Thursday, March 28, 2024

Overwhelming evidence

Milton Friedman didn't like too much money. But he also didn't like too little money. In chapter 2 of Money Mischief he wrote:

There is strong evidence that a monetary crisis involving a substantial decline in the quantity of money is a necessary and sufficient condition for a major depression.

Insufficient money can cause a depression. If the "monetary base" grows too slowly it can cause a major downturn. That happened twice in the past hundred years:

Graph #1: Growth Rate of the Monetary Base
From mine of 10 Feb 2021
The downtrend before the Great Recession runs from 2001 to 2008.


Graph #2: Federal Debt 1970-2023 and the 2001-2023 Exponential Trend
From mine of 7 March 2024
The "below trend" data before the Great Recession runs from 2004 (or before) to 2008-09.


Graph #3: The Quantity of Transaction money per Dollar's Worth of Output
The low before the Great Recession runs from 2004 to 2009.


Fiscal and monetary policy cooperated, creating a substantial decline in the quantity of transaction money, from the record low of 2000-01 to a level at which our economy could no longer function.

 "... substantial decline in the quantity of money is a necessary and sufficient condition for a major depression."

The evidence is overwhelming.


As an afterthought, I quote from Robert Lucas's "Ptize Lecture" of 1995:

Evidence on trade-offs is also marshalled, though in a very different way, in Friedman and Schwartz’s (1963) monograph A Monetary History of the United States. These authors show that every major depression in the United States over the period 1867-1960 was associated with a large contraction in the money supply, and that every large contraction was associated with a depression. These observations are correlations of a sort, too, but they gain force from the size of the largest contractions. In a period like the post-World War II years in the United States, real output fluctuations are modest enough to be attributable, possibly, to real sources. There is no need to appeal to money shocks to account for these movements. But an event like the Great Depression of 1929-1933 is far beyond anything that can be attributed to shocks to tastes and technology. One needs some other possibilities. Monetary contractions are attractive as the key shocks in the 1929-1933 years, and in other severe depressions, because there do not seem to be any other candidates.

Sargent (1986) also examines large, sudden reductions in rates of money growth (though not reductions in the levels of money stocks). In his case, these are the monetary and fiscal reforms that ended four of the post-World War I European hyperinflations. These dramatic reductions in growth rates of the money supply dwarf anything in Friedman and Schwartz or in the post-war data used by McCandless and Weber. Yet as Sargent shows they were not associated with output reductions that were large by historical standards, or possibly by any depressions at all. Sargent goes on to demonstrate the likelihood that these reductions in money growth rates were well anticipated by the people they affected and, because of visible and suitable fiscal reforms, were expected by them to be sustained.

... The observation that money changes induce output changes in the same direction receives confirmation in some data sets, but is hard to see in others. Large scale reductions in money growth can be associated with large scale depressions or, if carried out in the form of a credible reform, with no depression at all.

I'm not one to leap to Milton Friedman's defense. But in this case I have to. Robert Lucas presents a version of Friedman and Schwartz's view that is a satisfying equivalent of the Friedman quote in the opening of this post. But then Lucas seems to turn around and challenge Friedman's view:

These observations are correlations of a sort, too, but they gain force from the size of the largest contractions.

Lucas could have said they are correlations and they gain from the size -- thus adding the two positives together -- but he chose to say they are correlations but they gain from the size, as if the size subtracts from the significance of the correlations.

And they are only correlations of a sort.

Then, rather than appreciating Friedman's willingness to evaluate the largest contractions, Lucas minimizes Friedman's effort:

[A]n event like the Great Depression of 1929-1933 is far beyond anything that can be attributed to shocks to tastes and technology. One needs some other possibilities.

Lucas even takes advantage of an opportunity to criticize Friedman for using large contractions that were not large enough to disprove Friedman's view:

Sargent (1986) [studied] dramatic reductions in growth rates of the money supply [that] dwarf anything in Friedman and Schwartz or in the post-war data used by McCandless and Weber. Yet as Sargent shows they were not associated with output reductions that were large by historical standards, or possibly by any depressions at all.

Lucas provides the weakest support for Friedman that I have ever seen:

Monetary contractions are attractive as the key shocks in the 1929-1933 years, and in other severe depressions, because there do not seem to be any other candidates.

That's not support. Lucas is trying to undermine Friedman's view. But for crying out loud, what Friedman said (at least in Money Mischief) was

There is strong evidence that a monetary crisis involving a substantial decline in the quantity of money is a necessary and sufficient condition for a major depression.

Friedman did not say a substantial decline in the Q-of-M guarantees that a depression will follow. Lucas was being ridiculous, saying what he said. Worse than ridiculous, and I'm being polite.

Lucas concludes his thought by suggesting that good policy can prevent the depression, as if this somehow proves Friedman wrong. But surely, if policy was that good, it would not have created the substantial decline of money in the first place! And dumb luck doesn't count as good policy. It counts as luck.

Given what Lucas said, I must emphasize the point that economic policy is all-important. Lucas himself seems to miss that point, preferring to emphasize that Friedman was wrong. As if any fool could make good policy.

It takes a special kind of fool to make good policy.

Thursday, March 14, 2024

Employee Compensation as a Percent of GDP

Two questions:

1. Why does no one say the Biden economy is good?  One reason: As a percent of GDP, employee compensation has never been lower:


2. Is that Biden's fault?  No. Employee compensation has been all downhill for half a century. That's not Biden's fault.

 

Third question: What is to be done?  The first three steps in solving a problem are

  1. Correctly understand the problem.
  2. Correctly understand the cause of the problem.
  3. Correctly understand the cause of that cause.

Hint: Low employee compensation is not the problem. It's a result.

Tuesday, March 12, 2024

A Political Note

For those on CNN and MSNBC who repeatedly wonder why no one appreciates how great the Biden economy is...

and setting aside my ordinary response BECAUSE THE ECONOMY ISN'T GREAT...

there is this statement from J. M. Keynes, quoted by Milton Friedman in chapter 8 of Money Mischief:


There is no subtler, no surer means of overturning the existing basis of society than to debauch the currency. The process engages all the hidden forces of economic law on the side of destruction ...


Keynes said it, and Friedman quoted it. That makes the statement twice as valuable, no matter which side you're on.

Sunday, March 10, 2024

Don't brag about our economic growth, make it worth bragging about

My wife insists on watching CNN and MSNBC. I find the shows tiresome because their arguments are often weak. For example, a Biden supporter comes on the show and says the U.S. economy is doing better than any other country. I don't believe it. Last I checked, China's economic growth rate had fallen from 10% to 6%. That may be slow for China, but 6% annual growth is about twice the U.S. rate. 

But even if it is true, it is not a strong argument. A strong argument would be to compare recent U.S. economic growth to our own past growth, and point out how well we are doing now. But we cannot do this because, by that standard, we are not doing well.

To see how good our economy is now, I compare our current performance to the high points of our past performance. The red line on the graph below shows the trend of those high points:

Annual US Real GDP Growth Rate and the Trend of Peaks  1948-2023

I connected the 1955 peak (7.1% growth) to the 2018 peak (3.0% growth), then extended the red line out to 2023 and back to 1948. Three above-trend peaks are easy to spot:

  • 1950-1951: Probably due to Korean war spending
  • 1984: Reagan's "Morning in America"
  • 2021: Real GDP growth in the year after the covid shutdown

In addition, U.S. economic performance surpasses the trend of peaks briefly in the mid-1960s, and again in the latter 1990s. Other than that, all the peaks are at or below the trend line. Even the 2023 data (which is preliminary and will likely change) is below the trend of peaks. If I was Biden, I would be embarrassed that my people are bragging about how good the Biden economy is. His people are probably losing him votes. 

By the way, the trend line itself shows a downtrend nearly as long-lived as Biden. This downtrend is clearly not Biden's fault. Maybe he should say as much, and add that he wants to turn the trend upward in the next four years. 

That trend can't go much lower. It fell five percentage points in less than 80 years, and is now near 2.5%. If the trend continues another 40 years, it will be approaching zero growth. Zero GDP growth means the only way to get a raise is to take income from someone else. That's not how things are done among a civil people.

There are better ways to run an economy, other than expanding credit and accumulating private-sector debt until the economy wants to die. "Perhaps the U.S. needs another, more modern Keynes..."



I checked: They dunno 2023 yet. But for 2022, China's real GDP grew 2.99%. So, 3 percent in 2022. The U.S. GDP grew 1.9% in 2022, and 2.5% in 2023. In 2021, the bounce-back year after the covid-shutdown recession, U.S. GDP grew 5.8% while China's economy grew more than 8%. 

"Which country is doing better?" is the not the question to ask. Better would be "How do we fix this?"

 

To my good eye, comparing US growth to that of other nations makes for a weak argument. Comparing our current economic growth to that of our own past makes a much stronger argument. But Biden can't do that until he makes the trend turn upward. And the trend will not show growth improving until we start making better economic policy decisions.

Don't brag about our economic growth, make it worth bragging about. All we need is the right plan.

Thursday, March 7, 2024

Federal Debt Held by the Public, 1970:Q1 - 2023:Q3

So I happened to look at Federal Debt Held by the Public at FRED:

Graph #1: https://fred.stlouisfed.org/series/FYGFDPUN

It curves up a little from 1970 to 1990. It curves down a bit then in the 1990s, the Clinton years, as you may remember. Then, since 2001, it goes up for a short decade, and UP for a long decade, and then SCREAMING UP to end-of-data.

You may have heard some of that screaming in Congress in the past couple years.

Hey, I'm writing this because the federal debt in the current millennium is eye-catching (to say the least). My first reaction, looking at the years since 2001, was That's gotta be exponential. So I had FRED show the "natural log" values:

Graph #2: https://fred.stlouisfed.org/graph/?g=1hHJY

Now it doesn't go up and UP and UP. Shown as log values, it's almost a straight line going up from the low point in 2001, to end-of-data in 2023.

When log values show a straight line it means the rate of growth is constant. That's "exponential" by definition -- a constant rate of growth, and a constant doubling time. On this graph, since 2001, it looks pretty straight. It suggests that the growth rate of the federal debt has been nearly constant since 2001 -- something that is hard to see on the first graph. There is a bit more curve in the Obama years than before or after. But there was more fixing of the economy to do at that time, what with the financial crisis and all.

After 2020, the slope of the line looks about the same as the slope from 2001 to 2008. So, the rate of federal debt growth might be about the same in the Biden years as it was in the George W Bush years.

Sure, there is more debt after 2020 than there was by 2008. That's what happens when debt grows. You get more of it.


I wanted to put an exponential curve on that first graph, to see how it fits the plotted data. FRED doesn't let me do that. So I brought the data into Excel and did it there:

 
Graph #3: The red line is an exponential curve based on
data for the 2001:Q2-to-2023:Q3 period

The red line shows the "trend" since 2001; data from before 2001 was not used to create the red line. It shows the 2001-2023 trend of "held by the public" federal debt. If the red line matches the blue in the 1970s it is by chance, or due to the constancy of human nature maybe -- or my eye is off -- but it is not due the arithmetic making them match.

The lows of 2005-2008 (which probably contributed to starting the 2008 financial crisis), of 2016-2020, and from 2022 to end-of-data, together offset the high of 2009-2015 (a high which probably prevented a collapse of the banking system). (Dates approximate.)

And it is interesting to see that the very large debt increase of 2020 did no more than bring the federal debt back up to the trend.

Again, these dates are approximate: From 1975 to 1995, debt held by the public rose more and more above the trend. In other words, all through the Reagan years, and the first Bush, and the Clinton years before the 'New Economy" of the latter 1990s, the growth of federal debt ran increasingly above trend.

The next graph uses the same data as graph #3, but shows the "natural log" values of that data (as with graph #2):

Graph #4: The red line is an exponential curve gone straight
because the graph uses log values.

Based on what I see in these graphs, it still holds good that:

  • The below-trend federal debt growth of 2005-2008 probably contributed to starting the 2008 financial crisis. There was also the low growth of M1 money in the same years, and the low growth of base money in those same years as well. As Milton Friedman points out in Chapter 2 of Money Mischief, "a substantial decline in the quantity of money is a necessary and sufficient condition for a major depression." We almost had one in 2008.
  • The high federal debt growth of 2009-2015 probably prevented a collapse of the banking system.
  • And, again, the very large, covid-related debt increase of 2020 did no more than bring the federal debt back up to the trend. Graph #1 clearly shows the size of that increase and #3 shows the return to trend.

The graphs also show that all through the Reagan years, and the first Bush, and the early Clinton years (before the 'New Economy" of the latter 1990s), the growth of publicly held federal debt ran increasingly above the trend. People often talk about how the federal debt "exploded" after 1980 -- how it grew so much faster than GDP. Well, debt did grow faster for some years  after 1980 than before, but most of the difference we see on the graph is because the rate of inflation was coming down after 1980.

At the lower rates of inflation, nominal GDP increased at a lower rate, so the rapid growth of the debt was easy to see. Debt did grow faster after 1980 than before, but most of that was due to inflation coming down -- to disinflation -- not to accelerated debt growth. Anyway, as the graphs show, above-trend growth of debt started in the mid-1970s. So some, but not all of the rapid debt growth was due to Reagan spending like a madman. I know nobody wants to hear it, but that's how it is.


One more graph:

Graph #5: Five-Year CAGR Growth Rates
Each plotted point shows the end of a five-year period
and the compound annual growth rate for the period

Now I see that federal debt growth was all over the place. Some things are recognizable:

  • "Morning in America" around the time of the 1984 peak;
  • The federal budget falling toward balance, 1995-2001;
  • The response to the 2008 financial crisis, 2008-2013.

I find it interesting, too, that federal debt growth seems drawn to the 10% level: briefly, after 1975; momentarily after 1980; for the five years from 1990 to 1995; and since 2020. It is as if some policymaker said "I dunno, let's see what happens at ten percent." It is most strange to see such things in our economy. It doesn't fill me with confidence.

Nor does my interpretation of logged data fill me with confidence... So much for the growth rate of the federal debt being nearly constant since 2001.

Tuesday, March 5, 2024

"Is U.S. Economic Growth Over? Faltering Innovation Confronts the Six Headwinds" (2012)

Is U.S. Economic Growth Over? Faltering Innovation Confronts the Six Headwinds by Robert J. Gordon. 2012

http://www.nber.org/papers/w18315
https://www.nber.org/system/files/working_papers/w18315/w18315.pdf


Among the paper's "basic points":

Future growth in real GDP per capita will be slower than in any extended period since the late 19th century, and growth in real consumption per capita for the bottom 99 percent of the income distribution will be even slower than that.

But only if we fail to solve the problem.

Monday, March 4, 2024

"What Do We Know About Economic Growth? Or, Why Don't we Know Very Much?" (2001)

What Do We Know About Economic Growth? Or, Why Don't we Know Very Much? by Charles Kenny and David Williams, 2001.

https://faculty.nps.edu/relooney/KennyGrowthSurvey.pdf


From the Conclusion:

There are, we think, a number of conclusions and implications which follow from the analysis we have presented here. First, a review of the available evidence suggests that the current state of understanding about the causes of economic growth is fairly poor... What we are arguing is that we are in a weak position to explain why some countries have experienced economic growth and others not.

If even the views of experts offer a fairly poor understanding of the causes of economic growth, then what would it hurt to consider the views of a hobbyist like me?