We need a better plan

Two points to keep in mind:

• If our economy went into decline after 1973, then 2023 marks half a century of decline.
• Long-term economic decline is indistinguishable from the decline of civilization.
  
  

 
In Why Raising Interest Rates to Fight Off Energy Inflation is Counterproductive by Hielke Van Doorslaer we read that hiking interest rates, our go-to method of fighting inflation,

puts monetary policy at cross-purposes with other policy priorities (such as investing in renewables and energy-efficiency) and risks further entrenching years of public and private underinvestment. In this way prohibitive interest rates will further exacerbate the trend of secular stagnation (defined by low rates of growth, productivity, and investment) that has plagued advanced economies for at least a decade (with some even dating the onset of the declining trend back to the 1970’s).

I agree that our existing anti-inflation policy is "at cross purposes with other policy priorities". As new anti-inflation policy, I have suggested pruning back the policies that promote credit use, and creating policies that induce people and businesses to accelerate their repayment of debt by offering tax benefits for doing so. (But this is not today's topic.)

At the end of the excerpt Van Doorslaer points out that slow economic growth "has plagued advanced economies for at least a decade" and adds

with some even dating the onset of the declining trend back to the 1970’s.

This is today's topic: long-term slowing of economic growth.

A footnote attached to the Van Doorslaer excerpt links to Jack Copley's "Decarbonizing the downturn: Addressing climate change in an age of stagnation". Copley points out that "GDP growth and investment have trended downward globally since at least the 1970s".

Since at least the 1970s. Since the 1970s or before. Yep. I wrote for the blog one time since the 1960s but then I went back and changed it because people might accept the 1970s as start-of-decline, but it sounds crazy to say our long decline began in the 1960s. So ya knock off the sharp edges, and try to look sane. But it amazes me that so many people, even economists, regularly fail to see the long decline. People too often see what they've been told they see, instead of looking with their own eyes.

Copley also says

Brenner has labelled the years 1965–1973 a transition period from ‘long boom to long downturn’ (2006: 37).

And if I read Steve Keen right, Minsky put the end of the Golden Age at 1966.

Scott Sumner has said

I am not denying that growth in US living standards slowed after 1973, rather I am arguing that it would have slowed more had we not reformed our economy.

Yeah, yeah. If the economy slowed after 1973 despite the reforms Sumner mentions, then those reforms were not the right solution. 

If you don't know how and when the problem started, what the cause really was and what the problem really is, all you can do is guess at the solution.

America has had no doubt for half a century that we must balance the federal budget. But America refuses to acknowledge that the only way to balance the federal budget is by reducing private sector debt so that the economy can grow and people can afford to live. Once we get the economy growing we can easily balance the federal budget: Easily. It will be easy. And now you're questioning my sanity again.

We've been trying to balance the budget the wrong way for 50 years. We have not restored vigor to our economy. We have not even once actually balanced the federal budget: The last time the federal debt fell was in 1969. We need a better plan.

The economic decline of a lifetime

When I google long-term economic decline, most of the results focus on recession. A few focus on depression. After three sittings, I found one search result that actually considers long-term decline:

"Recessions are difficult, but stagnant growth could prove more challenging, Stanford economist warns", by Melissa De Witte, dated 7 December 2022. Almost current. From Stanford News. The article is an interview with John Cochrane. I know of him from The Grumpy Economist.

De Witte's opening:

While recessions are painful, they are only temporary interruptions to the economy, says John Cochrane, an economist at Stanford’s Hoover Institution, arguing that people should be paying more attention to long-term economic growth, which in the U.S. is currently stagnating.

Yeah. But I wouldn't say "stagnating". I'd say declining. And not just "currently".

But wait, this is a good observation:

Rather than focus on quarterly changes to growth rate, which is how recessions are currently gauged, the long-run growth of the economy matters more.

Most. Definitely. Yes. Long-run economic decline is like a weak, permanent recession that slowly gets worse and only gets worse.

 

Here's something. Cochrane says:

Currently, stagnant growth is our big problem. Long-run growth of the economy matters much more than year-to-year growth rates. Recessions are painful interruptions, but we should be paying much more attention to long-run growth.

"Currently" again? What does he mean to say? Only just now? Or for quite some time? For how long? Dunno, dunno, dunno. Plus, Cochrane needs more emphasis on "big problem".

But hey: If long-run growth is what matters, then maybe long-run growth is what we should look at. Let me start with a quick look at three graphs. We can dwell on the third one.

Graph #1: Up, Up, and Away! (Or so it appears)

Graph #1 shows Real GDP for the US in dollars, millions of dollars, from 1790 to 2021. Data from MeasuringWorth. "Real" means inflation has been taken out of the numbers, so we see the growth of output without the increase of prices.

Graph #2: The same data. A different glance.

Graph #2 shows "percent change" from the year before, for each year. Cochrane calls it the "year-to-year growth rates". The plotted line is jiggy as can be. And I don't see any definite pattern in it, so I don't know what the graph is telling me.

Graph #3: Again, the same data, but 20-year averages

Now we are looking at long-run growth. This graph shows long-term (20-year) average growth rates. Each point on the graph shows the average for the 20-year period that ends at that point.

A pattern begins to emerge in the third graph: 

• Increasing growth from 1820 to 1863
• Decreasing growth from 1890 to 1933
• Increasing growth from 1933 to 1954
• Decreasing growth from 1954 to 2020

Consider the decrease that ends in 1933. I am reminded of Milton Friedman in Free to Choose, saying "the economy hit bottom in 1933." Graph #3 shows long-term growth hitting bottom the same year, 1933. Coincidence, you think? Not entirely.

The graph shows sharp increase after 1933, rising to a peak in 1954. Recovery from the Great Depression feeds into that sharp increase. So does the second World War.

After the 1954 peak the graph shows persistent decline, from more than 6% growth to less than 2%. What was it Cochrane said?

Long-run growth of the economy matters much more than year-to-year growth rates. Recessions are painful interruptions, but we should be paying much more attention to long-run growth.

Cochrane said, and I can only agree that

Currently, stagnant growth is our big problem. 

But the problem is bigger than we think. If nobody's happy with our economy currently, maybe it is because long-run growth has been in decline now for seven decades.

Seven decades. It's a lifetime.

Why the long decline? Finance is the problem. Excessive finance. This graph is from Thomas Philippon:

Graph #4: from "Has the U.S. Finance Industry Become Less Efficient?"
by Thomas Philippon (2011)

Philippon's high points match up with the lows on Graph #3, and Philippon's low points correspond to the highs on Graph #3. I'm not comparing the years before 1890. But since 1890, Philippon's graph goes up, down, up while my graph goes down, up, down. Since 1890.

Finance goes up-down-up while Real GDP growth goes down-up-down. Economists might call that a negative correlation. You might think they would want to spend more time looking into excessive-finance-as-the-cause-of-slow-growth. Cochrane in the interview, for example, fails to mention finance as a possible cause of the stagnation.

Philippon describes his graph:

The cost of intermediation grows from 2% to 6% from 1870 to 1930. It shrinks to less than 4% in 1950, grows slowly to 5% in 1980, and then increases rapidly to almost 9% in 2010.

His turning points: 1930, 1950, 2010. My turning points: 1933, 1954, 2020. The big discrepancy, 2010 versus 2020, arises only because I'm using Philippon's ten-year-old graph. The other differences are rounding errors.

His graph shows two reversals of trend. My graph (again, since 1890) also shows two reversals of trend. There are only two reversals in 130+ years, and those reversals are almost simultaneous on the two graphs. That's gotta be significant!

I took Philippon's graph and erased the background, then "flipped" the graph to make the low points high and the high points low. Then I overlaid it on my Graph #3 and fitted his graph to mine to make the dates line up. Now we can look at the pattern since 1890 and see how the growth of finance compares to long-run economic growth:

Graph #5: Philippon's graph overlaid on my Graph #3
with dates aligned (note alignment at 1880 and 1980)

Remember: finance is flipped for figure five.

• From 1890 to the early 1930s, 20-year average RGDP growth shows decline. So does flipped finance.
• From the early 1930s to 1950 or so, 20-year average RGDP growth shows increase. So does flipped finance.
• From 1954 to end-of-data, 20-year average RGDP growth shows decline. So does flipped finance.
  

Flipping Philippon's graph makes it easy to see the similar pattern in the two graphs.

Philippon's original, un-flipped graph shows that the size of finance runs opposite the long-term growth of Real GDP. The more finance we have, the less economic growth we get. The less finance we have, the more economic growth we get. Since the late 1800s this is true. The graph shows it.

"Long-run growth of the economy matters much more than year-to-year growth rates," Cochrane says, and "we should be paying much more attention to long-run growth." Indeed it does, and yes we should.

Since 1890, as finance increases in size, Real GDP grows less. Why? Because finance is a cost, and more finance means more cost.

Hey -- We need finance for growth, yes. But we shouldn't be using finance for everything. We need just plain money -- money that doesn't cost money -- to maintain our existing economy at its existing level. 

We can use credit for growth. That's fine. But after the economy grows, the economy is bigger. So then, instead of the credit, we need more "money that doesn't cost money" to maintain the bigger economy. This solution, obvious as it may be, is not part of economic policy.

Under existing policy, old, well-used credit continues to exist as debt, public and private. The financial cost of debt interferes with the cost of living. As Vladimir Simkhovitch said of ancient Rome, things eventually get to the point where labor cannot support life. This is where we are today: Labor can no longer support life. So we live on credit, and our debt just grows and grows. This cannot end well. Policy must change.

Central bankers don't see anything wrong with the cost of finance, probably because the cost of finance is income for bankers. For the rest of us, the cost of finance is a cost. 

As long as economic policy continues to promote credit use and the accumulation of debt, we will continue to see the decline of long-run economic growth. 

 

The comparison of Finance to long-term Real GDP growth is not a good match before 1890. I did not research this discrepancy. But finance was small before 1890. Perhaps it was small enough, then, that finance was still beneficial, rather than harmful to growth. If so, the discrepancy is further evidence that excessive finance is harmful -- and that finance has been excessive since 1890.

Seeing growth slowing

At FRED, the annual version of Real GDP -- the one with inflation stripped out of the numbers so we can see a real (honest) measure of economic growth -- looks like this:

Graph #1: GDP with price changes removed, to show real growth

The blue line begins in 1929 and ends in 2022. It curves gradually upward over time, and growth appears to be good. But I think we know that growth is not good.

Graph #1 doesn't show the growth rate of GDP. It shows amounts of GDP. Bigger amounts are higher on the graph. But increasing amounts don't always mean better growth. It's like getting a 10-cent raise in your paycheck: It is an increase, yes, but you're not even keeping up with the cost of living.

There are two things: There's the data that you put on the graph, and there's the way the graph shows the data.

The GDP data we're using has inflation stripped away so that the numbers don't increase from prices going up. The numbers only increase from producing more stuff. You have to do that to see GDP growth. But to see or compare changes in growth, you have to do something else. This second thing is to tweak the graph to make the vertical scale a log scale.

Graph #1 shows GDP gradually curving up. The graph shows that we we are producing more and more. But we need to know more than that. We need to know if we're growing as much, now, as we did 10 years ago. Or 30 years ago, or 50. 

Are we growing faster than we did years ago, or slower, or about the same? You cannot tell by looking at the first graph. But you can tell by looking at this one:

Graph #2: GDP with price changes removed and the Vertical Axis showing a Log Scale

The second graph shows the same data as the first. I only checked a box to make the vertical axis show a Log Scale. Just the vertical axis is different. But the line now shows a downward curve rather than an upward curve. The downward curve on a log scale graph means growth is slowing.

On a graph with a log scale, a straight line indicates a roughly constant rate of growth. And the steeper the line, the faster the growth is.

To my eye, the blue line from 1950 to 1970 is pretty straight, and more steep than from 1966 to 2007. And then from 2010 to 2022 the line is even more straight, and even less steep. I added three trend lines by eye, based on these straights and dates:

Graph #3: Same as #2, with three trendlines added by eye

The early period (1950-1970) goes uphill faster than the midsection, and the midsection goes uphill faster than the period since 2010. Economic growth has been slowing. You can't see it on graph #1, but growth has been slowing. The log scale graph makes it easy to see.

You may notice that the middle period starts before the early period ends. Yeah. I look for straight-line trends from left to right, and I look again from right to left. Sometimes the right-to-left view suggests an earlier date, like the 1966 of the middle period. Also I think that when a trend changes, it often takes a few years for the change to be completed -- and 1966-1970 is one of those times.

I jotted down my dates for the three "straight" sections where the growth rate doesn't change much. Then I went back to the FRED data and figured the average growth rate for each of those three sections. There is more difference than I expected:

Graph #4: Average rates of Growth, where the plotted line is relatively straight

Just over 4% in the early period. Just over 3% in the middle years. And just over 2% in the most recent period. Growth slowed by about one percentage point after 1970 and again after 2007.

I know, one percent doesn't sound like much. But one percent every year from 1970 to 2007, compounded, does add up. In his 1995 book To Renew America, Newt Gingrich said that if we can get "a 1 percent increase in our economic growth rate" then "the Social Security Trust Fund never runs out of money". One percentage point more growth means a lot more GDP.

And yes, Gingrich was yearning for the good growth of the 1950-1970 period. As am I.

Trend lines by eye are nice, but trend lines by Excel are better. The next graph shows Real GDP (the same data as before) and a trend line based on the 1950-1970 data, using a log scale:

Graph #5: Real GDP and the 1950-1970 trend line

The trend line (red) is an exponential curve, the curve that curves uphill like crazy, like a covid graph. But an exponential curve shows a constant rate of growth, so on a Log Scale graph the exponential curve is a straight line. Yeah, I am still amazed by this -- and sometimes fooled by it.

Now, think of the red trendline as showing where Real GDP would be if growth kept to the 1950-1970 trend rate. And the Real GDP line, well, it's just sad and droopy. Real GDP is growing a lot slower now than it did 50 to 70 years ago:

"Not-strong-enough-to-say-NO" is *not* the problem

From a comment by David, on "Inflation: True or False" by David R. Henderson at the Hoover Institution:

Borrowing creates money out of thin air, which is an expansion of the money supply. The largest borrower/debtor on the planet is the U.S. Treasury. In 2017, the U.S. National Debt was $20 trillion. As if that wasn't bad enough, by 2020, it had grown to $28 trillion... The driver of all that borrowing is some combination of Congress and the White House--primarily Congress, as the Treasury has no option but to borrow to pay for the excess and largess of a Congress that is not strong enough to say, "No."

That seems to be what everyone thinks: Congress or somebody is not strong enough to say "No" to government spending. When you get right down to it, that's probably why people are insurrecting.

Here's the thing: 

• Cutting government spending will not fix the problem. 
• Raising government spending will not fix the problem. 

The size of government is not the problem. I know what Reagan said:

"Only by reducing the growth of government," said Ronald Reagan, "can we increase the growth of the economy."

I know. But Reagan was wrong about why growth was slow.

 

People sometimes measure the size of government by the size of government debt. That debt is huge: inexplicably, incomprehensibly, incredibly huge.

The biggest problem with government debt, I think, is that we have lost control of it. We cannot stop the increase.

But the reason we cannot stop the increase is simple: We have the wrong solution.

Everyone thinks we have to reduce government spending. As Rush Limbaugh said:

what actually causes budget deficits [is] spending more money than you have.

But Limbaugh didn't mention income. He only mentions spending and the money we "have".

The problem is with income.

Adam Smith wrote:

Every workman has a great quantity of his own work to dispose of beyond what he himself has occasion for; and every other workman being exactly in the same situation, he is enabled to exchange a great quantity of his own goods for a great quantity or, what comes to the same thing, for the price of a great quantity of theirs.

In the Project Gutenberg version of The Wealth of Nations, the phrase "what comes to the same thing" occurs 19 times. Smith must have thought the phrase important. A quantity of output comes to the same thing as the price of that output. It has the same value. In other words, income equals output.

GDP can be measured as the value of output we produce in a year, or as the value of income we earn in a year. The only difference between the two totals is due to measurement error.

The important thing, and what surprised me most in Econ 101, is that income equals output. So if GDP growth is slow it means the growth of output is slow, but it also means the growth of income is slow. And slow income growth is the kicker.

Reagan wanted to increase the growth of the economy so that income would increase. That is something we all want. How to make it happen is the question.

In my previous post I show this graph:

The graph shows the relation between the federal debt and everyone else's debt, for the US. Our debt is bigger than the federal debt, except for a few years at the end of World War Two. Our debt was more than five times the government debt in 1974 when GDP growth slowed. Our debt was more than five times the government debt in 1929 when the Great Depression started. Our debt was more than seven times the government debt in 2007, and then we had the financial crisis and the Great Recession. When our debt is too much more than the government debt, bad things happen.

Our debt was a bit over three times the government debt in 1919, and 3½ times the government debt in 1993. Our debt was not low then, but it was low enough that the economy could grow with vigor. After 1919 we had the Roaring Twenties, and after 1993 we had what Alan Greenspan called "the New Economy". Both bouts of vigor ended in disaster: high debt and disaster.

Our debt was less than the government debt in 1945, and after the war we had a "Golden Age" that lasted all the while our debt was less than three times the federal debt. We had troubles when the ratio went above three. The "Great Inflation" began around 1965, but our debt kept growing faster than government debt and the economy kept growing. Then in the mid-1970s, productivity growth slowed and GDP growth slowed and income growth slowed.

It was at this point that government debt started to grow faster. It grew slightly faster than our debt for 20 years. The red line wanders slowly downhill during those years. Finally, in the 1990s, our debt was low enough (relative to federal) that the economy could grow with vigor. 

But we always let debt-other-than-federal increase until it makes the economy go bad.

Really, the problem is policy. It's the government's job to make the economic environment a good one for growth. If you want to be angry with government for something, let it be this: not that the federal debt is too high, but that debt-other-than-federal is too high. Their policies encourage that. They should discourage it.

It is the times when debt-other-than-federal is low that economic growth is at its best. But of course the economy grows because we use credit, and our debt increases because we use credit. When our debt gets high enough, it brings trouble to the economy.

The saddest part of all this is that our solution is to reduce the federal debt. What we need is to reduce debt-other-than-federal -- just the opposite of what we are doing.

Evaluating credit as a tool to boost aggregate demand

You've seen the GDP equation Y = C + I + G + NX where

• C = Consumer spending
• I = Business Investment
• G = Government spending, and
• NX = Net Exports

C and I and G represent the three major domestic sectors of the economy, and NX covers trade with other nations.

Okay, I want to look at interest cost, not GDP. But I want to look at interest cost in our whole economy. So again there are the three major domestic sectors. In place of Net Exports we could have interest paid by "rest of the world" but I don't want to count that: I want to count the interest paid by us -- by US households and businesses and governments. And I want to count that interest, no matter who received the payment. So I have C and I and G but I don't need Net Exports.

To count all of G it is necessary to count two parts: the federal component, and the state-and-local component. So to count all interest cost, I count interest paid by C + I + F + SL.

I could use FRED's "Monetary Interest Paid" dataset, but that only goes back to 1960. C and I and SL go back to 1946, and F goes back to 1929. Evidently, FRED doesn't combine those datasets, probably because they are incompatible.

I'm going to combine them anyway, to get a rough idea of what the interest cost might have looked like all the way back to 1946.

Here is the combined data:

Graph #1: The orange data continues out to 2020. This graph stops at 1975.

The blue line is my estimate of interest paid during the 1946-1959 period. It includes payments by households, businesses, the federal government, and state and local governments.

The orange line begins in 1960 and shows the FRED data for the total Monetary Interest Paid.

That was more work than I expected, considering the simple thing I want to do with the resulting data: I just want to compare the cost of interest to the new use of credit each year.

I'm thinking of new-use-of-credit as extra money spent into the economy. And I'm thinking of cost-of-interest as a reduction of money available for current spending. I think the "extra money" will have to be greater than the "reduction of money", or the economy will get no boost from the use of credit.

 

My purpose in this post is to evaluate the flow of credit into the economy, and the flow of payment-for-that-credit out of the economy, or (more precisely) out of the producing-and-consuming sector and into the financial accumulation sector. My task is an attempt to see how useful credit is as a tool for expanding aggregate demand in the producing-and-consuming sector. 

My gut says it is no longer useful, because the accumulation of debt is so big that the interest we pay -- even at low rates -- is more than our new uses of credit. I figure it is the financial accumulation sector that benefits most by our use of credit.

The graph below shows for each year the total amount borrowed by all sectors of the US economy, minus the amount we paid as interest on our debt. Where the plotted line is above zero, it shows a boost to the economy from our borrowing and spending. Where the line is below zero, it indicates a drag on the economy arising from interest costs that exceed the amount of our borrowing. Where the line is at zero, there is no boost and no drag. Note: The values shown on the graph are annual, not cumulative:

Graph #2: That tall spike on the right is the covid-related borrowing of 2020

Until around 1978, the plotted line runs near zero. It says there was no net gain from our use of credit, but no net loss either.

After 1978 the plotted line is almost entirely below zero. Net loss almost all the time, except in the financial accumulation sector. This could explain the relatively poor performance of our economy for the last 40 years and more.

Below is a close-up of the early years, from 1946 to 1982:

Graph #3

It runs at or slightly above the zero level until around 1966; then at or increasingly below the zero level. In other words, there was a very slight boost to the economy from credit use, early on. But after 1966 the use of credit did more harm than good, as the interest cost was generally more than the boost we got from credit use.

That drag on the economy continued after 1982, as graph #2 shows. That graph can explain the slowing of the US economy since 1966. Thus one argument -- the growth of finance -- explains the long-term slowing of economic growth.

 

The Excel file at DropBox: Monetary Interest Paid datasets 3 Jan 2022 ArtS.xls