friedman says.

What?? There is no "twice". The economy started picking up steam in late '63, and continued picking up "at the same rate" (Friedman says) after the 1964 tax cut. From this, Friedman concludes that people double-count the effect of the tax cut when they say both that
  • the economy moved ahead in late 1963 in anticipation of the tax cut, and that
  • the actual occurrence of the tax cut fostered the growth afterwards.
Double-counting? Really? No, not really. We get a boost before the fact, from expectations. And then we get a boost after the fact, if the expectations pan out.

Friedman's "You can’t count it twice" argument looks like serious bullshit to me.

//

As Friedman describes it, "late 1963 and early 1964" was "prior to the tax cut." Going along with the semiannual rhythm Friedman uses, I assumed from his remarks that the tax cut came in "late" 1964 (and I wrote it that way). In proofread I checked this, and of course it was wrong. I found “The Kennedy Tax Cut of 1964”, section 27.2 from the book Theory and Applications of Economics, which says
As is usually the case when a major fiscal policy action is under consideration, there was a lengthy time lag between the initiation of the policy and its implementation. Even though the tax cut was proposed in 1962, President Kennedy never saw it put into effect. He was assassinated in November 1963; the tax cut for individual households and corporations was not enacted until early 1964.
Early 1964. Wikipedia confirms:
The United States Revenue Act of 1964 (Pub.L. 88–272), also known as the Tax Reduction Act, was a bipartisan tax cut bill signed by President Lyndon Johnson on February 26, 1964.
February 26. Early 1964. Maybe the tax changes of early 1964 didn't have an effect until later; maybe that's what Friedman meant. That's fine. After the changes took effect (whenever that was), you've got growth from the tax cut (according to Friedman) and before the changes took effect you've got growth from "expectations" of the tax cut. It's not double-counting to count both effects, if both occurred.

To call it double-counting and dismiss it, is bullshit.


Two paragraphs from that "Kennedy Tax Cut of 1964" chapter give a good feel for what was going on in those years:
Every president has a group of economists, known as the Council of Economic Advisors (CEA; http://www.whitehouse.gov/cea), that provides advice on economics and economic policy. The list of members and staff of the 1961 CEA reads today like a “who’s who” of economics. James Tobin and Robert Solow were prominent members of the economics team; both went on to win Nobel Prizes in Economics. The chairman of the CEA was Walter Heller, an economist known for a wide variety of contributions on the conduct of macroeconomic policy.

The economists in the Kennedy administration observed that there had been three recessions in the two Eisenhower administrations (1952–1960): one from 1953 to 1954 after the Korean War, one from 1957 to 1958, and one in 1960. You can see these in Figure 27.4 "Real GDP in the 1950s". The CEA members and staff thought that more aggressive fiscal and monetary policies could be used to keep the economy more stable and prevent such recessions. Their goal of moderating fluctuations in the economy was based on the framework of the basic aggregate expenditure model, which had been developed in the aftermath of the Great Depression, augmented by some developments in economic thinking from the 1940s and 1950s. Based on that analysis, they believed that fiscal and monetary policies could be used to control aggregate spending and hence real GDP.

That's what they were doing with "the new economics" of the 1960s. It doesn't sound to me at all like their focus was as Friedman describes: "changing the direction of the economy within a short span of time". Not at all.

Here's how Time magazine described the goals of the new economics in a 1965 article:

In Washington the men who formulate the nation's economic policies have used Keynesian principles not only to avoid the violent cycles of prewar days but to produce a phenomenal economic growth and to achieve remarkably stable prices. In 1965 they skillfully applied Keynes's ideas—together with a number of their own invention—to lift the nation through the fifth, and best, consecutive year of the most sizable, prolonged and widely distributed prosperity in history.
...
In the early 1930s, Keynes cried out that the only way to revive aggregate demand was for the government to cut taxes, reduce interest rates, spend heavily—and deficits be damned.
...
Keynesianism made its biggest breakthrough under John Kennedy, who, as Arthur Schlesinger reports in A Thousand Days, "was unquestionably the first Keynesian President." Kennedy's economists, led by Chief Economic Adviser Walter Heller, presided over the birth of the New Economics as a practical policy and set out to add a new dimension to Keynesianism. They began to use Keynes's theories as a basis not only for correcting the 1960 recession, which prematurely arrived only two years after the 1957-58 recession, but also to spur an expanding economy to still faster growth.

It surely does not sound like the focus of "the new economics" was to change the direction of the economy in a short span of time. Change the direction? Yes. They wanted to improve things. But Friedman does not focus on their efforts to improve things. He focuses on the "short span of time" and suggests that the New Economics was a failure because they couldn't get things done quickly.

Friedman's focus is flawed.

No comments: