In search of: Older Debt Data

Graph #1

I got this graph from bctich at Reddit.

I have to go looking for the numbers. I've used the Historical Statistics, one of the sources noted on the graph, but that one runs from 1916 to 1970. I need the older numbers.

ps, the exchange at Reddit is from a year ago. But I forgot about that graph until I rediscovered it by accidebt.

Wow, I'm not fixing that typo!


From George Washington at Zero Hedge:
138 Years of Economic History Show that It's Excessive PRIVATE Debt Which Causes Depressions

The National Bureau of Economic Research has published a new paper analyzing 138 years of economic history in 14 advanced economies, which proves that high levels of private debt cause severe recessions. As summarized by Business Insider: Through a series of tests run on a sample of 14 advanced economies between 1870 and 2008, Mr Taylor establishes a link between the growth of private sector credit and the likelihood of financial crisis. The link between crisis and credit [i.e. private debt] is stronger than between crises and growth in the broad money supply, the current account deficit, or an increase in public debt.

That's good, that's good.

The "new paper" is The Great Leveraging (PDF) by Alan M. Taylor. The link in the excerpt lets you to buy the PDF for $5, but you can get the PDF from BIS for free.

No data in the paper. A few graphs, including this, with source info:

Source: The sample period is 1870–2008. Bank loans are loans by banks in aggregate to the nonfinancial sector, excluding interbank lending and foreign currency lending based on Schularick and Taylor (2012). Public debt is total sovereign debt outstanding based on Reinhart and Rogoff (2009). See Jordà, Schularick, and Taylor (forthcoming)

A search for Jordà, Schularick, and Taylor turned up a page listing Alan M. Taylor's economic papers. First item on the list:

 NEW    American Economic Review

Schularick & Taylor “Credit Booms Gone Bust: Monetary Policy, Leverage Cycles and Financial Crises, 1870–2008”   NBER WP 15512    Dataset

Dataset. That sounds good.

The link wants to download a 1.3 mdg zip file. The zip is full of little files I cannot identify by type. But there's a "readme" that says you need STATA to access the files. STATA is a purchasable product, so I'm out.

(Not sure what happened next. I guess, when I couldn't use the dataset I left the page without looking at the PDF. Later I found another link to a PDF with the same title.)

Schularick, Moritz, and Alan M. Taylor. 2012. Credit Booms Gone Bust: Monetary Policy, Leverage Cycles, and Financial Crises, 1870–2008. American Economic Review 102 (2):1029–61.

link page: http://www.nber.org/papers/w15512

PDF: http://www.nber.org/papers/w15512.pdf?new_window=1

From the conclusion of that PDF:

Our ancestors lived in an Age of Money, where credit was closely tied to money, and formal analysis could use the latter as a proxy for the former. Today, we live in a different world, an Age of Credit, where financial innovation and regulatory ease broke that link, setting in train an unprecedented expansion in the role of credit in the macroeconomy.

© 2009 by Moritz Schularick and Alan M. Taylor. All rights reserved.

Exactly right: These days, we use credit for money. But Schularick and Taylor seem to think that the credit monster we've evolved into is a beast that can survive.

It cannot survive. We can go back to an Age of Money, or our civilization can die. It is that simple. It is that cut and dry.

What was it Toynbee said? Civilizations die by suicide.

Well I'd love to stay and chat, but the lawnmower is calling.

Having it both ways

I looked at one of Marcus Nunes' graphs earlier today:

Graph #1: Employment 1979-1989 (Marcus Nunes)

Marcus said the graph shows how lucky you would have been to graduate in 1982 (where the dotted line appears). I couldn't see it. To me, the dotted line comes at a low point in employment, an absolute bottom, the worst time to be looking for work. Employment was getting worse before the dotted line, and better after the dotted line. But all of 1982 and '83 was a bad time to be looking for work, to my way of thinking.

In my earlier post I said the graph makes it look like employment was running flat at a low level before the dotted line, and things suddenly got better after the dotted line. The graph creates a false impression, I said.

Marcus's graph shows only ten years -- only three or four of them before the dotted line. That is the source of the false impression, I said.

Marcus replied, saying the discussion was about "a very specific period of time. Why should I bring into the argument the distant past?"

In that reply, Marcus said the dotted line on his graph marks "the very beginning of a long boom."

So I got thinking about that, and how there is that boom there after the dotted line, and how there was a boom earlier, too, but it got cut off. Bobbitted. So I reattached it:

Graph #2: Employment 1975-1989

When I looked at this graph, it seemed strangely familiar. Suddenly, I knew why. So I eyeballed-in a trend line, in red:

Graph #3: Employment 1975-1989 & Trend Line

Graph #3 seems to show the employment trend (red) and an employment gap (blue). But Marcus assures us that what we're looking at is not an employment gap, but "the very beginning of a long boom."

I never looked at it that way. But perhaps that solves a problem that Marcus himself has been having. Here's a graph of NGDP and a brief remark from Marcus's WeaK Signals:

NGDP is “content” to remain on a “flight plan” far below any reasonable trend level.

Do you see it? This is exciting stuff! Marcus is worried about nothing! It's not that NGDP is running "far below any reasonable trend level." It's that we're at "the very beginning of a long boom."

Did I get it right, Marcus?

Chop Shops

For me, the goal is always to get as much data as I can. The more years, the better. Even so, I generally ignore the first ten years.

There's nothing scientific about that. It's just a habit I developed while looking at graphs of economic data. It's tempting to think you can imagine what the numbers were doing in the years just before the start of the graph, but I've been wrong often enough to know better. And since I cannot see what happened just before the first years of a graph, I don't trust the stories about what I do see in those first years...

I trust that what's there in those first years is valid. I don't trust the interpretations of it or the explanations of what brought it about. But that's just me.


Marcus Nunes (7 September 2013) quotes Ed Lambert:

Volcker put the breaks on the liquidity coming from the Fed rate in the early 80′s. And the babies born from 1960 on graduated from college into a world of tight liquidity due to tight Fed policy. They became disadvantaged. They did not find job openings. As a generation, they fell behind. I was one of those babies that graduated into the heart of the 1982 recession.

A little too personal for me, Ed, but I take your point. (BTW it's brakes, like brakes on a car: for slowing or stopping the car or (in this case) the liquidity.)

Marcus responds:

Actually it appears he had greater luck than those who graduated a few years earlier! The charts show how lucky Edward was!

Here's the second of Marcus's graphs, from that same post:

Graph #1 from Marcus Nunes

Marcus's graph makes it look like employment was stuck at a low level until Ed Lambert needed a job, and then things suddenly picked up. That's not really what happened. The graph creates a false impression because it only goes back to 1979 and only shows 10 years in all. Marcus uses the false impression to build his story.

Here's how FRED shows Employment for the 1979-1989 period, just like Marcus:

Graph #2: Employment, 1979-1989

Looks just like Marcus's graph. But wow, my numbers are way different from Marcus's numbers. Marcus, você pode explicar isso? Maybe we used two different data series? And yet, the patterns appear to be identical.

Here's the same FRED data for the full period, with Marcus's ten years in red:

Graph #3: Employment since 1939

You have to tip your head to the left a bit, so that the employment line since 1960 looks like a flat line. That's the "trend". (Employment was slowing before 1960, and also after 2000, relative to this trend.)

When you look at that trend since 1960, you can see a flat line with high spots and low spots in it. A low spot near every recession bar. One of those low spots, the big red one, occurred just when Ed Lambert graduated.

The low spots are not good times to go looking for work. Lambert went looking for work at one of the low points of employment. I don't see how Marcus can call that luck, unless he means bad luck.

Marcus chops his graph down to just a few years, making it look like the employment numbers support the story he tells. But what looks like a flat spot on Marcus's short graph is really a low point in a generally rising trend.


Ed Dolan (23 September 2013) does a pretty decent evaluation of the Money Multiplier, contrasting old and new views. He opens with a look at the older view:

The multiplier posits that there is a stable ratio between M2 ... and the monetary base

Dolan writes:

There is just one problem. As the following chart shows, something has gone badly wrong with the money multiplier in recent years. For most of the 1990s and 2000s, it was steady as a rock. From 1994 to 2007, the 12-month moving average of the multiplier stayed in a narrow range, between 8.0 and 8.4. Then it fell off a cliff. By July of this year, it had reached a record low of 3.24.

Yep. Dolan could have stopped the graph in 2007, if he wanted to defend the "stable ratio" argument. He didn't do that, maybe because he he's not defending the stable ratio argument. Or maybe because he wants to look at the graph and see what actually happened before he picks an argument to defend. That would be good.

Here's the FRED graph that corresponds to Dolan's graph:

Graph #5

Here's the full FRED dataset:

Graph #6

Who in his right mind would call that "stable"? And why does Ed Dolan not show the years before 1993?


Warren Mosler (24 September 2013) writes:

Interesting chart- inventory of existing homes for sale vs the labor force participation rate…
new home sales track closely as well…

I couldn't find data for inventory of existing homes for sale but FRED has new home sales, so I looked at that. Since 2005, like Mosler's graph:

Graph #8 (since 2005)

Next, the full dataset:

Graph #9 (all years)

"New home sales track closely," Mosler says. That wouldn't be my argument. My argument would be: You can take any little short period of time when two numbers are moving in the same direction, and zoom in on just that period of time, and pretend to show evidence of whatever you want to claim.


EDIT: Concluding sentence removed.

Oh, what a lucky man he was

Marcus Nunes (7 September 2013) quotes Ed Lambert:

Volcker put the breaks on the liquidity coming from the Fed rate in the early 80′s. And the babies born from 1960 on graduated from college into a world of tight liquidity due to tight Fed policy. They became disadvantaged. They did not find job openings. As a generation, they fell behind. I was one of those babies that graduated into the heart of the 1982 recession.

Marcus responds:

Actually it appears he had greater luck than those who graduated a few years earlier! The charts show how lucky Edward was!

The first of Marcus's charts shows the Unemployment Rate for 1979-1989:

Graph #1 from Marcus Nunes.
Oops, I erased most of 1989, there on the right.

How lucky Edward was! If he kept looking for work until 1984 his chances would have been as good as they were in 1980. If he kept looking till 1988 his chances would have been as good as they were in 1979. Lucky guy, Ed!

Here's how unemployment looks at FRED for the 1979-1989 period, same as Marcus:

Graph #2: The Unemployment Rate, 1979-1989

Here's the whole data series, with Marcus's bit of it highlighted red:

Graph #3: The Unemployment Rate, 1948-2013

Edward Lambert went looking for work just when unemployment was at the highest point of the whole 1948-2013 period. Marcus says Edward was a lucky man.

Hetzel (1998) on Burns

From Arthur Burns and Inflation (PDF) by Robert L. Hetzel:

In November 1970, the minutes of the Board of Governors show Burns telling the Board (Board Minutes, 11/6/70, pp. 3115–17) that

prospects were dim for any easing of the cost-push inflation generated by union demands.

In 1970, Arthur Burns attributed inflation to union wage demands.

One page later, Hetzel presents a more tangled tale:

How did Burns view macroeconomic policy as an economist? Most generally, Burns had a credit view of monetary policy. That is, monetary policy worked through its influence on the credit market. However, monetary policy was only one factor affecting credit markets. At times, in its influence on inflation, monetary policy could be overwhelmed by other factors. More specifically, Burns had a real or nonmonetary view of inflation. That is, inflation could arise from a variety of sources other than just money. He believed that a central bank could cause inflation by monetizing government deficits but did not attribute inflation to that source in the early 1970s. Instead, he attributed it to the exercise of monopoly power by unions and large corporations.

If conventional monetary policy weapons were powerless to deal with these forces, then perhaps direct controls might work. Accordingly, President Nixon imposed wage and price controls August 15, 1971. The experience with such constraints offered a tailor-made experiment of Burns’s views. The controls worked as intended in that they held down wage growth and the price increases of large corporations (see Kosters [1975]). Nevertheless, inflation rose to double digits by the end of 1973. So Burns attributed inflation to special factors, such as increases in food prices due to poor harvests and in oil prices due to the restriction of oil production. However, special factors are by nature one-time events. In 1974, inflation should have fallen as the effect of these one-time events dissipated, but it remained at double-digit levels that year. Burns then blamed inflation on government deficits. Although those deficits were small in 1973 and 1974, Burns was able to make them look larger by adding in the lending of government-sponsored enterprises like the Federal National Mortgage Association.

Burns said inflation was caused by wage demands. Wage and price controls worked as intended but inflation rose anyway. Then Burns changed his story and blamed special factors. But the special factors passed, and time went by, and inflation did not fall. So Burns changed his story again and blamed government deficits, even padding them to make them look bigger.

That's not my story. It's Robert Hetzel's, and it isn't kind to Arthur Burns. I don't know if the story's accurate. I know that other parts of Hetzel's PDF agree with things I hold true, things I read long ago, things written in the 1970s. Such old sources seem to me less tainted by modern views, and more reliable. And Robert Hetzel's story fits well with those.


Here is what I want to say:

Burns covers all the bases, all but one. He points the finger at workers. He points the finger at acts of God. And he points the finger at government. But he does not point a finger at finance, the rising cost of finance.

Maybe I'm wrong. Maybe the cost of finance is not the cost that's killing us. But we cannot possibly know that, if we refuse to look at it.

Bruenig on debt

Matt Bruenig has a post called Strike Debt calls for increasing inequality, again.
Strike Debt is some political organization, it seems. I only read a little of Bruenig's post. I had to stop when I got to this:

Needless to say, a focus on debt is pretty incoherent. The only way you make it less incoherent is to talk about income inequality and how that affects what debt means to a given class. But doing that shows you aren’t mad about debt; rather, you are mad about distributive inequality and are simply expressing that anger through the horribly clumsy proxy of debt.

Needless to say?

Let's get a definition out of the way: What is debt?

Debt is created when money is borrowed. Debt is the record of money borrowed and not yet repaid. That is what debt is. That is the definition. When most people talk about debt, as in "using debt" for some purpose, what they mean is "using credit" for that purpose. The debt is the record of the credit that was used, and not yet repaid.

If you think a focus on debt is "pretty incoherent" maybe it is because you don't know the difference between debt and credit. Or maybe it's because you have not noticed that accumulated debt, and the interest paid on it, together determine the cost of money -- a cost that competes with wages and profits, a cost that hinders production.

Matt Breunig brings up "income inequality" and "class" in his post. What he says there seems right: Some people, anyway, may be confusing debt-related problems with inequality-related problems, and the reverse. I had a thought on that. I want to take something I wrote for the 6 September post, and tweak it just a bit here:

If we spend the interest we receive, then it doesn't matter much what portion of our income is interest. But if the interest we receive stays in savings, then it matters very much. For even if interest expense and interest income both were owned and owed equally by everybody, but we tended to save regularly and never spend our savings, then we would eventually but inexorably create the sort of financial crisis that just ruins an economy for years and years and years.

If wealth and income are not equally distributed, that only speeds the process.

Debt is the measure of money that we have put into circulation, that we are paying (debt service) to keep in circulation. The problem with debt is the cost of it -- the cost it adds to using money in the macroeconomy. The reason we have so much debt is: people think it's a good thing. And policy is based on that thought.

I talk plenty about debt, and never about class. Why? Because debt, excessive debt, is a problem that affects the macro economy. Does it affect groups and classes and sectors within the economy? Maybe, but I don't have much to say about that.

"I'm going to talk about an economic cycle without defining or identifying it."

Graph #1: Accumulated Debt relative to the Quantity of Base Money
The Google Drive Spreadsheet.

The Economist: Wrong and wronger (2)

(continuing yesterday's critique)

Regarding the second topic, the purchase of existing assets: Sure, if we spend new money to buy old things, debt will increase faster than GDP. Okay, I'm not disputing the obvious. I'm disputing the significance of it.

I'm not going to call it buying existing assets. I'm going to call it the speculative use of debt. (It would be more accurate to call it the speculative use of credit, but people would mostly fail to understand.)

And I'm going to talk about an economic cycle without defining or identifying it.

I think the speculative use of debt in general happens late in the economic cycle. I see this as a result: a changed behavior arising in response to conditions that are created by the increasing reliance on debt that happens in the earlier part of the cycle.

People get more and more into buying existing assets, after the debt problem arises and before most people even see there is a debt problem.

I think people see the rise of debt being used for speculation, and recognize it as a problem, but fail to dig into it and fail to realize that the debt problem started much earlier and that speculation developed in response to that problem.

Thus, while debt-for-speculation is a significant problem, it is a resultant; to focus on it is to assign it too much significance, in the larger picture of causalities.

The Economist: Wrong and wronger

I want to take another look at the excerpt we saw yesterday, from The Economist:

When a bank makes a loan, it credits the money to the borrower’s deposit account. In so doing the loan adds to the money supply. If that money is spent on a new car, factory or other freshly produced good, it contributes to demand, helping the economy to make fuller use of its productive capacity. If the economy is already near full capacity, it will probably just raise prices instead.

The phrase "freshly produced" is a little creepy, I'd say, but it's easy enough to see the meaning. And as for the meaning, it makes sense to me. At least, the part of the excerpt repeated above makes sense. I think the rest of it is nonsense:

But either way, the bank lending will add both to debt and to nominal GDP, the money value of economic output, leaving the ratio of debt to GDP largely unchanged.

However, loans can also be spent differently. They can be used to buy existing assets, such as homes, office-blocks or rival firms. Since the asset already exists, its purchase does not add directly to GDP, which measures only the production of new goods and services. As a consequence, debt increases, but GDP does not.

There are two topics there in the nonsense part, in the two separate paragraphs:
1. The consequence of buying "freshly produced" stuff.
2. The consequence of buying "existing assets".

I want to treat these separately. But I had to leave 'em together, uninterrupted in the excerpt, so you could see that the first thought dies with the words "leaving the ratio of debt to GDP largely unchanged." There's nothing else there. There is no follow-up thought that you're not seeing because of some editing I did. It's not there, in the original article. It's just not there.


Regarding the first topic: The article says as long as we use our borrowed money to buy new stuff, the ratio of debt to GDP remains "largely unchanged".

Don't depend on it. I know a lot of people agree with the view expressed in the excerpt. I'm just saying, don't think the issue is settled.

Debt is a stock, and GDP is a flow. Next year, we still count the debt we created this year, or what's not yet repaid of it, but this year's GDP doesn't count at all. Next year, we start counting GDP at zero, but debt starts from a high number. Debt accumulates; GDP does not. It's that simple, really.

Why does debt grow faster bigger than GDP? Because debt is a stock and GDP is a flow.

On top of that, money tends to disappear. Some of it goes to trading partners. A lot is saved. And some slips down between couch cushions. All these things make money disappear from circulation, reducing the spending that is counted as GDP. (It reduces spending in general, including the spending that counts as GDP.) The disappearance of money undermines the expansion of GDP. The bigger the trade deficit, and the greater the saving, the more is GDP growth undermined.

Debt does not disappear like that. Therefore, contrary to what the article says, the ratio of debt to GDP does not tend to remain unchanged. In the U.S. experience, it takes a great inflation to stabilize that ratio.

The Economist gets it wrong

Sloppy work, and not just at The Economist. I've seen this argument made several times before. Here's how The Economist puts it:

When a bank makes a loan, it credits the money to the borrower’s deposit account. In so doing the loan adds to the money supply. If that money is spent on a new car, factory or other freshly produced good, it contributes to demand, helping the economy to make fuller use of its productive capacity. If the economy is already near full capacity, it will probably just raise prices instead. But either way, the bank lending will add both to debt and to nominal GDP, the money value of economic output, leaving the ratio of debt to GDP largely unchanged.

However, loans can also be spent differently. They can be used to buy existing assets, such as homes, office-blocks or rival firms. Since the asset already exists, its purchase does not add directly to GDP, which measures only the production of new goods and services. As a consequence, debt increases, but GDP does not.

Maybe I shouldn't generalize, but when the wife and I bought our new place, it was ten years old. In order to help us buy it, we sold our old place. And after we bought the new one, the people who sold it to us built themselves a new home. They relied on our buying their old house, just as we relied on the people who bought our old house.

There was a chain of transactions, three that I know of, of which only one was a quote productive unquote transaction that resulted in new construction.

Well yeah, but have you seen the price of housing? Today, a whole chain of transactions is required so that one fellow can afford to put up a new house! So don't tell me that my transaction was non-productive. Call it non-final, if you want.

This whole "buying an existing house is not productive" thing is a meme, spreading like a bad cold. Don't catch it. It's not right.

The Economist gets it right

What has changed in three years?

In mine of September 17, 2010 I looked at an article from The Economist:

From A special report on debt: Repent at leisure, The Economist, 24 Jun 2010

Hyman Minsky, an American economist who has become more fashionable since his death in 1996, argued that these debt crises were both inherent in the capitalist system and cyclical.

Inherent, and cyclical. Agreed. At The Economist, they know an impressive statement when they see it, at least if the speaker is fashionable.

Debt increased at every level, from consumers to companies to banks to whole countries. The effect varied from country to country, but a survey by the McKinsey Global Institute found that average total debt (private and public sector combined) in ten mature economies rose from 200% of GDP in 1995 to 300% in 2008... There were even more startling rises in Iceland and Ireland, where debt-to-GDP ratios reached 1,200% and 700% respectively.

"Debt increased," the Special Report says, but "the effect varied." The excerpt suggests the "effect" was that debt increased, and the variation was that it increased to various levels. This is not an impressive analysis.

Today (18 September 2013) a Reddit link brings me to The dangers of debt: Lending weight in The Economist of 14 September:

It was the growing rate of default on home mortgages in America that precipitated the financial crisis five years ago. These delinquencies, although not enormous in themselves, became impossible for some investment banks to bear, thanks partly to their own heavy debts. As the contagion spread throughout the financial sector in 2007-08, nervous or cash-strapped banks and other creditors stopped lending, thereby infecting the rest of the economy. Deep recessions and big financial rescues then led to a surge in government debt. That, in turn, raised fears about the solvency of various countries in the euro area, culminating in Greece’s default in 2012. Debt was, then, both a cause and a consequence of the crisis, and remains a big reason for its continuance.

Three years ago they couldn't even identify the problem with debt. Today they can identify it, they do identify it, and they use it as their opening statement! Even if The Economist is still just being fashionable, and nothing more -- and I'm not saying that's the case -- the story is definitely different now. That's a good thing.

Could've inspired Steve Waldman's Not a monetary phenomenon

Steve Waldman links to Kevin Erdmann of 6 September. That post links to Erdmann's earlier It's All Demographics, Again

Erdmann takes a population forecast for 2015-2060, projects labor force participation for that period by age and gender, and creates a labor force participation forecast through 2060. Then he combines this forecast with existing data and creates a graph showing more than 100 years overall. And he offers an insight that simply astounded me:

The most important thing to note here is that the current slope of the line basically tracks the slope we have seen since 2000. You don't need a disillusioned labor force that's given up in order to explain this decline.

I'm trying to say enough to get you interested, but not so much that you don't need to read Erdmann's post. I've probably not done it justice. Go read Kevin Erdmann's post for yourself, if you missed it.

Two Mismatches

Steve Waldman's key graph compares growth in the civilian labor force and growth of NGDP:

Graph #1: Steve Waldman's Key Graph

Thinking about the things he's comparing -- labor force and output at actual prices -- Okun's law came to mind. Okun's law says there's a relation between employment and output. And SRW's graph says there's a relation between employment and output.

Maybe that's a little sloppy. The graph shows a relation between output and employment plus unemployment, where employed and unemployed people together make up the labor force. Still, there might be a relation...

Okun's laws says if unemployment goes up one percent then output goes down two percent. And if unemployment goes down one percent then output goes up two percent. Stuff like that.

If we have a population of 100 workers and 95 of then are working, then we have 5% unemployment. By Okun's law, if unemployment goes up then employment will go down. Suppose 50 people join the workforce and none of them get jobs. Now we have 55 people unemployed out of 150 workers. That's like 36% unemployment -- an increase of 31 percentage points.

Okun's law says for every percentage point unemployment goes up, output goes down two percentage points. So with a 31-point increase in unemployment, output should fall 62% Yikes! But obviously that's not right. Output doesn't drop just because people try to get jobs. So Okun's law really doesn't apply.

Also, apparently Okun's law applies to "real" (inflation-adjusted) output, not to the "nominal" (actual-price) output Steve Waldman is considering.


Then it occurred to me that, yes, SRW is considering nominal output, not real output. And that I had no idea what the ten-year growth rate of real output looks like. Maybe it looks like the nominal rate, only lower? Nope:

Graph #2: 10-Year Growth Rates in the Civilian Labor Force (shaded) and RGDP (pink)

The civilian labor force pattern rises to a late-1970s peak and then declines, just as in Waldman's graph. But the pattern of RGDP is completely different. High early, in particular from the mid-1960s to the mid-1970s, and low thereafter.

Oddly, there is a low spot in the early years that's just as high as the plateau that runs from the latter 1970s to around 2006. And oddly, there is a brief low in the early 1980s -- looks like the combined effect of the 1974, 1980, and 1982 recessions -- that is higher that output has been since 2008.

But there's not much similarity between real output growth and labor force growth. Not that I see on this graph, anyway.

Can I duplicate that?

Steve Waldman's Not a monetary phenomenon struck me immediately as an unusually fascinating piece of work. It seemed to draw plenty of attention from others, too.

Graph #1: From Steve Waldman's Post

So after looking at a few preliminary graphs these past few days, I looked at the key graph from his post and wondered. Can I duplicate that? I got the Labor Force and NGDP numbers from FRED -- quarterlies -- and sat down to see.

Graph #2: Duplicate of SRW's 10-Year Growth Graph

Yup, nothin to it. So I looked at the graph a while, and then I thought: Well, that's the 10-year change. I wonder what the 5-year change looks like...

Graph #3: The 5-Year Growth Rates

... and the 3-year change...

Graph #4: The 3-Year Growth Rates

... and the one-year change...

Graph #5: The One-Year Growth Rates

... which should look like what I can get from FRED:

Graph #6: The One-Year Growth Rates from FRED

So, yeah. I can duplicate that.

Here's my spreadsheet.

Afterthought

So I looked at FRED and sure enough they have a series called Working-age Population in the United States (USAWFPNA).

Starting with yesterday's Graph #4 (renumbered) ...

Graph #1: FRED's Mystery Population as a Percent of Total U.S. Population

... I want to back out FRED's mystery population number, which I thought must be the working age population, and see if it matches this new-found series.

CIVPART is the civilian labor force relative to that mystery population number

CLF16OV/PXX
So if I divide CLF16OV by CIVPART, I get their population number:

CLF / (CLF/PXX) = CLF * (PXX/CLF) = CLF*PXX/CLF = PXX
That's the blue line on Graph #2 here:

Graph #2: Two Sources for Working Age Population

Looks same-same to me. Actually, my calc (blue) goes back farther in time than the official stat (red) so it might be useful.

Oh, and here are both versions, relative to the total U.S. population. Just so it's easy to see that both versions look like Graph #1 above and yesterday's Graph #4.

Graph #3: Working Age Population relative to Total U.S. Population (two measures)

So yeah, yesterday's Graph #4 definitely shows the working-age population as a percent of the total population. And FRED's CIVPART definitely looks at the labor force relative to the working age population -- not relative to total population.

I love all these little details.

Population and Labor Force and Population

Yesterday we looked at population and prices. But Steve Waldman was not looking at population. He was looking at "civilian labor force". That's a different animal. For one thing there's an 18-year lag, or so, from when a newborn joins the population, to when he or she joins the labor force. And then, the chance that he-or-she actually does join the labor force varies over time:

Graph #1: Civilian Labor Force as a Percent of Population

A significant increase in the probability of joining the labor force began in the early 1960s, slowed just before 1980, and died out around 1990. Isn't that strange? What would cause millions and millions of people to make decisions that, when we add them up and look at the result, look like an agreement to increase our workforce to half our population. What would cause that?

For me the answer is always economic -- economic forces, or the economic concerns which give rise to economic forces, or both. Something I got from Hayek.


Oddly, FRED's "Civilian Labor Force Participation Rate" shows a pattern comparable to Graph #1, but the numbers are much different:

Graph #2: Likely a better measure of participation than is Graph #1

FRED offers 88 pages of data sets in response to a search for civilian labor force. I can't say for sure, but I'm thinking that the U.S. civilian labor force is the U.S. civilian labor force -- age, education, race, ethnicity, geography, and gender aside. I'm thinking the civilian labor force is the same in both graphs above.

Graphs #1 and #2 combined:

Graph #3: Participation Rate from Graph #2 (blue) and my calc from Graph #1 (red)

If the labor force number is the same in both ratios, then it has to be the other number that's different: It has to be the population number that's different. My ratio (the red line on Graph #3) is lower, which means my population number is higher.

If my pop number is higher, and mine is the "Total Population for the United States" then FRED must not be using total population. Maybe they're using the working-age population. That would make sense.

If -- again, if -- we're using the same labor force numbers (CLF) and different population numbers (POP and PXX say) then my graph is CLF/POP and FRED's graph is CLF/PXX. If I divide my number by their number I get

(CLF/POP) / (CLF/PXX)
Dividing by a fraction, I invert and multiply:

(CLF/POP) * (PXX/CLF)
The CLF top and bottom cancel out, and I'm left with their population (PXX) on top, and mine (POP) on the bottom. So that way I can look at their population number as a percent of Total Population:

Graph #4: FRED's Mystery Population as a Percent of Total U.S. Population

Well this result surprised me. I guess I expected something more stable. Still... what have we got?

If my assumptions are valid, the blue line shows the working-age population as a percent of total U.S. population. Should we expect an increase like that -- from the early 1960s to 1990, just as on Graph #1 -- as a result of the Baby Boom? Let's see.

According to the Wikipedia article, the Census Bureau dates the baby boom from 1946 to 1964, and an "echo" boom ("the children of the post-WWII baby boomers") from 1982 to 2000.

16 to 20 years after the 1946 start of the boom puts us at 1962-1966. That's a good match to the first leg of the increase on Graph #4, an increase that only accelerated thereafter.

16 to 20 years after the 1964 end of the boom puts is at 1980-1984. That's just at the beginning of the slowdown visible on all these graphs.

And 16 to 20 years after the 1982 start of the echo boom puts us at 1998-2002. Between those dates, another uptrend begins on Graph #4. (The vertical at year 2000 is a break in the data resulting from revised accounting. But there is a clear downtrend into the late 1990s, and a clear uptrend since 2000.)

So yes, I think so. I think Graph #4 shows the U.S. working-age population as a percent of total U.S. population.

Hmm.

Population on a Log Scale

The numbers are missing between 1944 and 1952. Otherwise, this graph goes back to 1900:

Graph #1: U.S. Population since 1900

Off the top of my head, it looks like the Baby Boom (first part of the red line) was no more than a return to the rapid population growth of the earlier years, which had been interrupted by a Great Depression and possibly a War.

Graph #2 shows the same measures of population (blue and red) along with two measures of prices, with prices scaled up to put the numbers in the same neighborhood as the population numbers.

Graph #2: Population and Prices (All as Natural Logs)

Unsustainable

From the PDF by Piketty and Zucman, linked by Carola Binder.

Click for a Better Look

The only way to save more than you grow, continuously, for 40 years or more, is to replenish the circulating money by printing or borrowing more. Otherwise there's soon nothing to spend, and the economy shuts down.

The problem here is not the printing or borrowing. The problem is the saving.

// Related post: Disastrous, cumulative and far-reaching repercussions of saving

My best criticism of Scott Sumner's analysis

From Wednesday:

Sumner makes the monumental mistake of applying "other things equal" to the real world. He assumes that nothing else has changed that could possibly be depressing growth, so that if growth is slow it must be because money is tight.

Canaries

This is John Quiggin's afterthought, that I quoted yesterday:

The neutral value changes gradually over time in response to a variety of factors, but is sufficiently stable that it can be regarded, for most purposes, as a long-term average, typically assumed to be in the range 1.5 per cent to 3.5 percent.

Quiggin was talking about the interest rate controlled by the central bank. But if we left off the range of values and just looked at the words about how the value is stable and only changes gradually, he could have been talking about any of a number of "neutral" or "natural" rates. He could even have been talking about the velocity of money.

I just wanted to say, in response to Quiggin's thought, that there have been sudden shifts in a lot of "sufficiently stable" variables in our not-too-distant past. Variables are called "variables" for a reason.

The gradual changes that come before sudden shifts are canaries. Indicators. Warning signs. They are not something to be thought of as stable. They can't be safely ignored. Can it be that economists don't know this?

// Related post: Two Thought Experiments

Loosey Goosey and the Four-Letter Word

Quiggin (September 1st, 2013):

The idea that the stance of monetary policy can be assessed as expansionary, neutral or contractionary depending on whether the interest rate controlled by the central bank is at, above, or below its real long run average neutral value isn't just mine. It's that of nearly all economists, notably including the US Fed. Update The neutral value changes gradually over time in response to a variety of factors, but is sufficiently stable that it can be regarded, for most purposes, as a long-term average, typically assumed to be in the range 1.5 per cent to 3.5 percent. End update Sumner claims that “People get defensive when I make fun of the view that low interest rates mean easy money” but doesn’t name any names. Does anyone really deny that this is the standard view?

Timewarp Sumner (August 29th, 2013):

I have to admit that Quiggin is right. Most economists do equate low rates with easy money. That is the accepted definition. How that occurred, how the lunatics took over the asylum, is beyond my comprehension.

Here's the thing. A measurement is a comparison. To measure the size of the hole in my head you put a tape measure to it and make a comparison. Same is true with easy and tight. Tight money is tight compared to easy money. Easy money is easy, compared to tight money.

But what if you want to look at a moment in time -- like now, say -- and see whether money is easy or tight? How does that work? It's still a comparison, but what do we use for the tape measure?

Quiggin compares the interest rate to its "neutral" or "long run average" value to determine whether monetary policy is easy or tight. Scott Sumner compares NGDP growth to its long run average value to make that determination. Sumner writes:

Woodford and Bernanke are right; the stance of monetary policy depends on outcomes like NGDP growth and inflation, not interest rates and the money supply.

But Sumner makes the monumental mistake of applying "other things equal" to the real world. He assumes that nothing else has changed that could possibly be depressing growth, so that if growth is slow it must be because money is tight. But if there is some other factor causing slow growth -- or if there could be such a factor -- then Sumner's evaluation cannot tell us whether money is easy or tight.

If there is some other factor making NGDP growth sluggish, easy money is probably not the solution. If there is some other factor, Sumner's whole argument falls apart.

And what else could possibly be responsible for slow growth? The four-letter word: Debt. Accumulated private debt. With excessive debt slowing the economy, you cannot use NGDP growth as a yardstick to determine if money is easy or tight.

But what does Sumner say? He says "Forget about debt".

Christopher J. Neely: Would It Help To Eliminate Interest on Reserves?

From Economic Synopses at the St. Louis Fed:

How much would eliminating interest paid on reserves increase bank loans and thus monetary aggregates? The experience of the ECB provides a case study of such an effect. On July 11, 2012, the ECB reduced the interest it pays on excess reserves held at its deposit facility from 25 basis points to zero. The chart shows that the European monetary aggregates did not grow unusually fast in the months following this action.

Tulipomania

From the Crisis Chronicles at Liberty Street Economics:

By the 1630s, the market for tulips began to grow as florists started buying and selling tulip bulbs still in the ground using promissory notes. The notes provided welcome credit and liquidity to help finance planting and limited credit risk to a known borrower with the borrower’s bulbs as collateral. However, the notes created a limited opportunity to inspect bulbs or to see them flower, provided no guarantee of quality, nor proof that the bulbs actually belonged to the seller, or even existed.

The post also considers Lessons for Regulators.

// Related post: 300 Years of Financial Crises

The best thing I've read in a long time.

Reviewing a PDF by Alan Blinder and Jeremy B. Rudd, Arnold Kling writes:

It is an article of faith among economists that the 1970s inflation and the 1980s disinflation both came from monetary policy, but that does not make it a proven fact. Maybe we have too much faith. Instead, we should be willing to examine data and adopt a skeptical perspective...

Amen to that!

// Update (8 Sept 9:35 pm)

Steve Randy Waldman writes

I am writing in 2013 about choices made in 1973 because I think a mythology has developed around 1970s experience that is very harmful.

There seems to be a lot of re-thinking going on.

Saving and Investment

Graph #1: Saving = Investment (more or less)

Graph #2: Saving Relative to Investment, and Unemployment

Graph #3: Investment less Saving, and Federal Deficits