A while back I pointed to (and demonstrated with not very pretty pictures) Randall Wray’s rather stunning observation: every depression in American history was preceded by a large decline in nominal federal debt.
And I puzzled about why this wasn’t true of our latest little…event:
We saw a decline leading up to 2000, but federal debt was on the rise when the big bang hit. If that 90s decline was the necessary (if not sufficient) cause of the crash, why was there an eight-year delay, unlike all the other depressions in our history?
Various have suggested in various ways what I’ve also presumed: that private debt carried us this time. For a while.
I think this chart may make that point better than any I’ve seen (click for source):
Those earlier depressions weren’t blessed with a mortgage industry engineered to pump newly-created bank cash to anyone who asked through home- and home-equity loans (or corrupt ratings agencies that were the crux enablers of that dynamic). The false GDP from that new private debt issuance — new money flooding the system — floated us through those years. (This is just a variation on what Steve Keen’s been saying all along.)
We’ve been in this woulda-been-a-depression since 2001. We just didn’t know it.
So it seems that Wray’s pattern holds, except — to quote dear Ophelia just before she drowned her sorry self — we wear our rue with a difference.
Cross-posted at Angry Bear.
Comments
22 responses to “Why This Time Is Different”
We never ran chronic trade deficits back in the day, nor could we under the gold standard.
With a 0 CAD, paying down the debt meant draining domestic private net savings— like Mr. Smirnoff would say, in Soviet Russia the debt pays down you!
And when the budget deficit ($162B in 2007) is smaller than our trade deficit ($700B in 2007), even without paying down the debt, we’re draining domestic private net savings.
I suppose that counts as a “financial innovation”.
Hmm, it seems to me that there isn’t really any evidence that fiscal surpluses cause large contractions in output other than the simple fact that surpluses have occasionally coincided with the peak of the business cycle. No one is asking: do other coincident (and more typical) indicators cause business cycles? For example, you could just as easily claim that low unemployment causes contractions in output, since unemployment is always low just before the economy tanks.
Are there any good econometric studies that show that surpluses cause large fluctuations in output? What about studies on the effect of govt spending on real GDP?
Vim, I agree there needs to be more rigorous analysis of this claim, but I don’t think you can just as easily claim low unemployment causes large contractions in output in and of itself. That just seems ridiculous using common sense. However, fiscal surplus being a causal factor does not seem *as* ridiculous using common sense since we know that it is a factual drain of NFA from the non govt sector. This is one plausible channel by which output might decline.
However, I will say that low unemployment does cause the automatic stabilizers to move towards higher taxes and lower welfare – which helps drive the surplus. So the low unemployment might not be causal in and of itself, but it’s impact on the fiscal position could be causal.
Steve, what I would *really* like to do but haven’t had the time is to produce a graph like that brilliant one you had on cross-country house price indexes, but do it for government deficits over the last 10/20 years and compare that to how the financial crisis affects those countries. If memory serves (which it may not do well), Norway has either had small deficits or surpluses for most of the last ten years and is doing just fine. This kind of cross-country comparison (obvs. only looking at ones that have their own currencies) would be extremely interesting on this topic.
This is by way of saying… I’d really like you to make that graph 😀
@Richard Williamson
Hey Richard, thanks.
Give me a little more detail. Scatter plot? Line chart? What on each axis? Spec if for me and it might be pretty easy.
@vimothy “Hmm, it seems to me that there isn’t really any evidence that fiscal surpluses cause large contractions in output”
Wray’s observation seems to me to give pretty good evidence that a fiscal surplus is a necessary condition for a depression. Obviously, there hasn’t been a depression every time there’s been a surplus.
@Asymptosis
Steve, just a chart with government deficits over time for lots of countries would be where I’d start. Depending on what that shows, I’d go from there. I basically want to check Randy Wrays claim across lots of countries. So if we could see (say) the 1920s and 90s-00s that would be pretty interesting. However, I think it could be a real pain in the arse to put together (hence why I haven’t done it myself!). Maybe we can collaborate on it (shoot me an email at the address on my ‘about the blog’ page if you’re interested).
Here’s data from stats.oecd.com, 42 countreis since 1970, deficit as a % of gdp. Excel. OECDs excel files piss me off because they don’t open properly in my Mac Excel (I hate the latest version so have never loaded it), or in google docs. See if your excel will open/save it.
http://www.asymptosis.com/?attachment_id=4893
@Asymptosis
Many knowledgable commentators have gone on record that we are in a de-facto depression now and have been for a while…
Second paulie46.
There is no formal economic definition of “depression.” Rather, it is a historical phenomenon, and recessions are compared to events that have been classified as depressions historically. However, conditions change and the manifestation of the phenomena are therefore different. Here what Wikipedia has to say in the article on depression:
“Considered, by some economists, a rare and extreme form of recession, a depression is characterized by its length, by abnormally large increases in unemployment, falls in the availability of credit– often due to some kind of banking or financial crisis, shrinking output–as buyers dry up and suppliers cut back on production, and investment, large number of bankruptcies–including sovereign debt defaults, significantly reduced amounts of trade and commerce–especially international, as well as highly volatile relative currency value fluctuations–most often due to devaluations. Price deflation, financial crises and bank failures are also common elements of a depression that are not normally a part of a recession.”
Looks to me like a lot of these factors are present not only in the US but also in the UK and EZ. The situation in the US would be a lot worse if safety nets had not been put in place during and after the Great Depression in order to obviate a repeat of that terrible devastation. Same in the UK and EZ. That has worked to a degree to mitigate the disaster that is now occurring. but ask the millions involved how bad it is for them. In addition, we now have social protest rising and in some places violence in the streets. This is not an ordinary recession due to the business cycle. It’s a financial crash at the culmination of a long financial cycle that’s dragged the real economy along with it.
IIRC. MMT economists argue that the Clinton surpluses forced the domestic private sector into deficit in the face of a persistent CAD. The private sector deficit was finance by an explosion of private debt that continued through the Bush administration. While the government fiscal balance expanded in that period it was not enough to offset the effects of the fiscal surplus under the Clinton administration in the face of a persistently large CAD over this period. In addition, leveraged financial “investment” began to substitute for real investment as rent-seeking behavior grew; FIRE began to account for an increasing GDP share, and profit share increased while wages stagnated. Finally, Ponzi finance lead to a Minsky moment when the RE bubble burst.
Wynne Godley predicted the impending crisis in August, 2005, in “Why Gordon’s Golden Rule is now history“
@wh10
According to the MMT economists, this is because the tax rate is too high, so that as the economy recovers taxes drain NFA too much and too quickly. The tax rate is the issue here.
@Richard Williamson
However, I think it could be a real pain in the arse to put together (hence why I haven’t done it myself!).
May already have been done. Ask Randy or one of the other MMT economists.
@Asymptosis
Awesome, it does open on my Excel (2010 all the way… I’m a big fan :P). Shoot me an email (shewingthefly@yahoo.co.uk) and we can discuss before putting anything out there.
Mark Twain: “There are lies, damned lies, and statistics.”
@Euro2012:
Obvious proof that any use of numbers is a lie.
@wh10
“I don’t think you can just as easily claim low unemployment causes large contractions in output in and of itself. That just seems ridiculous using common sense.”
Well, exactly–it is ridiculous. But the logic is just the same. My point is that a bust (or “depression”, however defined) naturally follows a boom. During a boom certain things tend to happen, e.g., tax receipts rise, unemployment falls so transfer payments fall, etc, etc.
So how do you tell what direction the causality is running? Is it the case that a sufficiently large boom can cause the govt’s fiscal position to go positive, or is it the case that a sufficiently large surplus can cause the economy to go from a boom to a bust?
It seems to me that the simple observation made by Wray in Modern Monetary Theory doesn’t really tell you anything very interesting about the relationship between the govt’s fiscal stance and the business cycle.
@Asymptosis
“Wray’s observation seems to me to give pretty good evidence that a fiscal surplus is a necessary condition for a depression.”
Perhaps. Although Wray is only looking at data for a single country, and, as Tom Hickey points out, there’s no agreed upon definition of “depression”, so that term is doing a lot of work in the above sentence.
E.g., I think you could fairly describe the last four years of UK economic performance as a depression, but we weren’t running a surplus in 2008.
Another interesting question to consider is: why should we be interested about the nominal level of debt and not any of the other measures?
For instance, if you look at the two graphs you have at the top of the page, you can see that the real level of public debt is in a slight but fairly continuous decline from the mid 40s to the mid 70s. And no depression. The picture is even starker if you look at the proportionate level of debt, which drops precipitously from the end of the war to the start of the 80s. You’ll note that this period is generally regarded as a golden age of American growth. Strange, no?
@vimothy “E.g., I think you could fairly describe the last four years of UK economic performance as a depression, but we weren’t running a surplus in 2008.”
Can you pull a time series for the UK? (Try http://www.ukpublicspending.co.uk/) I’d be interested in seeing it. Does a similar pattern hold, where surpluses preceded depressions (except this time for some reason there was a lag)?
“why should we be interested about the nominal level of debt and not any of the other measures?”
Who said we weren’t interested in other measures? See: necessary but not sufficient condition.
“real level of public debt is in a slight but fairly continuous decline from the mid 40s to the mid 70s.”
The top graph shows nominal, which is what Wray was discussing; the second graph just tagged along from wikipedia.
“It seems to me that the simple observation made by Wray in Modern Monetary Theory doesn’t really tell you anything very interesting about the relationship between the govt’s fiscal stance and the business cycle.”
I disagree. Just because it doesn’t explain everything doesn’t mean it’s uninteresting, or not useful. Smoking doesn’t of itself explain lung cancer. But it’s certainly “interesting” when discussing the subject.
@Asymptosis
Steve,
Top graph appears to be constant (2010) dollars. Second graph is proportionate level of debt.
In the UK we have run fiscal surpluses from time to time. IIRC, there was one in 2001, one in the early ’80s, etc.
My point is that busts are preceded by booms–that’s what we mean by business cycle. In a boom, the economy does well, which can lead to fiscal surpluses. But this doesn’t mean that fiscal surpluses cause the business cycle, or are necessary for the existence of a business cycle.
Now, you might say that you’re referring to a small subset of cycles, but again, it looks to me like a surplus is just an indicator of the peak of a cycle and that you can redefine “peak” and pick a sample or definition of depression such that it satisfies the condition.
If you believe that surpluses help you to predict depressions and/or recessions, then I’d be interested in seeing some econometric evidence.
By “we” I meant MMTers, and those of us who consume their academic and non-academic output. In my assessment MMT is wholly focused on the analysis of nominal stocks and flows. (YMMV, of course). For instance, I’ve never seen anyone address the question of why a nominal decline is very bad, but real and/or proportionate declines don’t appear to have any noticeable effect. Isn’t that odd? Why should the nominal value of govt liabilities held by the private sector matter but not the absolute level of real debt or its value proportionate to real output?
(I’m not saying that this hasn’t been addressed, BTW–but if it has, it doesn’t feature prominently in any of the papers or posts that I’ve read.)
@vimothy
I feel like you ignored the rest of my comment. As I said, I agree more analysis needs to be done, but to equate the surplus hypothesis with your unemployment hypothesis is taking it too far. The logic is not the same since one can actually postulate a reason why a surplus might cause a downturn (NFA drain). That warrants further research.
“So how do you tell what direction the causality is running? Is it the case that a sufficiently large boom can cause the govt’s fiscal position to go positive, or is it the case that a sufficiently large surplus can cause the economy to go from a boom to a bust?”
I agree we can’t be sure until the data is rigorously analyzed, but I don’t think you’re making the proper dichotomy. Those two can go in sequence. Just because the cycle leads to the fiscal position doesn’t mean the fiscal position cannot feed back into the cycle. See: “However, I will say that low unemployment does cause the automatic stabilizers to move towards higher taxes and lower welfare – which helps drive the surplus. So the low unemployment might not be causal in and of itself, but it’s impact on the fiscal position could be causal.” Meaning, “a sufficiently large boom can cause the govt’s fiscal position to go positive,” and as a result, that “sufficiently large surplus can cause the economy to go from a boom to a bust.”
Furthermore, I think Wray’s hypothesis was specific to significant and quick reductions in the debt ( perhaps “shocks”). I don’t believe slow declines over a long period of time applied.
@wh10: “Furthermore, I think Wray’s hypothesis was specific to significant and quick reductions in the debt ( perhaps “shocksâ€). I don’t believe slow declines over a long period of time applied.”
I actually don’t think he fleshed out his thinking to that extent. Which is why I went and pulled the data. I don’t think the “eyeball analysis” I provided either supports or rejects your suggestion. I couldn’t actually figure out how to do that analysis. Greater minds than I?
@Asymptosis
You’re probably right, but it’s something I inferred from “Since 1776 there have been six periods of substantial budget surpluses and significant reduction of the debt.” I took that to mean ‘significant’ reductions over (relatively) short periods of time.
BTW, however, looking at the data, I found 2 more periods of significant debt reduction. They were from 1805-1812 (45% reduction) and 1844-1846 (33% reduction). I don’t know what the economic climate was like then, but why doesn’t Wray address these time periods?
In any case, I definitely agree with Vimothy that some sort of econometric study is needed before the claim is validated; I just think there is more support for it than the ludicrous claim that low unemployment causes large contractions in output.