To find out if productivity is really demand constrained, let's look to see what happens when productivity is up against the effective demand limit. We will find that productivity stops and sits for a number of years. First we get the data and build the model.

The data will come from this graph at FRED. The graph shows real GDP (Y), effective demand (E) and total labor hours (L) all indexed to 2005=100.

The light orange line is labor hours (nonfarm business sector), which are still at the same level of 15 years ago. Yet, real GDP (blue line) has risen over that time. So the basic story is that we have been more productive with the same labor hours. Here is a graph to show that real GDP used to rise with increased labor hours. However, since the 1990's, real GDP has increased even though labor hours have not (total plot 1967 to 2013).

OK... then how can productivity be demand constrained when real GDP keeps rising in spite of the fact that labor is not increasing their hours to earn more income? In other words, wouldn't a demand constraint be dependent upon labor hours? Well, no...

Productivity is normally determined by dividing real GDP output (Y) by total labor hours (L).

Productivity = Y / L

Now to show when productivity is constrained by effective demand, we divide effective demand (E) by total labor hours (L) to get the effective demand limit per labor hour. The reason is that what is produced in an hour cannot surpass the hourly effective demand limit. We will check the data to see if this reasoning holds up.

Real hourly effective demand = E / L

(Note: Effective demand has the equation, E = Y * e/T ............. e = effective labor share, T = TFUR, total factor utilization rate (employment rate * capital utilization)

Real hourly effective demand, E / L = Y / L * e/T

Real hourly effective demand = productivity * e/T

Real hourly effective demand allows us to compare effective demand with hourly measurements, like productivity, real hourly compensation, capital used per hour, etc.

Now, what happens if we graph productivity (Y/L) against real hourly effective demand (E/L)? (Data in graph is given by quarters from 1967 to 2013.)

This graph is a scatter plot using the data from graph #1. Let's first look at the red line which represents the effective demand limit. A basic principle of effective demand is that real GDP is constrained below effective demand. Thus the red line sets the theoretical effective-demand limit for productivity.

The blue line is how productivity has moved with real hourly effective demand. Productivity is how much production is produced (in real 2005 dollar terms) per hour. Real hourly effective demand is the potential demand per hour for hourly production. In theory, productivity per hour should be limited by the hourly effective demand limit; Production would not go over the demand constraint. Thus, the plot in graph #3 should stay below the red line. In other words, productivity should stay below the effective demand constraint.

And what do we see in graph #3? The plot of productivity does in fact stay below the effective demand limit (red line). Productivity will bounce along the effective demand limit.

Using effective demand gives a wonderful way to view the behavior of productivity. Normally real GDP is plotted against effective demand. However, productivity moves differently than real GDP because of the variability in labor hours.

Yet, the most interesting part of this graph is how productivity behaves when it is close to the effective demand limit. Productivity stalls for a number of years... 3 to 4 years. We find that during those 3 to 4 years, productivity does not increase much at all. Effective demand will not move much either.

It is impossible to see in the graph, but there are numerous dots all bunched together in the spots where productivity stalls at the effective demand limit. I only count the dots up against the red line. You cannot see them all in the graph. For example, between 1994 and 1997, you see what looks like two dots peaking over the red line. There are actually 16 dots in that little space between those two dots; that is 16 quarters... 4 years of productivity being completely stationary and demand constrained.

If you look at the plot of 1977 to 1979, you will see a line that heads straight toward the effective demand limit and comes straight back. There are in fact 20 dots in that little line that sticks out. That is 20 quarters or 5 years worth of data. In other words, productivity and hourly effective demand moved in a perfectly straight little line against the effective demand limit for 5 years. This relationship between productivity and effective demand has never been seen before. It certainly is interesting.

When productivity increases up and toward the left in the graph (meaning effective demand is declining while productivity increases), productivity eventually hits the hourly effective demand limit and stops. Productivity sits against the effective demand constraint for a number of years until the tide turns and effective demand starts to increase. Then productivity can start increasing. In effect, effective demand has to start increasing first in order for productivity to start increasing. Then, hourly effective demand and productivity will increase together moving in the direction of the effective demand limit (red line).

Lots of words to describe a simple process. Let me boil it down. Productivity is often constrained by effective demand.

Let's look at current data at the end of the plot. We see that productivity has come close to the effective demand limit again and has stalled out in the same spot for 2 years. Productivity itself has stalled for over 3 years. I will state three conclusions...

- Productivity is being compressed by the effective demand limit again.
- Productivity is not going to be increasing soon unless effective demand reverses its decline.
- For those, like Ray Dalio, who say that productivity will increase as the economy recovers, they will be disappointed. The implication is that the economy instead will have to grow upon an increase in credit-fueled consumption.

Other relate posts

I think you've got it. Productivity gains are there to be realized , but only if the spending power is there as well. We've gotten used to the idea that debt could be an effective substitute for sharing those gains in productivity , replacing the wage gains that should have accompanied the productivity increases. Now that it's more difficult to ramp up the debt again , we're stuck.

You might like to read this article by Crystia Freeland about the UK , where a similar dynamic has unfolded :

http://articles.chicagotribune.com/2013-06-27/news/sns-rt-us-column-freeland-20130627_1_middle-class-economy-labor

Posted by: Marko | 07/27/2013 at 02:42 PM

I have only just discovered your effective demand concept, so forgive me if I have misunderstood it in any way, but I have a few thoughts at this point:

1) If I understand what it is supposed to mean correctly, I think you should rename it potential demand. Even if potentialdemand.typepad.com is taken.

2) Something strikes me as wrong with it just looking at the graph you present of ED over time. In particular, the way it shoots up in recessions and remains flat or declines in booms. If this is some kind of lid on the economy, it makes sense that we would grow toward the lid during a boom, then hit it, then drop away during a recession, but why would the lid race away from us in a recession, and why would the potential demand be attracted downward toward GDP during growth periods? To me, this suggests the equation isn't quite right; that it does tell us when we're up against the lid but isn't giving a reasonable estimate of where the lid is when we're not against it.

3) It surprises me that the balance of trade doesn't enter into effective demand in any way. Or for that matter, government. Do you have a post where you explain this?

4) So a little algebra gets me to the law: capacity utilization * (1 - unemployment) <= labor share. Would I be correct in understanding this to mean that rising profits don't boost demand?

Posted by: Eric L | 07/28/2013 at 10:15 PM

Marko,

I checked out that article. Their explanation for problems in productivity point to flexible labor markets. They don't have a way yet to determined a demand constraint on productivity. We do now.

Posted by: Edward Lambert | 07/29/2013 at 08:11 AM

Eric,

Answers to your points.

1) Effective demand is potential demand and more. It would be better to place potential demand equal to the demand at potential real GDP. Potential demand would then be the center of the business cycle. Effective demand would be the top limit of the business cycle.

2) Effective demand shoots up during a recession because capacity utilization takes a hard hit. For example, in the crisis, capacity utilization dropped by 16%, while employment and real GDP fell by 4.7% each. If capacity utilization had fallen 4.7% too, effective demand would have stayed stable.

and you are right, effective demand acts as a lid, but also shows the extent that labor and capital is not being utilized.

3) Balance of trade and government point to leakages and add-ins in the circular flow of the economy.

Take China for example, they have a huge export industry with low domestic consumption. The result is a low labor share of income. If labor received much more of that export market, labor would have lots of money to buy things. There would be a big inflation problem for one, and less capital investment for two.

I am not answering your question very well. I probably need to do a proper post to explain this.

4) Your algebra is correct. If profits rise and labor share stays the same, labor income will rise with profits. Therefore real GDP will rise. But... the effective demand limit upon utilization of labor and capital will not change.

So rising profits will boost aggregate demand, real GDP and effective demand, but not the effective demand limit on utilization of labor and capital.

Again, I am not answering clearly. I just woke up.

Let me know if my answer makes sense.

Posted by: Edward Lambert | 07/29/2013 at 08:29 AM