Falling Labor Demand
I have had a series of posts (one, two) in response to Simon Wren-Lewis‘ post on an aggregate demand constraint. My posts give another version of his model which puts real wages on the y-axis and Qty labor employed on the x-axis. I put labor share on the y-axis and employment rate on the x-axis. I make these changes based on the following equations.
Labor share = real wages/productivity
Employment rate = Qty labor employed/labor force
So labor share is positively related to real wages… and employment rate is positively related to Qty labor. (I convert the labor share index into an effective labor share by multiplying by 0.762.)
In the second post, I posted this graph of data since the 1950s.
As labor share has dropped overtime, the model suggests that the employment rate has a lower and lower limit due to the sloping labor supply curve.
But now I want to add some views about the labor demand curve. The drop in labor share may actually be a drop in the labor demand curve in the model.
The model for the 1950s and 1960s. We see a boundary for the labor market in terms of supply and demand. I will include these lines in the following graphs as reference points.
The model for the 1970s. We see that labor demand curve shifted down and the labor supply curve shifted left. In the following graphs, we will see the labor demand curve continuing to shift downward. Business is basically less willing to demand labor at any given level of sharing revenues with labor. The labor supply curve may have shifted left due to the baby boomers and women entering the labor force creating the potential for more unemployed workers at any given level of labor share.
The model for the 1980s. We see labor demand continuing to fall. We see labor supply continuing to represent more potential for unemployment especially due to the manufactured recession by Paul Volcker.
The model for the 1990s. Labor demand stayed the same from the 1980s, but labor supply shifted back right.
The model for the 2000s. After a surge in labor share around the turn of the century, employment rose. Then labor demand fell through the 2001 recession, which explains the slow recovery of employment. Labor supply returned to its original point from the 1950s and 1960s.
The model for the 2010s. Labor demand has fallen much more since the crisis. Just like the 2001 recession, this fall in labor demand explains the slow recovery of labor after the crisis. Lower labor share represents a lack of valuing and demanding labor. Business is less willing to share its revenues with labor. Labor supply looks to have the same curve as it did in the 1950s and 1960s.
A falling labor share leads to a tighter effective demand constraint. Yet, the real problem in the labor market is labor demand, not labor supply. Lower labor demand is creating a tighter effective demand by including less labor in the national income.
The concern going forward is whether labor demand will continue falling. That has been the pattern for decades. The result would be an even higher natural rate of unemployment.
I wonder if you could do this for the Asian nations or the primary areas where labor need here has been off shored if there would not be an offsetting effect.
it’s not like labor is not needed. It’s that the need went shopping outside of this nation since the 80’s.
The great thing about the US is the data for labor share. It is consistent and detailed. I do not trust yet labor share data to be this good for any other country yet. For example, look at labor share for United Kingdom…
It is absurd. i suspect the Simon Wren-Lewis doesn’t use labor share for the bad data in the United Kingdom.
You might think Germany was better… not much
According to the graph for Germany, labor share hasn’t changed since the crisis. They have no clue about labor share.
The same futile data exists in Asian countries too.
The Untied States gives labor share data on a quarterly basis too. It is great.
Here is China…
Were you expecting labor share to go up in China?
Ed: Were you expecting labor share to go up in China?
I would not expect share to go up unless the number of people being employed is rising faster than income rose. I would expect with their economy structure, labor share is rather flat in that the ratio remains unchanged, income paid to labor vs total income. That is, any rise in demand for labor since the great migration of global companies to China et al has not created a greater share of the income to labor but instead just more people employed.
I was thinking about this statement: Yet, the real problem in the labor market is labor demand, not labor supply.
We have and they have plenty of labor supply. Business has seen their need for labor globally continuously rising even with the recession. American companies did not just market their rising need for labor, they marketed just about everything they could to a lower cost region of the world.
So, had that rising need been met with US labor, what would the charts look like? And, if you could get clean data from other nations, would we see our declining need looking back at us as a rising work force population over seas?
It’s not like the global economy hasn’t been growing.
Labor share fell in China in the last 15 years by quite a bit. That puts pressure on the US to lower labor share too. Labor is less valued there too.
It would be the dream of every business to lower labor share and have the labor supply curve completely inelastic (vertical). I imagine that the labor supply curve of the model above would be fairly vertical in China. Yet in the US, when labor share drops, or just real wages drop, then people are less willing to supply their labor. and we have more of an up-sloping labor supply curve.
The drop in labor force participation over the years and even before the crisis is no surprise when you see the graphs above in this post.
So, labor share declines, resulting in lower labor demand? Yet the economies are growing. Labor is butting up against productivity offsets via technology? Yet, people at least in this nation are working more hours and have less leisure time.
If labor has less to spend for consumption, then sure labor demand ultimately declines, but would that not result in a recession/depression or at least effectively zero economic growth? Assuming tech has not reached the point of offsetting human labor.
I’m not sure I see how you get that the real problem in the labor market is demand for labor unless we have crossed that point that technology has finally reached a point to which it produces productivity offsets to human labor equal to or faster than products are created (not produced but actually created).
I mean, you have certainly documented what is happening and the relationships, but I’m not certain it’s at the point of suggesting a policy direction? How do you increase labor demand if tech is producing productivity offsets to human labor as fast as products are created? If tech is not actually offsetting the demand for human labor then the charts are predominately the results of policy.
Good morning Daniel,
Basically I say that labor demand is the problem because it has fallen since the 1950s but labor supply has maintained itself in the same position. So it was demand that backed out of the deal.
But you bring up productivity and there is more to say about that.
Start with the equation…
labor share = real wages/productivity
We have all seen that over the years, real wages have not kept pace with productivity. OK… we see the fall the fall in labor share.
As productivity increases, labor share falls. So we move down and to the right on the labor demand curve. More labor is demanded with productivity increases. Labor is relatively cheaper. More productivity would imply more employment from a demand perspective.
Yet what happens on the labor supply side? More productivity will lead to less labor being supplied. For one, labor is cheaper and some people choose not to supply their labor at a lower price. They may choose to be employers of labor instead. Also,
But there is a flip-side to the model above which is for the labor market. The flip-side is the consumption market. The labor supplied in the model above becomes the consumption demand in the consumption market. Lower labor supplied becomes lower demand for consumption. That is where the effective demand limit comes in.
For example, let’s say that labor share is dropping and labor demand increases to the right beyond the crossing point of supply and demand. Labor demand would be happy right? Isn’t labor supplying all that the firms desire? And still, more people are being employed right? Isn’t this what krugman and others want?
Put your mind around this thought… the labor supply curve is actually the effective demand limit curve. I don’t know if I can write this yet. It may be a concept too much for others to grasp.
But you will not be able to have a “free-market” situation where labor is supplied to the right of the labor supply curve. Why?
Firms would be producing beyond the capacity of labor to purchase the production. Profit rates would start to drop and firms would back off from hiring the relatively cheaper labor.
So productivity has its effective demand limit.