Chart of the day: Growth and debt

By Felix Salmon
March 16, 2012
Greg Ip has a fantastic blog post on the subject of America's GDP growth and the potential thereof.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Greg Ip has a fantastic blog post on the subject of America’s GDP growth and the potential thereof. He’s talking about this chart:


The blue line, here, is actual US GDP. The green line is what’s known as “Potential Gross Domestic Product” — the amount of output that the Congressional Budget Office reckons that the US could produce, if it were running at full capacity. (Don’t be confused by the weird formula: potential GDP is released in real 2005 dollars, so I’ve multiplied that data series by the GDP deflator to convert it to nominal dollars. You’ll see why in a minute.)

The main worry in this chart, of course, is the fact that the blue line — actual GDP — is so far below the green line, which is where by rights we should be. Up until the Great Recession, the two series tracked each other very closely. Now, however, they diverge by some $890 billion. That’s $7,800 per household, per year.

Greg’s point is that the green line might well overstated: that the economy can’t in fact grow, sustainably, at the kind of pace that the CBO is assuming it can. As he puts it: “both the level and growth rate of American potential output is much lower than we think”.

I think this theory is entirely plausible. And to demonstrate why, I’m going to take the exact same graph above, but add one more data series to it — this time the amount of credit outstanding in America.


There’s a whole narrative in this chart. From 1970 through the beginning of the crisis in 2008, GDP grew at a pretty steady pace. But the amount of debt required to generate that output just got bigger and bigger — the rate of growth of the credit market was much faster than the rate of growth of GDP. In 1970, GDP was $1 trillion while the credit market was $1.6 trillion: a ratio of 1.6 to 1. By 2000, when GDP reached $10 trillion, the credit market had grown to $28.1 trillion: a ratio of 2.8 to 1. And by mid-2008, when GDP was $14.4 trillion, the credit market was $53.6 trillion. That’s a ratio of 3.7 to 1.

In other words, in order to keep up a steady rate of GDP growth, we had to saddle ourselves with ever more cheap and dangerous debt.

Then, suddenly, the growth of the credit markets screeched to a halt, and we had a major recession. And since then, the size of the credit market has been roughly flat.

It makes sense that if we needed ever-increasing amounts of debt to keep up that long-term GDP growth rate, then when the growth of the debt market stops, our potential growth rate might fall significantly.

I’m glad that we’ve finally put an end to the credit bubble, which had to burst at some point. But it’s naive to think that we can do so without any adverse effects on broad economic activity. So we might indeed have to resign ourselves to lower potential growth going fowards. If only because we’re taking ourselves off the artificial stimulant of ever-accelerating credit.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

You seem to look at things from a consumption point-of-view, much like a household. “Of course, we’d be able to buy more stuff if we could take on more debt.”

But from a production point of view (which I think is consistent with the idea of “potential output”), that’s not as clear. If the U.S. has a fixed base of physical capital and labor, why would more or less debt enable that capital and labor to produce more?

Or are you making a more subtle argument that more debt led to a greater stock of physical capital and labor? But if so, why would that stock have disappeared?

I don’t really understand the logic you’re laying out.

Posted by Beer_numbers | Report as abusive

As a supporter of NGDP targeting myself, I find the second chart very interesting.

One thing economists undervalue as a cause of both the 1920′s stock market bubble and the credit bubble of the 00′s is income inequality. Whether increased income inequality is a good or bad thing, if the rich take home a higher share of the GDP, the propensity to save a dollar of GDP goes up.

If markets were always efficient, the increased propensity to save wouldn’t lead to unsustainable credit bubbles because no rich person would lend to somebody else if they wouldn’t get paid back. With efficient markets, the propensity to save gets offset with lower interest rates until more money gets spent on consumption now instead of earning low interest rates for the future.

However, deregulated financial markets somewhat predictably find themselves producing asset bubbles and, if the rich’s propensity to consume all their wealth is extremely low, then GDP growth may come from investment with negative NPV instead of only consumption, positive NPV investment and government spending.

For an economy to try to get NGDP growth back on trend merely by pulling the same lever, i.e. by increasing investment only and not consumption, then disillusionment by rich investors from past asset bubbles will prevent further negative NPV investment as a driving force. Furthermore, there may be no more investments with positive NPV’s at zero risk-free rates plus the investment’s risk premium.

That’s where Paul Krugman typically comes in and says that Keynesian spending is the only reliable way to go, since that’s the only way to increase positive NPV investment at the ZLB (positive NPV for the Treasury buyer today, if not for the future taxpayers).

But there’s another lever: increasing consumption. If the Central Bank sets expectations that things tomorrow will be more expensive than today, no matter what, than at least some money-holders will increase consumption today until that becomes a self-fulfilling prophecy. The increased consumption will furthermore generate more positive NPV investment opportunities, as the huge boom in business investment 1933-36 shows after FDR adopted price level targeting.

In short, the fact that increased income inequality and financial deregulation can create asset bubbles does not mean that we have to accept that the Central Bank is powerless at the zero lower bound. As much as those distortions, inequality and deregulation, are issues, they have solutions through the tax code and regulation, not through monetary policy. Their issues still take the same bite out of RGDP regardless of central bank policy. With no structural (not political) bound to how much the Central Bank can inflate a currency, there is no reason for the Central Bank to limit itself to targeting short-term interest rates.

Posted by mwwaters | Report as abusive


It’s the propensity to consume vs. the propensity to save as I argue above.

America does have a fixed capital and labor stock to produce what’s at the green line. What has happened in the past though is that the consumption has happened by those who can’t afford it because those who could afford it direly wanted to lend instead of consume.

In short, for production to recover back to potential levels, those who can afford stuff need to start actually buying stuff. This isn’t about exporting to China, but to start exporting to all those people who have built up wealth and decided to save. Essentially we have a current account crisis, but with Americans owing Americans instead of nations owing other nations.

Also, as I said, lower consumption also, paradoxically, drive investment lower since any investments needs future consumption to have a positive NPV. If the Central Bank credibly targeted *levels* (not rates) of either prices or NGDP, then both private consumption and investment would go up in tandem.

There is no structural, economic reason for the Fed not to target levels. There is certainly a political one, with the current members of the FOMC and Congress, but so often commentators try to make that into a structural, irreversible issue. No, it’s purely a political issue that can be easily reversed once the politics change.

Posted by mwwaters | Report as abusive

youre again hitching your wagon to a financial constraint, which is largely a fiction…

credit is a barbaric relic…

Posted by rjs0 | Report as abusive

It’s a great topic, but it needs to be pushed a bit further.

Of course you need to look at debt to GDP across the whole economy (govt, domestic non-financial, financial), but you also need to look at servicing costs.

Debt levels in the 70s and 80s were impeded by high interest rates, so you could say that debt levels would have been higher if rates were lower. Of course there were no toxic debt products back then, but the long-term trend of declining interest rates since the early 1980s is a huge part of this equation.

Higher debt to GDP levels need to be considered in the context of overall asset values and the % of disposable income required to sustain it.

I’m not saying leverage isn’t a problem, it is, but these ratios don’t tell the whole story.

Posted by David93 | Report as abusive

Not the crux of your argument, but neither graph is log scale, so in regard to saying the output gap is currently bigger than it ever has been is a bit inaccurate. Maybe nomiminally it is, but I’d be curious how high, percentage wise, the gap was in the early 80′s compared to now.

Posted by timothydh | Report as abusive

It makes for a nice story, but correlation does not prove causation.

It may well be that some 3rd factor (e.g. great stagnation in innovation, regulation, or whatever else rocks your boat) explains a drop in potential GDP. This in turns leads to downward reassessment of future expected disposable income by households who take on less credit than they used to. That’s a basic permanent income story that does not involve any causal effect of debt on potential GDP.

Posted by Turpentine | Report as abusive

I am sympathetic to your story, but simply comparing a stock (credit outstanding) and a flow (GDP) might not be the best way forward.

A more reasonable exercise is to look at changes in nominal debt versus changes in nominal GDP. Over the period from 1949 to 2000, the role of debt in the path of real GDP, whether in levels or per capita terms, is sizeable, but not overwhelming (contrary to my own expectations).

The story from 2000, however, is very different: observed real GDP per capita is about 8% higher than it would be without net additions to outstanding consumer and government debt.

(Wish there was some way I could post the figures for this here…)

Posted by dsjacks | Report as abusive

There’s an ambiguity in measuring the concept of output — actual or potential. This is often ignored by those who believe that there is such as thing as an output “gap,” and that this gap is measured by the distance between the slope of two curves that diverge from the moment when an asset bubble bursts.

In one sense output is obviously measured in physical capacities — factories, machines, fields cleared, products produced. For the moment we can say that this capacity remains the same, no matter how completely it is utilized.

But how to cumulatively measure the sum of all these various capacities — acres, volumes, weights, etc.? The common measure is what they are worth, and this gets us GDP. But in monetizing in order to cumulate, we introduce the notion of value. And that is what can change, suddenly. If those same fields, factories, etc. are worth only half of what they were worth in the bubble, then GDP has fallen by half.

That’s why GDP should be considered as a flow. So, what kinds of flows produced that level of value of GDP? As Felix’s interesting graph shows, it took lots (and lots, and lots) of credit. Above and beyond income, it took credit to fuel the consumption that built factories, made products, etc. that produced GDP.

So the upper curve that is called “potential output” is really the level of output that could be achieved at that time, given that utilization of resources, technology, labor, etc. And just as important: given that distribution of income, level of wealth, and…amount of credit. There is always a gap between actual and potential output. The real “potential output” of any economy isn’t a curve on a graph; it’s the application of all the energy (labor, technology) to all the resources of the economy. Not infinite, but some very, very high value.

Could the pre-bubble level of GDP be recreated? In the absence of continuing private sector credit, only the government could replace all the lost debt and the reduced income and wealth that results from de-leveraging and unemployment in order to reach the same GDP. But the government wouldn’t have to stop there. If it succeeded in maintaining that level of utilization, it could go even further (unless there were scarcities that produced bottlenecks, and in a richly endowed economy it could probably bypass them as long as there was unemployment).

And if the government did try to recreate the output level indicated by the top curve, it would still have to use — or destroy or give away — the surplus products produced, because lots of people will still be de-leveraging and won’t be able to afford them without new debt. Take World War 2 as the example: full employment, government producing war materials, buying it up, GDP soaring (and limiting consumption, so people de-leverage) — and having the “surplus” (planes, guns, manpower) destroyed and have to be built all over again.

Anyone think this problem is going to be solved soon?

Posted by jbernar | Report as abusive

“Could the pre-bubble level of GDP be recreated? In the absence of continuing private sector credit, only the government could replace all the lost debt and the reduced income and wealth that results from de-leveraging and unemployment in order to reach the same GDP. But the government wouldn’t have to stop there. If it succeeded in maintaining that level of utilization, it could go even further (unless there were scarcities that produced bottlenecks, and in a richly endowed economy it could probably bypass them as long as there was unemployment).”

This is exactly the argument my posts above tried to refute.

There seems to be this view of a credit bubble where the credit came down from the sky and once the credit stops, that source of money for GDP has been vanquished forever.

No, the credit bubble was tightly related to the “global savings glut,” which in turn was a result of greatly increased economic inequality. Inequality was particularly bad if you looked at wealth. For every American deeply in debt, there has to be an American* “deeply in assets.” That’s simple double-entry accounting.

* And yes, that would almost certainly be an American on the other side of the transaction. The net capital flows across the borders were only a small part of the increase in credit. Nearly all of it was Americans lending to Americans.

The conventional wisdom is that the deleveraging follows after a credit bubble. The debtors scrounge up cash and debit their debt on the right side of the balance sheet to somehow become whole. Then the creditors just pocket the cash or invest it in Treasuries, since even at zero interest rates, there is very little loan demand from creditworthy private borrowers. The creditors still have the same wealth but they are keeping cash instead of buying stuff.

In fact, all the deleveraging seen in the graph above has gone into exploding excess reserves at the Fed. But the Fed has also paradoxically been paying interest at reserves, starting at the worst possible time. The government is paying those who have money *not to spend it.* Why everybody doesn’t see this for how utterly asinine it is is beyond me.

The Fed has no reason to even agree to keep excess reserves on its accounts at zero percent. It can charge penalty rates, like any other bank does through banking fees, on cash kept on its accounts by member banks. Since only the Fed has the authority to convert electronic deposits to newly printed money, the only alternative for those wishing to hold cash is to hold literal cash in vaults. That has risks and costs and the Fed can even charge fees or limit how much cash it prints as people convert excess reserves to physical notes.

There are further nuances here, but I’ve gone on too long already. I hope you can see how horribly, horribly misguided this conventional wisdom is that investment is the only lever. Should the central bank, any central bank, really want to increase the “flow” of GDP, there is no structural or economic boundaries, only arbitrary political ones.

Posted by mwwaters | Report as abusive

The *inverse* is what makes sense:

First, there is no correlation to interest rates (go back to the 50s) or the velocity of money, or money base, its not a credit phenomenon.

Second, you are comparing the pile of bricks to the flow of bricks being built. In an economic sense the pile is capital so the appropriate metric is the interest rate on the pile vs. current income. Or, GDP/TCMDO {because you want nominal to nominal actual comparison}. Technically you want to remoce financial debt because both the assets and liabilities of banks are debt, so you are double counting.

GDP/TCMDO is about .3
GDP/TODNS is about .4

So if the “interest rate” on the pile is 4% we the “debt service” is 10%-13% of GDP. Yawn.

Maybe that big pile of debt represents investment in something like, you know, capital that the baby boomers will need to retire (houses, machines to build cars, oil wells…). .png?g=5L1

Posted by dwb3 | Report as abusive

I have to admit I’m with the other commentators who are puzzled by this line of reasoning. Output is measurement of the physical capacity of a country to produce goods and services.

Now, you can argue that output capacity has degraded. You can argue that a declining population has shrunk the potential market. You could argue that technological shifts have rendered portions of the market redundant.

But “debt” doesn’t do anything to output. Debt is an allocation of capital. You can say that without debt the consumer class isn’t broad enough – but again, that’s a capital issue, not an argument about output. The answer isn’t “it appears we should just give up on GDP trend” but rather, how can we allocate capital more efficiently to maintain a level of output our country is perfectly capable of producing.

Posted by strawman | Report as abusive

Breathtaking really – the number of posters on this thread who look at that red line on chart 2 and dismiss it as meaningless.

I’m with the author – it’s frightening. Somebody has to service it, and somebody has to pay it off – or somebody has to write it off. As one poster mentioned, it’s Americans who lent the bulk of it, and who stand to be stiffed if write-offs happen. Wonder what that will do to GDP growth.

Posted by MrRFox | Report as abusive

MrRFox, wouldn’t write-offs tend to *improve* GDP growth?

Consider mortgage principal reductions, for example? The usual argument runs: lower debt –> lower payments –> more money left for spending.

I agree that the rising debt burden is unhealthy for the economy, but write-offs take that red curve in the opposite direction.

Posted by TFF | Report as abusive

Oh God, Felix, you’ve fallen for the old stock versus flow comparison, as somebody else has pointed out earlier. If we assume that some debt is necessary in all economies, the question becomes what is prudent in the aggregate, so why don’t you take the PV of each years increment in income and call that the borrow-able base, then take the delta on debt and see what kind of aggregate “Loan-to-Value” ratio you get. I think you will be surprised by how low it comes out.

This is the same canard that every hedge fund guy who got his greek short right keeps inserting into the public discourse as a metaphor for what is wrong with America. Instead, focus on the cretins who live in the Reality Distortion Field that is the modern Republican party, and the inability of the US to reach escape velocity since the Great Recession becomes abundantly clear. Pointing to debt accumulation of yore is a total red herring.

Posted by Hookahboy | Report as abusive

@ TFF, 17th, 2:23pm

I don’t see write-offs as a positive for GDP growth, unless it’s part of an overall restructuring package, and even then it’s dubious, IMO.

Each dollar written-off brings the borrower’s debt down, but he’s still always “broke”. His purchasing power in cash terms doesn’t improve at all, though his new debt-free future is perhaps brighter than his debt-burdened past.

The creditor, on the other hand, takes an immediate hit to his net worth, dollar-for-dollar with what the borrower receives in forgiveness. That’s probably likely to put a crimp in the lender’s consumption spending, isn’t it? Or, is the “wealth effect” just an illusion?

Posted by MrRFox | Report as abusive

@MrRFox, I thought that restructuring the cash flows was supposed to be a key part of principal write downs? After all, if the principal has been written down, then you either make lower payments or pay off the loan more rapidly.

I’m not sure I believe in the “wealth effect”, especially when the wealth is concentrated in so few pockets.

Posted by TFF | Report as abusive

@TFF – Um … well, you’re half-right, I suppose. A restructured loan may well have lower payments than the original, and that benefits the borrower. But what about the lender? – every dollar of reduced d/s comes right of his pocket, doesn’t it? Sounds like “zero sum” to me.

Of course, someone else could make-up the lender’s loss, so nobody loses anything, right? Wonder who that Tooth Fairy might be? Got any suggestions?

Here’s a wild idea – find the most indebted single individual in the country, and have the Treasury print up enough $100-bills to cover his entire net debt – and give them to him; print enough to give every other man, woman and child in the country each the exact same amount. Now everyone is solvent and everything is perfect and fair – right? Krugman would love it, wouldn’t he?

Posted by MrRFox | Report as abusive

@MrRFox, I agree that it is ultimately zero sum on the national balance sheet, however it is very different from being a net zero for GDP.

If the households with the reduced payments are more likely to spend that marginal dollar than the households that eat the writeoff, then the net effect on GDP is positive. That’s the theory, at least. I see a couple flaws in the theory, personally, but you haven’t been pointing them out so I won’t drag them into the discussion.

I wasn’t proposing “making the lender whole”, nor was I proposing blowing up the currency. Not sure where you came up with that one? For that matter I wasn’t even proposing this as a GOOD solution. Like I said, I see flaws that seem to have escaped you (or you wouldn’t be making stuff up).

Posted by TFF | Report as abusive

@TTF – I see some of those flaws that you alluded to as well, and I surely don’t buy the idea that spenders are “good” and savers are “bad”. If Danny Black called himself “old fashioned”, I can too – I think saving is a virtue; every right-thinking, non-Keynseian should too.

Didn’t mean to suggest that money-printing foolishness was your proposal – but it kind of is the Krugman/Keynes attitude, isn’t it? Truth be told, inflating our way out of this debt-trap is the only way out that I can see. You?

Posted by MrRFox | Report as abusive

We save as much (proportionately) as anybody, so you know where I stand on THAT.

I agree that inflation is the only way out — but if you announce that you will be aiming for higher inflation for a decade, then interest rates adjust and the massive present debt load kills you.

The trick is running a moderate rate of inflation while publicly arguing a lower rate. An insidious policy, if that is actually what they are thinking.

Whatever the “solution”, it is essential that people not lose faith in the house of cards.

Posted by TFF | Report as abusive

“Whatever the “solution”, it is essential that people not lose faith in the house of cards.” (TFF)

I love your tag line here ^ ^, TFF – it’s perfect. Everything will be OK so long as everyone (except you and me, TFF) continues to believe in the financial Tooth Fairy and the economic Easter Bunny that our economic rabbis preach to us.

Posted by MrRFox | Report as abusive

MrRFox, I don’t exempt myself… I, too, believe in the Easter Bernanke. After all, what choice do I really have?

I do make some effort to risk-proof my portfolio, and my financial picture in general, but if things start to topple I’ll lose plenty along with everybody else.

Posted by TFF | Report as abusive

Bull’s-eye again, TTF. I too can’t see a plausible way to escape the train wreck, if that happens – and I live on the other side of the world. Even here, and no matter what one does, it stands to be a lifestyle-changer if things come unwound in the US.

Hate to say it, but maybe the “money printers” have the least bad of the solutions (all of them ugly) on tap.

Posted by MrRFox | Report as abusive

Felix, I find it amazing that this topic can be bandied without referring to Steve Keen’s work showing the correlations between debt levels and house prices, unemployment, and stock prices.

Actually, a second derivative: the rate of change in the rate of change in debt — what he calls the credit impulse or the credit accelerator. He’s demonstrated strong correlations in the U.S., the U.K, and Australia: teve-keen-dude-where%E2%80%99s-my-recove ry.html

Posted by ReutersRat | Report as abusive

GREAT chart — we have been on an unsustainable OVER-employment “trend” since … Nixon took the US totally off the gold standard.

No “good” way out, but the first point is to get balanced gov’t budget. Freeze gov’t spending, and taxes, until balance.
Then print enough cash to retire all gov’t debt — stop borrowing. Let the investors get stuck with too much capital looking for places to invest. And as interest rates rise, the lack of gov’t debt will be fine.

A bit of rough ride, no matter what, but then more stable growth that is sustainable.

Posted by TomGrey | Report as abusive

Your basic intuition is betraying your reasoning. The point you are making in this post is profoundly wrong for two main reasons. And I have seen recently such kind of argument so many times, that I am becoming paranoid about it. So here is my contribution.
Firstly, as one of the earlier comments highlight it is not clear that a reduction in the pace of debt will lead to a lower capital stock, lower productivity, to lower technological knowledge, or to a lower labour supply. What is the connection between the two sides of your reasoning: debt and potential output? I don’t see any: prices, wages, interest rates may change, but those REAL factors remain the same. I suspect that your problem is related to the second point I am going to highlight next.
You may be confusing the impact of debt with the impact of some other aggregate. The central bank and the banking sector can create money out of the blue. But no one can create debt out of the blue.
For the sake of simplicity, imagine a closed economy (the argument is easy to develop this way).There is a certain level of savings which end up having a certain monetary value. If the level of savings increase (and Price level constant), the monetary value of savings goes up as well. The point is that the level of savings may remain constant, but (apparently) de amount of debt may go up. Why “apparently”? Because in this economy (a closed economy), the level of all debt (negative savings), has to be TOTALLY EQUAL to all loans (positive savings).
So for this economy as a whole, it is as if net liabilities (debt) are zero. In the end, the LEVEL of debt is irrelevant, but not the way debt is USED for. So, if I borrow money and just throw it down the tube, I will not be able to pay it back. But this is another problem which has nothing to do with the level of debt itself, but rather with the efficiency of allocating those loans to specific aims.
So unless, we have someone coming from Mars or any other planet, borrow and disappear, or we have a large proportion of debt badly used (which I am not ruling out, for obvious known reasons), we do not have (as a matter of fact) any problem with the fact that the LEVEL of debt may grow faster than GDP. It is one major mark of the very history of modern capitalism. There is plenty of evidence showing that financial intermediation has grown faster than GDP. And in the world of economic theory there are already convincing arguments showing that if the costs of intermediation go down for a while, yes, we should expect intermediation to grow faster than GDP. There is an interesting recent paper by Edward Prescott and associates on this issue.
Do we remember what happened in the 1990′s with a similar story about the “Paper Tigers”? A huge amount of people were sentencing them to oblivion. It was a fashion that proved wrong.

Posted by Viriat | Report as abusive