Saturday, October 1, 2011

Parsing Waldman (1): Twentieth Century Policy

Liminal Hack points to a post by Ashwin at Macroeconomic Resilience. Part of that post, as I said to Liminal, "is pretty much exactly in line with my thinking."

The "pretty much exactly" part is where Ashwin links to Steve Waldman's Monetary policy for the 21st century. Waldman's post is the object of my attention now and for the next few days.

I printed a copy of Waldman's post and numbered the paragraphs for reference. To avoid being confused by my attention to detail, all you need to know is that Waldman's 'paragraph one' is the one-liner --

Twentieth Century monetary policy can be understood very simply.

-- an opening which contrasts nicely with the title of his post. Paragraphs 2, 3, and 4 provide Waldman's simple understanding, one I'm not sure I can agree with:

"Prior to the 1980s, the marginal unit of CPI was purchased from wages..."

Then came the “Great Moderation”. The signal fact of the Great Moderation was that the marginal unit of CPI was purchased from asset-related wealth and consumer credit rather than from wages.

So Waldman sees a shift from "wages" to "consumer credit" arising with the Great Moderation in the mid-1980s. A shift in policy, he says. Prior to the 1980s, "central bankers had to reduce the supply of wages" to fight inflation. But under the Great Moderation, they could "inflate asset prices and credit availability" or alternatively "[restrain] asset price growth and credit access," rather than manipulating wages.

"Central bankers had to reduce the supply of wages." But in paragraph four, Waldman undermines his own analysis:

I want to emphasize, because it always comes up, that it was not central bankers primarily that suppressed wages during the period.

See, now that I agree with.

If Waldman's analysis is correct, we should be able to see the effects of central bank policy in the numbers.

#1: Suppression of Wages

Graph #1

There is indeed a perceptible decline in wages since around 1980. But this fact is not evidence that the decline was a result of central bank policy. More likely -- as Waldman himself observes -- "Globalization and declining union power did most of that work."

#2: Asset-related wealth

Graph #2: "Assets"
Certainly there was a change in these assets, beginning in the early 1980s. But I don't think this graph shows much that could be called "suppression" since that time.

#3: Consumer Credit

Graph #3: Consumer Credit
Consumer Credit does seem to increase faster after 1980 than before. But there is no definitive change in this picture, that we may associate with the Great Moderation. There is no distinct change from flat to suddenly rising, as there is on Graph #2. I don't see a change associated with the Great Moderation here. I see a long-term upward sweep beginning in the 1950s or earlier.

#4: A Log Look at Consumer Credit

Graph #4: Log Consumer Credit

This graph also shows a significant increase in the growth of consumer credit before 1960. Since then, it shows a pretty consistent trend. If anything, there is a slight lessening of consumer credit growth since the 1990 recession. But there is no indication of slow consumer credit growth before the 1980s. And there is no hint of a speed-up related to the Great Moderation.

These four graphs do not show that the central bank suppressed wages. They do not show that anyone suppressed assets. And they do not show an acceleration of consumer credit that is associated with the Great Moderation.

On the other hand, the constant, continuous, critical increase in consumer credit-use does stand out. An increase began immediately after World War Two and continued without relief to the crisis of 2008. An increase that was itself part of the underlying cause of that crisis. An increase of credit-use. An increase of debt.

Changes in the '80s? Of course there were changes in the '80s, because there were problems in the '70s. Remember stagflation? Remember double-digit inflation? Those problems were caused by increasing credit-use. And by the rising factor-cost of money, which was a result of the excessive accumulation of debt (yes, already in the 1970s) which was itself caused by the constant, continuous, critical increase in consumer credit, and by other uses of credit.

There were changes in the '80s, like special tax treatment for capital gains. A shift away from a focus on profit, toward the focus on wealth accumulation. Changes that can be seen in Graph #2. Changes put in place in the 1980s, in response to prior problems.

Did the Fed change its behavior? Of course. It did what it had to do, to make policy effective. But it did not change its support for credit-use and the growth of debt. Rather, it continued to appease the need for credit-growth -- a need created and enhanced by Congress.


LiminalHack said...

Art, ashwins article positions the use of transfers as a temporary measure to prevent collapse, not as a long term policy. Check the comments section to verify that.

His subsequent comments concur with my analysis which is to say that for a reslient economy what needs to happen is the people be weaned off the notion of 'risk free' assets.

Since "money", specifically base money, is a nominal risk free asset, his thesis, like mine suggests we should have less of it, not more.

That's not to say transfers (aka social credit) is not compatible in theory with our position. It just means that citizens (and certainly not limited liability corps) wouldn't be eligible for deposit insurance in large amounts.

Are you familiar with CH Douglas and his "social credit" theory? I imagine you must be given your dislike of debt...

The Arthurian said...

In the 60 years between 1947 and 2007 our economy transformed from low reliance on credit to extremely high reliance on credit. I think people tend not to prefer an economy they have not experienced.

Because they came along later than I, younger people are less willing to accept a low reliance on credit, and instead come up with notions like "for a reslient economy what needs to happen is the people be weaned off the notion of 'risk free' assets."

I am not interested in transfers. I am interested in a balance between credit-in-circulation and non-credit-money-in-circulation, and the relation of this balance to economic performance.

I am not familiar with Mr. Douglas. I am familiar with the Debt-per-Dollar graph. I do not try to read books. I try to read the economy.

Nobody gets that.

jim said...

Hi Art,

I always thought this commercial says it all:


Steve Waldman said...


First, thanks for taking such a careful look.

A few things:

1) I think you can see a regime change in the data w.r.t. the credit-intensity of new purchasing power. Interestingly, the regime change looks to have started in the mid-1970s rather than in the 1980s.

Try this graph:

It shows the change in household indebtedness (including both consumer & mortgage debt) divided by the change in household indebtedness plus wages. The intuition is fairly straightforward: An increase in indebtedness generates a positive cash flow to households, just as wages do. In a wage-centric economy, most consumer purchasing power would come from wages. In a credit-centric economy, wages represent a smaller fraction of annual cash flows to households.

Through the late 1976, the new debt component of total purchasing power ranged between between 2 and 8%. From 1977 through 2007, the new debt component was above that previous maximum in 20 out of 31 years.

I think these graphs understate the change, as increasing wage inequality over the period means an increasing share of wages saved (which bids up financial asset pries rather than consumer prices).

2) It is not my view that central banks engineered this shift. I think they fell into the new regime, and then backstopped it. I don't claim, or mean to claim, that this was a shift in policy (although I think central bankers took credit for it as though it was a shift in policy). The economy simply became more amenable to management via interest rate policy without disruptive and politically difficult recessions. I do think central bankers understood that this happy new circumstance was related to the decline in the wage intensity of GDP (fell through its traditional range at about the same time as the credit uptick began, ), and which why monetary policy during the "great moderation" was ostentatious allergic to any rises in unit labor costs (or equivalently, increases in the wage intensity of GDP). As you note, central banks did not _cause_ this decline - the decline of organized labor, globalization, and technology played larger roles. But nor was central bank behavior irrelevant. The fact that any rise in unit labor costs was sure to be met with tight monetary policy is likely to have conditioned the behavior of policymakers and labor activists. In Bill Clinton's America, pushing policies that would shift GDP-share towards labor meant inviting the tight money that GHW Bush blamed for his defeat. That might have affected all kinds of policy calculations (e.g. NAFTA, support of organized labor, etc.)

Thanks again. I'm really flattered that you are taking the time to read my effluent so closely.

The Arthurian said...

Hi, jim. Oh, I remember that Visa ad. A good one. It seems something like that happened to our economy, three or four years back. The sudden stop.

I also remember an automobile ad, Honda I think, the businessman walks up to the window and looks at the car, and suddenly his business suit morfs into a lumberjack outfit and he springs a beard. And presumably, he walks away from the high-pressure, high-velocity-of-money rat-race life. Sort of like that PIMCO guy did, come to think of it.

The Arthurian said...

Steve Waldman. Thanks for the visit, sir. Actually, I've got five posts lined up, four on yours and one on the comments that follow it. In my posts, I disagree with your analysis of the past, but strongly agree with your proposal for the future. Or, with what I see as your proposal for the future, which seems to be different from what a lot of other people see. (I would be interested to know whether you think i have *that* right). Your proposal is *HUGE*.

RE your point #2: "It is not my view that central banks engineered this shift."

I can see that. In mine of 3 October I say as much. I think you over-simplified your summary of 20th century policy, much as I have over-simplified my criticism of it. And I think we both did it to make important ideas stand out.

RE your point #1: "a regime change in the data"

You write "Interestingly, the regime change looks to have started in the mid-1970s rather than in the 1980s."

Short response: Yeah, but keep looking. It goes further back.

Long response: I'm a sucker for graphs.

So, this "regime change" does not start with the Great Moderation, but before it? Agreed!

In the mid-1970s came the end of the "golden age," and the recession of 1974 which changed everything. Income grew slower after that; surely this affects your graph. (My argument would be that the accumulation of debt before 1974 is the reason everything changed. The camel's back broke in 1974. And of course, the growth of debt in the 1950s and '60s was the source of the golden performance. It's payback. Ouroboros. Yin-yang.) What you see as "regime change" seems to me to be the result of prior debt growth, coming home to roost.

Trend lines through peaks on your graph might put the turning point at 1969, even.

I certainly do agree that "credit intensity" (good phrase!) has increased. That is the whole focus of my work. I do dispute that increasing credit intensity began in 1969 or the 1970s.

Looking at your data a different way seems to show that the trend-change occurred after the near-recession of 1966-67.

And looking at the same data even more simply -- simpler arithmetic, I mean -- clearly shows an increase in consumer debt (relative to wages and salaries) that begins with the earliest FRED data, around 1952.

The fact that the slope of the exponential growth curve was quite flat in the early years is not evidence that the trend of debt growth was not a problem. Debt itself was not a problem until some years later. The trend was a problem from its inception.

Steve Waldman said...

I've noticed you like graphs!

I certainly agree with you that the change in trend in credit intensity occurred prior to what we now call the great moderation. How you want to date the regime change (and whether you want to impute a sharp or gentle transition) is pretty tricky. My intuition is that, starting in the late 1960s and 1970s (when we began to have banking crises again after a 20 year lull), credit intensity picked up while expectations of both inflation and wage growth remained high. This helped create the inflationary 1970s, as credit + expected wages +expect devaluation + expected cheap repayment all support current spending.

The great moderation began when expectations of inflation and wage growth were blunted, and that was the joint work of Volcker and Reagan. With inflation "tamed" in expectation (and interest rates deregulated, so real interest rates could be made high), the Fed found traction. The 1970s combination of loosening credit, low real interest rates, and expected wage growth were gone, people could be induced to save, and induced to spend, by interest rate policy.

So perhaps we have three regimes: postwar stability, loose credit without policy traction, and great moderation.

I'm having a bit of a hard time with your graphs. Your first new graph, from the image, looks like it's graphing the same data I've graphed, just as a ratio of debt growth to wages rather than as a fraction of debt growth to (debt growth + wages). But your graph seems to be telling a substantively different story than mine, even though it should be a monotonic transformation of the data. When I graph what I think you are graphing, I get something different than you did (and more like my original graph). Here's a link to the FRED graph page, rather than the image. Do you see what we're doing differently? Your second graph I have a hard time interpreting, because I'd say it mixes a stock and a flow. Wages and change in debt represent a flow of new purchasing power, but the stock debt doesn't map to that.

My sense is the late 1960s is the earliest you'd date the increase in credit intensity, and that it only becomes a plain feature of the data in the mid to late 1970s.

But you might be able to tease out a better story!

The Arthurian said...

Steve, my first new graph there is the "percent change from previous year" transformation of my second new graph. So, both of them mix stock and flow... Both of 'em divide a level by a rate.

If the water level in the reservoir is [low or high] and the rate of water usage is [low or high]... There might be some use for dividing a stock by a flow.

Lenders are said to look at a potential borrower's income and existing debt, while deciding whether to lend more to that person. The lender's concern is therefore the relation between a stock and a flow.

Also, anyone who has expressed concern about the level of debt relative to GDP is also looking at a stock/flow mix. But I am not among those people. I prefer to look at total debt relative to M1 money. (The Historical Statistics gets me back to 1916.) This is apparently a stock/stock relation.

I would bet my Debt-per-Dollar graph is a better measure of "credit intensity" than your flow/flow graph.

Calgacus said...

LiminalHack:my analysis which is to say that for a resilient economy what needs to happen is the people be weaned off the notion of 'risk free' assets.
Since "money", specifically base money, is a nominal risk free asset, his thesis, like mine suggests we should have less of it, not more.
Why on earth would one want that? I want the dollar in my pocket or my bank account to be basically risk-free. If someone has a dollar addiction & hoards a billion - take it away the old-fashioned way - taxation or other forms of state control (for financial institutions).

Arthur is absolutely right, and in accordance with the entirety of human history, that mo' money (and monetary stability) & less, controlled, credit is a Good Thing.

The Arthurian said...

Hey Cal. Thanks for the thumbs up. And I like the link to history.

But I am surprised to see you distinguish between "money" and "credit" in your remarks. ??

Calgacus said...

I was trying to not be pedantic. Money is a form of credit/debt, but all credit/debt is not the same. The basic difference is who owes/issues the debt.

Just agreeing that it is better if people have lots of state credit/debt = base money, government bonds, evenly distributed, in relation to the amount of debt they have to each other - so if their businesses go broke, they can pay each other off with state money, and you don't get chain reactions of bankruptcies. Control credit, so control inflation. Short, shallow recessions, good growth.

That's what the post-war world was like. Lots of risk-free assets. High public debt/GDP (OK maybe not 5000%) is/was a good thing.