Arnold Kling  

Potential Output Straw Man

An Early Stanley Fish Moment... Murray's WSJ Solutions ...

Recently, bloggers have been talking about potential output or potential GDP. I find the discussion to be frequently misleading. For example, Mark Thoma quotes Tim Duy:

If we claim the economic potential of the nation has declined - that in aggregate, we can not make as much stuff as we did a few years ago...

But nobody claims that. The pessimists on potential output only claim that it is growing below some previous trend. They never claim that it has declined in absolute terms.

I do not blame Mark or Tim for making misleading statements on this topic. I think there is something inherently misleading about the concept of potential GDP. We talk about it as if we were living in a one-good world; however, in the real world there is heterogeneity of workers and and goods.

(PSST is most explicit about heterogeneity. The "representative agent" DSGE model is most explicit about ignoring heterogeneity. Other frameworks are somewhere in between.)

From a PSST perspective, employment and output decline when patterns of production become unsustainable in some sectors before new patterns have been established. This makes the concept of potential GDP untenable.

Let us say we are at full employment in 2007 and that we will return to full employment in 2015 (I am not predicting that, just using it as an example). What is potential GDP today? Is it the GDP we could have had by keeping everyone working exactly as they did in 2007? Or is it the GDP that we will produce in 2015, even though the production techniques, patterns of trade, and even some of the goods themselves have not all been discovered as of today?

The standard practice is to draw a straight line between GDP in 2007 and 2015, look at the current date relative to that line, and call that potential GDP. The problems with this are obvious. First, we do not know what GDP will be in 2015 (or whenever it is we might return to full employment.) Second, just because we draw a line between 2007 and 2015 does not meant that we belong on that line right now. The PSST story is that we are where we are because we constantly have to discover and tinker with the processes and patterns of specialization and trade that make the best use of resources.

Comments and Sharing

CATEGORIES: Macroeconomics

COMMENTS (10 to date)
Mark Thoma writes:

See here:

MattW writes:

In one sense potential GDP given real-world limitations is always going to be actual GDP.

Bill Woolsey writes:


We do not draw straight lines. In the example you describe, potential output grows more slowly (and possibly falls) and then rises as resources are reallocated.

The live issue is whether output falls below potential due to too little spending.

If spending falls more than potential, does output fall below potential, or do the prices of goods and services and wages all fall enough so that real expenditure remains equal to potential.

The new new classical, market clearing model assumes that perfectly flexible prices keep real expenditure equal to potential output at all times.

Lord writes:

But we do know what gdp will be when we return to full employment. One only has to look at the past two centuries of data to see how stable this is. For all the deviations such as depressions, we do revert to trend eventually.

UnlearningEcon writes:

Dr Kling,

Your PSST could be interesting but so far it seems to be little more than 'let's do nothing, we can't understand how complex the market is so, erm, get the government out of the way'. Perhaps developing some more rigorous theory and incorporating a role for both the private and public sector could get people on your side?

Dave writes:

How much of the spending that was associated with the pre-crisis GDP level was propped up by unsustainable debt accumulation?

Many seem to believe we can get right back to the trend we were at, ignoring the role that debt played in boosting spending and providing ample opportunity for low skilled workers to produce goods with inflated values (houses).

William Bruce writes:

To second Lord's point, how does PSST reconcile with or depart from the "Plucking Model"?

Moreover, how does PSST address inadequately compensated demand for money? I would have assumed that the market monetarists and fractional-reserve Austrians made more intellectual headway here, to the point that PSST would be, at most, complementary to the monetary issues -- with the exception of Austrian capital theory. What am I missing? Why aren't we either filing PSST under "structural," or going whole hog for an Austrian capital-theory story?

Bill Woolsey writes:


In naive Keynesian theory, production is limited by demand. Firms can produce whatever they can sell.

You are adding to the theory a claim that "debt" is what generated this demand. People borrowed more than they could afford to buy things, this generated demand, and the firms produced would they could sell.

This is a theory that demand has fallen, and so firms will produce less.

Potential output is not how much we can produce because someone will buy it. It is productive capacity. It is limit on the economy due to the scarcity of resources--land, labor, and capital.

Suppose money is simply printed and given to people to spend. They don't have to borrow. It is just a free gift and no one worries about what the money will buy in the future. We could increase "demand" without limit. Now, getting this exactly right to some particular amount of demand would be very difficult, but we could easily double or triple it.

How much could be produced? Abstracting away from the issue of "demand," what people will buy, how much can firms produce and sell?

That is what the concept of potential output is about.

When you say, how much of production was based upon unsustainable debt, you are asking about how much of the demand in the past depended on unustainable debt.

One more aside. When people borrow to consume, they are dissaving. Other people are saving. When some people don't borrow, perhaps because their debts are unsustainable, this leads to more saving on net--less dissaving by some failing to partially or completely offset the saving of others. On the whole, this unsustainable debt argument is that people are saving too much.

And so, it is just the most simplistic Keynesian argument. Debts are unsustainable, so households as a group are saving too much, so firms can't sell as much, so we must produce less.

It is an argument that too much saving leads to a permanent unemployment equilibrium, with production permanently below potential.

It is nothing to do with potential income, which grows due to population growth (more workers,) investment (more capital goods,) and improved technology.

Joe Eagar writes:

Potential GDP was invented so central banks (and the non-monetarist economists who watched them) could understand economic stabilization.

Potential GDP is the level at which the economy begins to overheat and inflation becomes a problem, and the central bank has to tighten. This is an informal definition that central bankers themselves use; they're well aware that the economy changes over time.

Estimating potential output is notoriously difficult, which central bankers are also well aware of (they seem to spend an inordinate amount of time in their meetings discussing estimates of the output gap, and how it evolves over time).

Joanne writes:

As a tangent conversation of that article that led to the review of Arnold Kling’s proposed “Patterns of Sustainable Specializations and Trade” I want to disprove the assumption within the paper that “all production technologies are unknown” although I can see where Kling is coming from, I believe that the focus of his assumption should not be that all production technologies are unknown, but that: a growth path of existing production technologies can help explain PSST and that the assumption of unknown production technologies should only be confined to the notion of unforeseen disruptive production technology.

Comments for this entry have been closed
Return to top