Sunday, July 25, 2010

Krugman, Output Gaps, and Oil Prices

Paul Krugman had a post the other day with the following picture:
Of which he says:
I show an index of eurozone real GDP, from Eurostat, with 2007 fourth quarter — a quarter in which Europe was doing OK, but certainly not experiencing inflationary overheating — set at 100. And I compared it with a trend assuming 2 percent annual growth in potential output, a fairly conservative assumption even for Europe.
So a huge gap has opened up between a reasonable estimate of potential output and actual output. Even if you believe that growth in Europe has picked up since the first quarter — and that the pickup will continue — it will take years to close that gap.

The only way you could justify not doing more to promote growth is to assume that potential output has been drastically reduced by the crisis — and if you believe that, you should be working day and night to reverse that decline.
In a later post, he backed off the 2%, and suggested 1.5% might have been more appropriate for European trend growth, but that's neither here nor there for my point today.

Which is: I have a beef with the "certainly not experiencing inflationary overheating".  It's true of course that the economy was not experiencing broad inflation (eg as measured, for example, by core cpi) either in Europe or here, but commodity prices were shooting through the roof in late 2007/early 2008, and in particular, oil supply was showing extremely severe strain at trying to keep up with global growth.  It's not at all clear that the trajectory shown in Krugman's red dotted line could ever have been achieved.

To look at this more closely, let's go global.  Neither the boom or the crash were primarily European phenomena, but rather due to global trends in debt, leverage, etc.  Repeating Krugman's exercise for the IMF's global GDP numbers, I first constructed this index for world GDP at constant prices, normalized to the last quarter of 2007 being 100.

(I used the WEO Outlook annual series through 2006, and then reconstructed quarters from the data in Figure 1 here for 2007 through 2010 Q1).  The trend from 2000 - 2006 was 4%, and if we then project this out from the 2007 Q4 peak, we get this (just focussing on the period after 2000):

The global output gap is about 5% (how much higher the red line of potential global growth is, than the actual line, by the early part of 2010).

If we now look at global oil production in a similar way (in the interests of time, I took the BP data for global production, set 2007 = 100, and put it on the above graph), we get this:

You can see that, following the 2000 tech crash, oil production was flat for a couple of years, then rapidly accelerated in line with global GDP growth.  However, in 2004, we hit the infamous plateau in global oil production, which then increased very little between 2004 and 2008 (after which it fell in the great recession).

The effect of the plateau in oil production, while global GDP continued to grow, was to cause oil prices to explode to drive the necessary oil-efficiency gains required for demand to meet flat supply.  They more than doubled between 2004 and 2008:

The red dotted line shows the approximate trend of prices, and extrapolated roughly what might have happened had the great recession not occurred.  Note that these are annual average prices - daily prices would have gone much higher due to volatility.  For example, the $97 average price in 2008 conceals a daily price that reached over $140 during the course of the year.  It doesn't seem likely that the global economy would have continued to grow much longer in the face of these kinds of price increases.

In another follow-up post, Krugman suggests that potential output of the economy continues to grow during a recession:
Why does potential growth continue in a recession? Part of the answer is that a lot of capacity growth reflects investments made some time in the past, which come on line only gradually. A larger part of the answer is that the potential output of the economy as a whole reflects more than just business investment; it reflects growth in the working-age population, rising education levels, improving technology, and more, all of which continue even if business investment is depressed. For those who know their aggregate production functions, growth in the stock of physical capital accounts for only a fraction of long-run growth in real GDP, so an investment slowdown doesn’t bring potential growth to a halt.
There may be some continued growth in output potential, but I don't think the oil constraint has continued to improve at the same rate as it would have done.  The drop in prices has caused a drop in the rate of improvement of the oil efficiency of the economy.  If we use the IMF global GDP (in constant prices) and oil production to produce a measure of output/barrel, and then take the year-on-year changes, we get this graph:

After the 1970s oil shocks, the economy was becoming more efficient very rapidly in the early eighties. This dropped off sharply during the '90s. However, efficiency improved strongly again during the 2005-2007 price spike, but has since dropped off the rate of increase.  This translates into real differences in physical capital - there are fewer Priuses and more big SUVs than there would have been had oil prices stayed high or gone higher in late 2008 through the present, as well as more airplanes and fewer video-conference systems.  That ground will have to be made up in the future to further increase oil efficiency.

In short, I don't think it's plausible that the global economy could have continued to grow at the rate it was prior to 2007, and I don't think it's plausible that we could now recover that trajectory, given the lower oil prices in the interim.


Burk Braun said...


Thanks for an interesting post, though I am not sure all the conclusions completely stack up. Increases in oil prices may have helped pop the balloon of the recent economy, but that does not mean that the prior growth was illusory, just that the demand for all the stuff that all the busy bees were making before has dried up. Now they are unemployed, meaning directly that economic activity is that much less than it could be.

There also seems to be a small issue with your definition of inflation, of which there are two species. There are supply shocks, where the price of some commodity like oil goes up. That causes an increase in related prices to work its way through the economy. But once the commodity price plateaus at a new level, no more price rises. True inflation, on the other hand, is a continuous increase in prices, month over month, year over year, based on a fundamental imbalance of demand and supply in an economy where credit is being created and people paid on the expectation of more inflation, and round and round, in a perpetual process.

This could be caused by the government consistently creating more economic demand (by its spending) than it is liquidating by way of taxes. Or it could be caused by bad monetary policy with the "printing" of too much money, (which would be issued by government spending, so they end up being the same process, really). While commodity price shocks can be construed as small scale supply-demand imbalances imposed on the economy in a similar way as a government imbalance, they don't have the same kind of perpetual quality (or danger) to them as more fundamenatally mismanaged macroeconomic policy.

rks said...

Excellently argued as always. Another way to look at it is this. For 200 years we’ve had cheap energy and a shortage of skilled labour, making labour productivity key and driving energy prices to the floor. Now that there is an energy shortage it is energy productivity that is key, and labour costs will be driven down if it goes on for long.

Stuart Staniford said...


I'm not arguing that the prior growth was illusory, just that it couldn't have continued much longer given the seeming constraints on oil supply, and that the trajectory it was on cannot now be recovered (which is what Krugman is saying we should try for). In other words, the recent recession was primarily caused by the bursting of a credit bubble, but if that hadn't happened, growth would instead have been crimped before too long by an oil-shock induced recession.

I agree that there is a useful distinction to be made between general monetary inflation and commodity price shocks. Nonetheless, the latter have a proven ability to cause recessions (particularly when it comes to oil). At the end of the day, the economy runs on physical materials, and if it can't get enough, it either has to come to use them more efficiently or stop growing, and there are limits to how fast the former process occurs.

Michael Cain said...

Absent the oil price part, this is the classic "unit root" problem in macroeconomics. Does output recover to the red line after the recession (no unit root) or is some part of the decline permanent? Opinions based on historical data are mixed: some analysts claim to show a unit root, some don't. As best I can tell from reading him over the years, Krugman is a no-unit-root guy; he believes that output will eventually return to the red trend line, rather than continuing on the parallel blue line.

Myself, I'm inclined to believe that "this time it's different." I like the Ayres-Warr production function, which implies that output growth is driven by (a) the availability of external energy, (b) equipment (capital) that lets us make use of that external energy, and (c) labor. In previous recessions, energy availability was not a binding constraint: the amount of energy use could return to the trend, so output could too. This time, it appears that energy may be a constraint, and not only can we not get back to the red line, but the inability to continue growing available energy means we'll steadily diverge from it as Prof. Krugman's chart suggests.