Labor share affects the potential of investment to raise GDP
The circular flow is a model used to show how money and products move through an economy. I am going to use a simplified version to show the effect of labor share of income on GDP. Many people think that labor income is not quantitatively different than capital income. I will show that the equilibrium level of GDP is affected by a change in labor share of income. This would have important implications for monetary policy and expectations of GDP potential.
Basic Equilibrium
Let’s start out with a basic model. It will only involve labor, owners of capital, firms and a financial sector. The only injection into the circular flow will be investment. The only leakage from the circular flow will be savings. The model does not include the government sector nor imports/exports.
Start at the out-going income from firms. Firms are paying out $1 million in income. 80% will go to labor (labor share of income of 80%). 20% will go to capital (capital share of 20%). The GDI (gross domestic income) of the economy is $1 million. Capital takes all of their income and saves it for investment. Labor takes 85% of their income and consumes production of finished goods and services (marginal propensity to consume of 85%). Labor takes the remaining 15% of their income and puts it into savings.
Capital savings and labor savings then add together for the total of savings in the economy. Then those savings are invested in firms to maintain and increase the means of production. Consumption is directed at the finished goods and services produced from the capital used in the means of production. The in-coming money to firms is the sum of investment and labor’s consumption.
In the graph above, the economy is in equilibrium. (Savings = investment) (In-coming = out-going)
In equilibrium…
GDP = consumption + investment
GDI = labor income + capital income
GDP = GDI
Investment rises
Money increases in the economy when banks make loans. This is how money is created in the economy. Banks can just create money when they make loans. So let’s change the graph above to show an increased level of investment by banks.
Look in the green boxes. Investment has risen to $520 thousand from $320 thousand. Investment is injecting an extra $200 thousand into the economy. Maybe QE is finally providing liquidity. The result is that there is now $1.2 million in-coming to the firms. Consumption has not changed because the extra investment has not been paid out yet as income.
The economy is going into a disequilibrium. (Savings ≠ investment) (In-coming ≠ out-going)
Firms now pay out that “in-coming” $1.2 million as income maintaining the same labor share of 80% and capital share of 20%. We allow that money to flow through the economy. Let’s see what we end up with.
We see the out-going $1.2 million go through the circular flow and leads to a total savings of $384 thousand, which is still less than the investment level. Consumption has risen to $816 thousand. Total in-coming into firms has now risen to $1.336 million from the increased level of investment and the new increased consumption.
The economy is still not in equilibrium. (Savings ≠ investment) (In-coming ≠ out-going)
Firms now pay out the increased in-coming money of $1.336 million in another round as income to labor and capital.
We see that total savings continues to increase and that consumption continues to increase. The in-coming money to firms continues to increase. And so the economy goes through numerous rounds like this until an equilibrium is reached. All this is due to the original disequilibrium created from increased investment by banks. Here is what the economy will look like once it reaches equilibrium, where (Savings = investment) and (In-coming = out-going).
The important thing to see here is that the equilibrium level of GDI (gross domestic income) has increased by $625 thousand as a result of increasing the investment level by $200 thousand.
But now let me ask you… What equilibrium level of GDI would this model have reached if labor share of income had fallen to 75%? If you think there is no quantitative difference between labor and capital income, then the resulting GDI should be the same, right?
Lowering labor’s share of income
I will now lower labor’s share of income in the model. Why? Because labor’s share of income has fallen 5% since the crisis. This fall is unprecedented. And I present a simplified simulation of how that fall in labor share affects investment… and how it affects the efficacy of monetary policy… cough cough.
I will simply change labor’s share to 75% using the increased level of investment of $520 thousand. Then I will solve for the equilibrium level of GDI. Here it is…
GDI has fallen from $1.625 million to a lower equilibrium of $1.434 million. We can see that the economy is in equilibrium (Savings = investment) and (In-coming = out-going). The only difference was lowering labor’s share to 75%.
Conclusions
- After the crisis, labor’s share of income fell 5%. The result is that the ability of investment to increase GDP output is muted. Investment has had less power to boost GDP. The truth is that the economy needs even more investment to get the same result it would have had with a higher labor share.
- Investment returns to firms in smaller amounts. Firms receive less in-coming money. Demand would appear low, but it’s just a labor share constraint on demand.
- Labor’s consumption and labor’s savings have decreased. This is obvious from lower labor share and lower GDI. Yet, look closely at the capital income in graphs #5 and #6. In graph #6, capital’s income at equilibrium with a lower labor share has actually risen to $358 thousand, in spite of GDI falling too. This model shows how corporate profits can reach record heights and Main Street struggles when labor share falls.
- A lower labor share affects the efficacy of accommodative monetary policy. How can you expect monetary policy to put money in the hands of people when the falling labor share is taking money out of their hands? Again, demand would appear low.
- Why has the economy responded so poorly to loose monetary policy? Monetary policy is heading toward a lower equilibrium level of GDP due to the lower labor share. We can clearly see above that a lower labor share mutes the potential of investment to raise GDP.
- We would also find that the natural rate of unemployment would be higher if we were to associate the rise in real GDP with utilization of labor and capital. Once GDP is limited to a lower equilibrium level by a lower labor share, employment also gets limited. (My view is that the natural rate of unemployment has risen to around 7% due to the fall in labor share.) Keynes sought to show that employment gets limited by effective demand. This model supports that view.
Does wage stagnation along with increasing productivity result in a lower labor share?
Jerry,
Yes, according to the equation…
labor share = real wage/productivity
If real wage stagnates, and productivity increases, labor share comes down.
For accounting precision, may want to use Saving rather than Savings. (I believe you were discussing flows rather than stocks.)
The trickle-downers would argue that the higher gdp of graph #5 would be maintained , as the capital income earners would divert 85% of their new 5% larger income share away from savings and towards consumption , allowing the system to balance at the higher level.
The counter-argument is that the propensity to spend from capital is lower than from labor , but few will concede this point , and fewer still can back up their claims – on either side of the argument – with hard numbers. Such is the dead-end we call economics.
KP,
You are absolutely correct. Saving is the word that should be used.
Marko,
Capital is not diverted into consumption of finished goods. It is used for the means of production.
The propensity to spend by capital income does not apply to consumption, because capital income is used for production not consumption.
If an economist says that capital income will increase demand, they are right, but not for finished goods, only for capital, buildings, land, equipment, means of production.
If a business uses its retained earnings for consumption instead of investing in growth opportunities for production, most shareholders don’t like that, because it squanders growth of the business. If businesses en mass started to convert retained earnings into labor income (personal income) for consumption, then labor share would rise.
” Capital is not diverted into consumption of finished goods. ”
Edward,
That may be the specification you’ve chosen for your model , but it’s not the standard adopted by BLS and others when they discuss labor vs capital shares. Dividends , interest , and capital gains fall under the capital income designation , and the shills for the elites will argue to the death that rich people consume with abandon from these income streams.
This is where the battle should be waged , because that’s where the “demand insufficiency” argument will be settled , if ever. The reason Keynes’ work was so actively suppressed was because he had the temerity to discuss “marginal propensities to consume”. That’s like kryptonite to the elite’s economic supermen.
Marko,
I acknowledge that the elite consume from capital income. Even labor income is used for investment in capital.
Many of the variables in the model stay steady, investment level, labor share, overall propensity to consume…
Ok… in equilibrium, capital income is used for investment and some of it for consumption… granted. So if capital’s share rises, does the “extra” capital income all of a sudden have a much higher propensity for consumption? Does the equilibrium level really change due to an overall increase in propensity to consume?
Another alternative is that the money would end up in savings and then sent oversees for investment? That would still be a leakage from the economy to offset domestic investment and domestic consumption.
Let’s do a quick back of the envelope calculation, capital income is about $4 trillion in 2009 $$. Labor income is about $11.5 trillion. From the model above, a 5% drop in labor share led to a 17% drop in consumption. … 17% of $11.5 trillion is $1.5 trillion.
Capital income rose 10% in the model. Then capital income would rise $400 billion.
As you can see, an estimated drop in labor consumption of $1.5 trillion can not be covered by a rise of $400 billion from capital income even if the propensity to consume with that capital income is 100%.
The model still shows that lower labor share will lower the equilibrium level of GDP, no matter how the rich spend their capital income.
Marko,
I responded too fast… if that extra capital income was used for consumption, the equilibrium level of GDP would rise back up.
The questions are… What percentage of the $4 trillion in capital income is being used for consumption? and has that percentage changed over time?
“The model still shows that lower labor share will lower the equilibrium level of GDP, no matter how the rich spend their capital income.”
Edward,
No , it does not show that. I’m surprised that you don’t see this.
Let’s take an extreme example to prove the point. Let’s assume labor share goes to zero , and capital share goes to 100% , and that capital spends their increased income share at the same 85%/15% consumption/savings ratio as labor would have.
The easiest way to visualize the result is to take graph #5 and relabel all the orange labor boxes – since labor is now slave labor – instead as ” capital income pot #2 “. All the numbers in all the boxes remain the same , we’ve just dropped out the labor income share and redistributed it to capital earners that dispense with the extra income just as if labor had received it.
The important point , as I emphasized above , is that earners of capital income to not spend just like labor income earners , and this is the crux of the demand shortage problem or the one-and-the-same “overcapacity” problem .
Edward ,
I guess while I was crafting my response you took a second look , so my post just above is unnecessary.
Marko,
Once we can distinguish labor income from capital income, the model works. If we start seeing capital income used for consumption, the model wouldn’t work. Just like you say, “earners of capital income do not spend just like labor income earners.” In that difference, the model will work to some extent.
An interesting graph posted by Mark at the ‘Illusion of Prosperity’ blog , showing the trendline of U.S. imports per capita from China :
http://illusionofprosperity.blogspot.com/2013/08/chinas-growth-story-in-one-chart.html
It looks like growth in our imports from China petered out when the housing bubble ATMs stopped spitting out money. We’ve created more problems with aggregate demand than just our own.
Notably , China seems to have a better grasp on the nature of the problem than we do , judging from official pronouncements and even some recent policy moves. They know they have to ramp up domestic demand , so they’ve been aggressively raising the minimum wage , and they understand that a stronger social safety net would be beneficial. They also seem to want to get away from debt-dependent growth.
Maybe they can show us how economics works , since we can’t seem to figure it out for ourselves.
Ed, I would delete a portion of this as unnecessary and incorrect:
“Money increases in the economy when banks make loans using a fractional reserve system. The banks can lend out more money than they have in reserves. This is how money is created in the economy. Banks can just create money when they make loans.”
So:
Money increases in the economy when banks make loans. This is how money is created in the economy. Banks can just create money when they make loans.
Steve,
I made the change just as you wrote.
So it’s velocity of money. If labor were able to borrow at zero rates, there would be no impact on GDP (labor would spend all they want to). Because money is not available at zero rates, their spending is constrained. You can increase labor share to decrease this constraint, or you can lower interest rates to reduce the constraint.
J Goodwin,
Excellent point. I see that the natural rate of interest for labor is around 1%. This means that they would need low interest rates to be in equilibrium. Your idea fits that view. But labor has to be careful of increasing their debt, even at zero % interest rates. There aren’t better prospects into the future for higher income.
The model above shows the injection of extra investment. It could also show the injection of extra credit to labor for consumption, like you say. But once the economy reaches equilibrium, labor would have to start paying back that extended credit on balance. Then you set the economy up for a recession and people with more debt.
@J.Goodwin:
If labor could borrow at zero rates, they’d have no reason to ever pay it back. Hey wait, that’s sounding a whole lot like helicopter money!
http://macromarketmusings.blogspot.com/2013/08/helicopter-drops-as-insurance-against.html
(Follow the link therein to SRW’s post.)
“The economy is going into a disequilibrium. (Savings ≠ investment) (In-coming ≠ out-going)”
Yes! This graph from Keen and Bezemer shows the instantaneous jump in potential spending when banks lend ex nihilo:
http://www.asymptosis.com/wp-content/uploads/2013/08/Screen-shot-2013-08-08-at-12.06.32-PM-480×473.png
From here:
http://debunkingeconomics.com/wp-content/uploads/2012/10/TowardsUnificationMonetaryMacroeconomicsMathArgument.pdf
Ed, isn’t this model just dying to be modeled in Minsky?
Steve,
You must write an article on modeling this model in Minsky.
Dammit I was afraid you were gonna say that.
“the natural rate of unemployment has risen to around 7% due to the fall in labor share.”
This rather profound: low wages cause unemployment.
Wouldn’t it be more accurate to say “the fall in labor share is due to the rise in unemployment?
Steve,
I hear you…
Also, Fed rate would have to go lower and lower as labor share goes lower, in order to increase investment injection into the economy just to maintain real GDP at a certain level.
If labor share goes low enough, Fed rate would hit the zero lower bound just to try and maintain optimal employment.
It’s just like Keynes said, the economy can get stuck in a sub-optimal state where full-employment is not reached. We can see this happen easily in the model of this post.
Jerry,
It is accurate to say that a fall in labor share can make the economy get stuck in a sub-optimal state where unemployment stays elevated. This is just as Keynes described, but never modeled.
Is it possible to have a low labor share and high employment at the same time?
Jerry,
Let’s take the current low effective labor share of 74% with an unemployment rate of 4%.
Then we have to figure out the limit for capital utilization.
Capital utilization limit = 0.74/0.96 = 77%
With these numbers, capital utilization would not go above 77%. It is now 78%. So unemployment is not going down to 4%.
So unemployment could drop more, if capital utilization was much lower than it is right now.
Hard to distinguish the effects of low wages on unemployment from unemployment on low wages.
Low wages means lower labor force participation because some people will choose to do unpaid work that allows them to avoid other expenditures that they could not pay if they worked (for instance in many parts of the country, childcare in particular is so expensive that after-tax and support programs, it is financially non-viable to work, i.e. you are paying more for childcare and food/insurance because you work than you would pay in total if you did not work). Even excluding support programs, in two-income families, it is still sometimes just cheaper for one person not to work if they are in a part-time job and the other has full employment.
High workforce availability in certain professions due to unemployment also depresses labor costs (as does cross-border work shifting, international visas for inbound workers from LCCs, etc). That’s the normal supply and demand effect.
We just had our first child, and it brings all kinds of these issues into focus in ways that previously I just wasn’t aware of. The cost of insurance (when you are single with a professional career employer, insurance deductibles are negligible in many cases, until you have a “+1” to deal with) plus you have all those doctor visits, not to mention the medical services immediately related to the birth, the relative cost of childcare (we shipped in a grandmother to stay with us and help care for the baby for a year, next year…maybe we need an au-pair part of the year, because after you start looking at day cares, the au-pair is actually less expensive…crazy as that sounds, and if you have more than one child, you definitely save with an au-pair). When our daughter is 3 she will start going to pre-school for two years, which is on the order of 8k a year and up (depending on whether you use the extended day options) because we don’t qualify for payment support and we don’t live in a universal preschool state.
After those first 5-6 years things ease up a bit because you are only paying for part time childcare and you have public schools to back you up the rest of the day, but I’m sure there is always something.
If you’re the kind of people who believe your child should have a non-public education, you can see why a lot of those people elect to have one parent stay-at-home and home-school their kids, because private religious schools are 15k a year or more per child.