Scott Sumner  

When supply is demand

PRINT
Krugman on Health Care: I Advo... Practical Guidance for Prudent...

In order to do good economic analysis, it is important to distinguish between supply and demand shocks. Commenter SG pointed me to this curious post by Paul Krugman:

For those new to or confused by the term, secular stagnation is the claim that underlying changes in the economy, such as slowing growth in the working-age population, have made episodes like the past five years in Europe and the US, and the last 20 years in Japan, likely to happen often. That is, we will often find ourselves facing persistent shortfalls of demand, which can't be overcome even with near-zero interest rates.

Secular stagnation is not the same thing as the argument, associated in particular with Bob Gordon (who's also in the book), that the growth of economic potential is slowing, although slowing potential might contribute to secular stagnation by reducing investment demand. It's a demand-side, not a supply-side concept. And it has some seriously unconventional implications for policy.


SG wondered why slowing population growth is not an adverse supply shock. Of course it is an adverse supply shock, more specifically it causes the vertical LRAS curve to shift right more slowly, or even move to the left.

Krugman's a smart economist, and certainly knows this. So why does he consider it a demand shock? Because under certain policy regimes an adverse supply shock can shift AD to the left. If you are foolishly targeting interest rates, as the Keynesians advocate, then a reduction in population growth will cause the Wicksellian equilibrium interest rate to fall relative to the target interest rate, and this will depress NGDP growth. If you are foolishly targeting a monetary aggregate, as older monetarists suggest, it will reduce NGDP by reducing velocity (due to lower nominal interest rates.) If you are foolishly targeting inflation, as the Germans advocate, then an adverse AS shock forces the central bank to tighten money and reduce AD.

On the other hand if you sensibly target NGDP, as market monetarists advocate, then a supply shock such as slower population growth will not impact AD at all. With sound monetary policies all of macroeconomics becomes much easier to understand. The book Krugman touts in his post is full of arguments that are almost incomprehensible, because you can't tell whether the author is talking about a supply-side problem that cannot be addressed with monetary stimulus, or a demand-side problem that is nothing more than monetary policy failure.

In a world of NGDP targeting, level targeting, internet macroeconomics debates would be 100 times more transparent. People would stop talking past each other. The relevant concepts would be clear.

PS. I have a related post at MoneyIllusion.


Comments and Sharing





COMMENTS (13 to date)
Market Fiscalist writes:

"On the other hand if you sensibly target NGDP, as market monetarists advocate, then a supply shock such as slower population growth will not impact AD at all"

What if the supply shock causes businesses to reduce investment for any given level of NGDP ? Then targeting NGDP will result in sub-optimal RGDP. Targeting nominal investment spending (or perhaps nominal wage spending) would then be optimal relative to NGDP targeting.

Whatever you target there would probably be some scenario where it turns out to be the wrong thing to have been targeting.

Dave writes:

Sorry to hijack this particular post, quick question from a newbie:

What path would you recommend someone with little education in economics (think micro/macro 101 level) follow to get a better grasp of MM ideas? Any books/textbooks/blog post series?

Thanks.

Michael Byrnes writes:

Dave,

I suggest you read some of the intro stuff on Scott's site (Check the links under "Quick intro to my views" on the right side of the page):

www.themoneyillusion.com

Vox's Tim Lee did a nice feature and interview with Scott also:

http://www.vox.com/2014/7/8/5866695/why-printing-more-money-could-have-stopped-the-great-recession

Dave writes:

Thank you, Michael.

I had never even noticed that section on Scott's blog.

It should probably be bumped up in the side bar.

Scott Sumner writes:

Market Fiscalist, You said;

"What if the supply shock causes businesses to reduce investment for any given level of NGDP ?"

Why do you assume that would be a suboptimal response to a supply shock?

Dave, I would give the same answer as Michael. Also look at other MM blogs.

Britmouse writes:

Targeting NGDP would still fail for some AS shocks, say if the population doubled overnight (open borders!)... perhaps unrealistic cases, but cases where targeting per capita NGDP or nominal wage targeting would be expected to keep the labour market stable?

SG writes:

Scott,

Thanks! That's a very helpful answer.

I'd like to suggest that secular stagnationists also add tsunamis, OPEC, housing bubbles, financial crises, tax hikes, occupational licensing, Richard Fisher, the EU and the BIS as additional causes of secular stagnation. I'm sure there are plenty more.

This is great. It makes me wonder whether every business cycle since 1913 can just be explained by incompetent procyclical monetary policy.

Market Fiscalist writes:

"Why do you assume that would be a suboptimal response to a supply shock?"

My (possibly flawed) thinking was as follows:

- Lowered 'animal spirits' cause suppliers to supply less at any given level of demand than in the previous period of full employment

- If the CB hits it NGDP target then supply will be lower that the previous period and this will cause RGDP to fall below this optimal levels

- If the CB had targeted investment or wage spending rather than NGDP then they would have maintained RGDP at a level closer to optimum (at the expense of greater inflation).

Philo writes:

Do you advocate targeting NGDP, or *NGDP per capita*? In the face of a sudden big change in the rate of population growth neither seems optimal, though I suppose NGDP would be better. Does it matter whether the change in population growth was expected or unexpected?

Scott Sumner writes:

Market Fiscalist, If you target a real variable then the price level becomes indeterminate--you end up with hyperinflation or hyperdeflation.

Philo, I prefer per capita NGDP.

Market Fiscalist writes:

"Market Fiscalist, If you target a real variable then the price level becomes indeterminate--you end up with hyperinflation or hyperdeflation."

investment and wage spending are nominal variables aren't they ?

Scott Sumner writes:

Market Fiscalist, I thought you meant real investment. Yes, you could target nominal investment or nominal wages. I would much prefer nominal wages, as it would smooth out the business cycle.

Philo writes:

But surely if the U.S. suddenly let in an enormous flood of poor immigrants, targeting *per capita* NGDP wouldn't work well. How about targeting *NGDP produced by people who were here throughout the period (from beginning to end)*? (For example, if expected NGDP one year in the future were being targeted, the "period" would be the coming year.)

Comments for this entry have been closed
Return to top