Sunday, September 5, 2021

The mirage: Thomas Palley's "Financialization revisited", part two

 

Key points:

  • People say debt was low and stable until the 1980s when financialization arose and suddenly increased the debt-to-GDP ratio.
  • But inflation affects debt and GDP in ways that reduce the debt-to-GDP ratio. The great inflation severely reduced debt-to-GDP between the mid-1960s and the early 1980s. Paul Volcker and the disinflation of the early 1980s are the true sources of the sudden increase in the ratio after 1980.
  • Financialization did not arise in the 1980s. It has been with us at least since the 1960s. To de-financialize the economy it will be necessary to reverse, revise, or eliminate not only policies created since 1980, but also many policies created before 1980.

 

 
"Financialization revisited: the economics and political economy of the vampire squid economy" by Thomas Palley.


 

This is my one complaint: Thomas Palley cannot see anything before 1980. He refuses to look. For example, he refers to the "era of neoliberalism (1980 – today)". Financialization also began in 1980, he says:

The first stage corresponds to the period 1980 – 2008. The second stage corresponds to 2009 – today.

His definition:

“financialization corresponds to financial neoliberalism which is characterized by the domination of the macro economy and economic policy by financial sector interests.”

This definition, he says,

identifies financialization with the period 1980 – today, which distances it from the history of financial deepening and increased financial sophistication.

Palley expands on that thought in footnote 1:

Graebner (2005) has shown civilization is marked by the increased use and presence of finance. Prompted by that, Sawyer (2013, p.6 cited in Epstein 2015, p.5) asks whether financialization has been on-going throughout most of the history of civilization. This paper would definitively answer “No” to that question.

Definitively: "with absolute certainty."

Palley rejects the idea that financialization may have existed before 1980. With absolute certainty and the shortest of answers, he dismisses the possibility. For Thomas Palley, financialization began in 1980 by definition. End of story.

 

I can see that Palley defines financialization as having started in 1980. I can see that he is obligated therefore to provide an unequivocal "No" to the question in the footnote.

What I don't see is evidence. I see only insistence.

Evidence

Palley offers unsatisfactory evidence. "Prior to 1980," he says on page 31, "the domestic non-financial debt-to-GDP ratio was stable" -- and after 1980, it was not. But that is only more or less true:

Graph #1

His statement ties in with the idea that the rising level of debt is evidence of financialization. As he says in the Abstract: "financialization rotates through the economy loading sector balance sheets with debt."

For Palley, rising debt is evidence of financialization. From this, it follows that a low level of debt indicates the absence of financialization. And that, perhaps, is how Palley arrives at the idea that financialization began in 1980. And yet, the FRED graph shows rising debt all through the 1950s.

Palley doesn't consider the 1950s. The number sequence 195 does not even exist in the PDF. Nor does the PDF show a graph of domestic non-financial debt to GDP. Instead, Palley writes:

Table 1 shows sector debt-to-GDP ratios for selected years and provides evidence ...

Table 1 shows seven years of data: 1960, 1969, 1980, 1990, 2001, 2007, and 2019. Seven years scattered across six decades. Palley presents data for seven years out of 60, and calls it evidence. 

 

Palley's seven data points take us back only to 1960. He provides no evidence regarding the 1950s -- evidence, such as it is. Palley ignores the decade-long increase of debt-to-GDP in the 1950s, and insists that financialization -- rising debt -- did not exist before 1980. 

It just doesn't make sense. The ramping-up of debt in the 1950s contradicts any claim of stability before the 1960s; therefore it contradicts Palley's claim of stability before 1980. 

But was the increase of the 1950s significant, really? 

Yes. Graph #1 has debt-to-GDP at 1.25 in 1951 and 1.4 in 1961. That's a 12% increase in 10 years. If we got a 12% increase every time 10 years went by, in 2021 our debt-to-GDP ratio would have been 2.76, about halfway up the covid increase of 2020, upper right on the graph. You have a calculator. Check my numbers.

The debt-to-GDP increase of the 1950s was most significant. If the 1950s rate of increase had continued, debt today would be little different from the debt we ended up with.

 

If an increase in domestic non-financial debt-to-GDP is evidence of financialization, then the evidence shows financialization in the 1950s. This definitively contradicts Thomas Palley.

And not only Thomas Palley. In the 2011 paper The real effects of debt, Cecchetti, Mohanty and Zampolli (page 6) write:

It is sufficient, however, to look back at the history of the United States (for which long back data are easily available) to understand how extraordinary the developments over the last 30 years [1980-2010] have been. As Graph 2 shows, the US non-financial debt-to-GDP ratio was steady at around 150% from the early 1950s until the mid-1980s.
The authors generously allow the reader to make an eyeball estimate to confirm that the graph shows debt-to-GDP steady in the 1950s, based on this disproportionately wide graph:

The wider and squatter you make the graph, the more flat, steady, and stable the red line will appear. If you make the graph squat enough, there is zero increase of debt. This does not work with numbers, of course, but you can probably create in readers' minds the impression you want to create, by making your graph more wide and less tall.

Cecchetti's glaringly wide graph provides airquotes-evidence for the assertion that the debt-to-GDP ratio was "steady" from the early 1950s to the mid-1980s. But that doesn't make the assertion true.

Hey, I'm not saying the slope of the line in the 1950s is as steep as the slope after 1980 or the one after 2000. But the red line is not as flat in the 1950s as it is from the mid-1960s to the early 1980s. Around the 1950s it slopes up for a decade and more. Debt-to-GDP was rising. And if increasing debt implies financialization, then there was financialization in the 1950s.


It's not just Palley's paper and Cecchetti's. It's everywhere you look. Some years back in Finance is Not the Economy, Dirk Bezemer & Michael Hudson wrote:

Growth in credit to the real sector [the "non-financial" sector] paralleled growth in nominal U.S. GDP from the 1950s to the mid-1980s — that is, until financialization became pervasive.

It's the same story Palley tells: the same non-financial sector debt; the same debt-to-GDP ratio;  the same three decades; and the same stability before 1980 and financialization after.

Bezemer and Hudson also quote Richard Werner (2005) and Godley and Zezza (2006) making similar observations.


Evidence

Graph #3 below is another look at the same debt we saw on graph #1 above. But just the debt this time, not debt-to-GDP. And this time I show the debt on a log scale.

On a log scale graph, a straight line at any angle is a constant rate of change. For any two straight lines, the steeper one is increasing faster; the flatter one is increasing more slowly.

On graph #3, the slope of the line -- the steepness of it -- indicates how fast domestic non-financial debt is growing. The steeper the line, the faster the growth of debt. The flatter the line, the slower.

In the 1950s, the line goes up. In the 1960s the slope is almost the same but it goes up a little faster than in the 1950s. In the 1970s the line definitely goes up faster than in the 1960s. So does the debt. 

We didn't see increase in the 1960s and '70s on the first graph. And yet, here it is:

Graph #3: The graph shows accelerating debt growth from
1951 to the mid-80s, and slowing debt growth since.

After the 1980 and 1982 recessions the line runs faster uphill for a few years, but slows in the mid-80s and goes uphill at this slower pace until after the year 2000. These are the years that debt is imagined to be growing very fast according to Palley, according to Bezemer and Hudson, and according to Cecchetti, Mohanty and Zampolli. All of them.

Don't get me wrong. I think they are good economists. I love reading their stuff, and I agree with most everything they say. But they have picked up the idea that debt was "stable" before 1980, when in fact it was not. And they build that error into their arguments and their theories, creating an unsound foundation for their work. As a result they tell us silly things, like financialization started in 1980.

This is only a minor point. Or rather, it would be a minor point, if it wasn't wrong. But it is wrong. And it is a most dangerous thing, to let wrong ideas become embedded in economic thought.

 

Convince yourself

The level of non-financial debt accelerates upward through the 1950s, '60s, and '70s. It wiggles and writhes in the 1980s. By 1990, the acceleration of debt is a thing of the past, and the level of debt is rising only about as fast as it did in the 1950s. This is not the story economists tell. It is, however, the story the numbers tell.

Click on graph #3 to see it bigger. Print it out, put a straightedge on it, and see for yourself when debt growth was fast, when it was slow, and when the changes occur. Eyeballing this stuff is better than taking somebody's word for it. And, frankly, it is fun.

Again, #3 shows just the debt, not debt-to-GDP. But the underlying perception is that debt grew slowly before 1980, and fast after. And graph #3 is the one that shows how fast debt is growing. Graph #1 doesn't show debt. It shows a ratio.

Notice that graph #3 shows increase from 1951 to 2021. It is uphill all the way. Here is what I find with my straightedge:

  • Domestic non-financial debt shows a straight run in the 1950s, then turns up a bit more just after the 1958 recession
  • A straight run in the 1960s, with an upturn just after the 1970 recession. 
  • A short, straight run, slightly humped, and an upturn after the 1974 recession.
  • A slightly humped run, and an upturn after the 1982 recession.
  • This is the fastest increase in debt so far, after the 1982 recession, but it starts curving down almost immediately. Then there is a straight run from 1990 to 2000 or say to about 2003. Then there is a small up-jog.
  • The straight run of the 1990s is less steep than that of the 1960s. It is about the same as that of the 1950s but to my eye the 1990s increase is slightly slower: Debt growth was slower in the 1990s than in the 1950s and '60s and '70s!
  • Note that when debt is fairly small (as in the 1950s) a significant increase may pass unnoticed. But when debt is already uncomfortably large (as in the 1990s) the same rate of increase seems massive because the numbers are so big and debt is already problematic.
  • Finally, the straight run between the 2009 recession and the covid year runs uphill at the slowest rate on the graph.

Don't take my word for it. To convince yourself, do it yourself. It's important, because everybody says debt was "stable" before 1980, but it wasn't. Before 1980, debt was growing more rapidly with each new decade. It slowed in the mid-80s, and after the 1980s debt increased at a slower rate.

Granted, we're not looking at debt-to-GDP here. We're just looking at debt. But the growth of debt was faster in the 1960s than in the '50s, and faster in the 1970s than in the '60s. After that was there the rapid increase of the 1980s, for a very short time. Debt growth slowed with the Savings and Loan crisis, beginning in the mid-80s. And then, in the 1990s debt growth was slower than debt growth in the 1950s and '60s.

The "debt was stable before 1980" story must be taken with a grain of salt.

 

The inflation complication

Who tells the story that I tell? Nobody I know. When other people tell the story, it is always non-financial debt relative to GDP. And no matter who tells the story, the story is that debt growth was stable and steady, relative to GDP, until the 1980s -- even though in the 1950s it was not -- and that after 1980 debt growth accelerated:

  • Palley, p.31: "Prior to 1980, the domestic non-financial debt-to-GDP ratio was stable."
  • Mason and Jayadev, p.2: "Between 1950 and 1980, the ratio of total nonfinancial debt to GDP was quite stable around 1.3 ..."
  • Benjamin Friedman, 1981, p.1: "The aggregate outstanding indebtedness of all nonfinancial borrowers in the United States has been approximately $1.40 for each $1.00 of the economy's gross national product, ever since World War II."
  • Cecchetti, Mohanty and Zampolli, p.6: "As Graph 2 shows, the US non-financial debt-to-GDP ratio was steady at around 150% from the early 1950s until the mid-1980s."
  • Bezemer and Hudson: "Growth in credit to the real sector paralleled growth in nominal U.S. GDP from the 1950s to the mid-1980s — that is, until financialization became pervasive."

But no matter who is telling the story, I never find them saying something like this: It might be that debt grew slowly before the 1980s and fast after. But it might be that GDP grew fast before the 1980s and slowly after.

But GDP growth is not the topic of the moment. The topic is the complication that GDP brings into the ratio: The inflation complication: Inflation changes GDP much more than it changes debt. Inflation increases the whole GDP number. But inflation increases only about one-tenth of the debt number, maybe less.

Here is the complication, in two bullet points. Bear with me:

  • Inflation changes prices and values for the current year's transactions. It affects all of GDP because the current year's GDP counts the current year's "final" spending, but does not count the final spending of other years. (Last year's final spending was counted in last year's GDP and was inflated by last year's inflation.)

  • Inflation changes all of GDP, but it changes only a small part of debt. Inflation only affects the current year's purchases on credit. Debt includes the credit purchases from previous years (unless they've been paid off) but inflation doesn't change them. Inflation doesn't change existing debt. (Additions to debt from prior years were inflated when those years were current.)

Here is the important point: Inflation affects all of GDP but only a small part of debt. Therefore, inflation increases the debt number a lot less than it increases the GDP number. This lowers the debt-to-GDP ratio.

Inflation reduces the debt-to-GDP ratio. 


Inflation reduces the debt-to-GDP ratio

The big inflation that I remember is the one called "the Great Inflation" of 1965-1984. If we never had those years of high inflation, the debt-to-GDP ratio for those years would not have been reduced, and graph #1 would have turned out more like this:

Graph #4: Inflation messes with the debt-to-GDP ratio

The debt-to-GDP ratio ran low in the 1960s and '70s because of the long period of severe inflation. Somebody laid an egg.

Graph #5: I want you to picture this whenever you see a graph like #1

Inflation increases the GDP number more than it increases the debt number. Therefore, inflation lowers the debt-to-GDP ratio. This is what we see on Graph #5: not that "debt was stable" in those years, and not that the growth of debt was low, but that the ratio was reduced by inflation.

Can it be that economists are unaware of this? I doubt it. But somehow, they can't even seem to find the increase of the 1950s on their debt-to-GDP graphs. Maybe they should make their graphs taller and not so wide.


Why it looks like debt growth was low before 1980

Graph #3 shows debt rising in the 1950s, rising faster in the 1960s, and rising faster yet in the 1970s. Graph #1 tells a different story.

The increase of the 1950s can be seen in the debt-to-GDP ratio on graph #1, yes. But no comparable increase shows up on that graph in the 1960s and '70s. How can this be?

The answer? Inflation. Inflation reduces the debt-to-GDP ratio. In the 1960s and '70s there was a "Great Inflation" that kept the debt-to-GDP ratio low, and made the ratio look "stable" in the years before 1980.

Funny thing. Everybody knows about the horrific inflation of the 1960s and '70s. But those years of inflation never gets mentioned when people are talking about how wonderfully low and stable debt was in those years.


1980 as a turning point

A big part of Thomas Palley's attention in the PDF is directed to the 1980 start-date of financialization. For Bezemer and Hudson, it is the mid-1980s and "pervasive" financialization. Cecchetti doesn't use the term "financialization", but the data he studies begins in 1980. Mason and Jayadev's paper is a study "in explaining rising debt levels, especially between 1980 and 2000." And Benjamin Friedman followed up on his 1981 paper in 1986, saying "The U.S. economy's nonfinancial debt ratio has risen since 1980 to a level that is extraordinary in comparison with prior historical experience."

Why is 1980 such an important turning point?

Back before 1980, our growing debt wasn't very troubling because inflation was driving income up. But then a wonderful thing happened: Paul Volcker conquered inflation. 

As a result, debt started growing faster than income; faster than GDP. And on the debt-to-GDP graph, you see debt shooting up because of Volcker's success.

 

The mirage

Thomas Palley attributes the sudden rapid growth of debt after 1980 to financialization. But there was no sudden rapid growth of debt after 1980. The sudden growth of debt is an illusion created by the ending of the Great Inflation.

1 comment:

The Arthurian said...

Building on Palley's view that even neoliberalism experienced financialization after 1980, I would argue that the Great Inflation of the 1960s and 1970s taught finance the lesson that its approach in those years was incomplete, and that to prevent inflation and preserve their gains it was necessary to take control of economic policy.