Tuesday, June 14, 2011

"There is no spoon"

Yesterday I reviewed one of Noah's posts. He linked to one by Mark Thoma. Today I'm looking at Thoma's post.

Thoma asks: Will the Economy Return To the Old Normal?

I don't know if that is "passive voice" or what, but it sure leaves out any participation by economists, don't you think?

Thoma writes:

This is a debate, in part, about whether the economy returns to trend after a shock, i.e. whether shocks are permanent or temporary. It is also a debate about the nature of the trend itself...

Oh... See? This is what economists do. Thoma is asking whether there is some metaphysical driving force that makes the economy always "return to trend." As if what happens to the economy has nothing to do with policy. It's a crock.

The nature of the trend itself?? There is no trend. The trend is an observation. It is an observation of what happened up to now. The trend is a figment, a conceptual tool. It's not in the economy. It's in the mind.

...the nature of the trend itself, i.e. whether the trend rate of economic growth is a smooth process that can be approximated by a trend line (with demand shocks responsible for most of the variation around the trend), or if the trend is a variable series subject to both permanent and temporary shocks (so that a substantial part of the variation in output over time comes from the trend itself...

Reject this.

If "the trend rate of economic growth is a smooth process" then we don't have to do anything by smile, and the economy will recover.

If "the trend is a variable series subject to both permanent and temporary shocks" then there is nothing that can be done to help the economy recover.

Nothing that can be done, or nothing that need be done. These are our only options?

I reject this.

Thoma quotes Mankiw:

According to the conventional view of the business cycle, fluctuations in output represent temporary deviations from trend.

The business cycle is the result of human interaction in a world of finite resources, or what economists call scarcity. Human nature is fixed and unchanging, but a sloppy fit. The position human nature takes, the "fit" of it, varies over the cycle. Sometimes we are more political, sometimes more social, sometimes more religious, sometimes more militant. Why we act as we do is not a question for economics. But we tend to act in certain ways at certain points on the cycle. That is what makes the pattern a cycle.

There are many cycles, or many interacting and overlapping trends. If we think of these cycles as waves, and it we think of sound as waves, we may compare them. Waves can be added together, or subtracted. More accurately perhaps, the effects of overlapping waves combine to produce a unified effect.

The longest business cycle I know of is the cycle of civilization, which is driven by the concentration and distribution of wealth.

Now, how does one look at the cycle of civilization and reduce it to temporary deviations from a trend of perpetual increase? Only near the economic peak of the cycle could such a concept even exist. Only after the peak would perpetual increase be such a pressing concern. Only in the decline would it matter.

Almost fell for it.

Thoma presents two graphs near the end of his post. One shows GDP varying about its long-term trend. The other, the "plucking model", shows GDP running at or near its long-term trend except when some shock or "pluck" pulls it down from trend.

Interesting. Indeed, that's why I almost fell for it. The plucking model shows GDP (after the pluck) returning to trend. The other model shows GDP always returning to trend.

But there is no trend.

There is no trend.


Jazzbumpa said...

Not much time, so I'll be brief. I disagree emphatically with everything you've said here.

Furthermore, you're dithering.

Your "Nothing that can be done, or nothing that need be done" pseudo-conclusion is baseless. Plus, you aren't offering anything.

A trend is not an observation, it is a thing that is observed. A falling tree makes noise, whether you are there to hear it or not.

Trends are not infinite. When a trend is broken, it is replaced by a new trend. That does not mean the old trend is invalidated. I just means it's done. And yes, there are trends at different levels. See my post on the dollar index.

The U.S. economy is in big trouble because of trends - slowing growth, rising wealth disparity, disappearance of the middle class, dominance of the financial sector.

"Only near the economic peak of the cycle could such a concept even exist."

Nonsense. I think you need to have a look at Elliot Wave theory.


Alas, not so brief after all.


Jazzbumpa said...

My post looks really cranky. Sorry for that. I woke up with a headache, and that will be my excuse for everything today.


The Arthurian said...

Well, your 2nd entry had me laughing, so I'm not even gonna use the word curmudgeon!!! Hope the headache went away.

The "pseudo-conclusion" is Thoma's, unless I misunderstand you.

I was a little "iffy" myself about this post. How can I deny that a trend applies to the future and at the same time speak of a cycle of civilization? Such a cycle is itself a trend!

Trends are memories. They exist only in the past.

I do not object to looking at trends of the past, and I find your glimpses of a trend-of-trends most informative. But I object strongly to basing one's views of the future on the assumption of a straight-line trend reaching back to the 1800s and continuing into the future, which is what Thoma does in his post.

Regarding GDP, you say we seem to have established a new trend line which is lower than the old trend line, which is lower than the even older trend line.

Regarding profits, you see an increasing slope over time.

So we have a series of decreasing GDP trend lines, and a series of increasing corporate profit trends which, evidently, didn't do all that much to boost GDP.

But Jazz, you are not claiming there is a single, straight trend-line for GDP like Thoma.

The general pattern of your GDP trend lines is like consecutive tangents coming off a circle. That circle is the cycle of civilization.

Jazzbumpa said...

I thought the psuedo-conclusion was yours. I didn't read Thoma as saying that.

The whole cycle of civilization thing is Elliot wave, a la Prechter.

The highest level trend (with a period of millennia) is up, but within it are lower level trends with periods down to fractions of seconds.

The next-highst level oscillations define the trend channel, and you can do that sort of thing at every level of trend. That's what I was alluding to here, though I only hinted at it.

I didn't notice how highly similar the various graphs were until I sliced them up for the post. That is the nature of fractals. They repeat (similarly, not identically) at every level of trend. Slice the first graph at about Oct, '10, and you see the same basic pattern yet again.

The subtle point of that post is the striking similarity of all the graphs.

I agree that the straight-line interpretation of GDP is incorrect. I don't understand how a careful observer could ever think otherwise.


Jazzbumpa said...

BTW, I still had the headache when I woke up. Made an unscheduled visit to the chirocracker, and he worked on everything from the shoulder blades up.

Wow, what a difference!


LH said...

JB I don't see any fractals in charts of world population growth since 30,000BC.

If it was a fractal process then the fractal nature would be observable regardless of how far one 'zooms out'. If that property doesn't hold then its not a fractal process.

Regarding trends, they are found by people who are looking for a trend.

Scientists don't really do trends, they have theories which predict the exact future path, or fail to.

Jazzbumpa said...

LH -

The 30,000 BC census is known to be highly unreliable, and the 20,000 BC census was lost in a volcano incident. The 7000 BC census was lost in a flood. The 1890 U.S census was lost in a fire. It was all very sad. You're not going to find fractals in a data set made up of wild-ass guesses.

But I will tell you that, Euro-centrically speaking, there were serious population declines in the first millennium AD, and again in the mid to late 14th century. The 1930's and 40's were not good time for population growth, but the 50's were.

Your description of fractals is spot on, though. My dollar index charts are good examples. Stock market charts are great examples.

Take a look at my trend channels, and tell me they are imaginary.

You have science exactly wrong. It is probabilistic, not deterministic.
Predicting future paths is the purview of mystics and charlatans.