The global decline of the labor share

by on September 5, 2013 at 3:39 am in Data Source, Economics | Permalink

That is the new paper by Loukas Karabarbounis and Brent Neiman, and the abstract is this:

The stability of the labor share of income is a key foundation in macroeconomic models. We document, however, that the global labor share has signi cantly declined since the early 1980s, with the decline occurring within the large majority of countries and industries. We show that the decrease in the relative price of investment goods, often attributed to advances in information technology and the computer age, induced fi rms to shift away from labor and toward capital. The lower price of investment goods explains roughly half of the observed decline in the labor share, even when we allow for other mechanisms influencing factor shares such as increasing pro fits, capital-augmenting technology growth, and the changing skill composition of the labor force. We highlight the implications of this explanation for welfare and macroeconomic dynamics.

In other words, capital-labor substitutability is very real.  The full piece is here (pdf).

1 GiT September 5, 2013 at 3:59 am

The organic composition of capital strikes?

2 Ray Lopez September 5, 2013 at 4:33 am

“At least since the work of Kaldor (1957), the stability of the labor share of income has been a fundamental feature of macroeconomic models, with broad implications for the shape of the production function, inequality, and macroeconomic dynamics” Strawman to a degree? How can labour-capital not be substitutes? The economics profession is in worse shape than I thought if this fact is not acknowledged. Unless these models are very short -term, and hence not meant to be used more than a few months or years out (akin to Keynesian economics being short-term oriented, with the long-term a series of short terms). I suspect this is the case, so this is a bit of a strawman by the authors.

3 HM September 5, 2013 at 11:48 am

Definitely not a strawman. Constant labor share is one of the big stylized facts in macro.

Your remark on labor-capital not being substitutes also misses the mark. There can be constant factor shares and labor-capital substitutability at the same time. The Cobb-Douglas production factor has this. The key is to have exactly the right amount of substitutability so that changes in capital-labor ratios are perfectly by changes in the interest rate-wage ratio.

Lastly, you remark that this must be short term is also wrong. The reason macroeconomics has gone this road is long-run empirical: the labor share has been fairly stable over time and even between countries.

4 Ray Lopez September 5, 2013 at 9:46 pm

Thanks, but unfortunately your third sentence in your second paragraph is grammatically incorrect, seems to be missing a word, and hard to understand the point you are making. But I did find this on Wikipedia and it’s quite good, making the points you seem to be making: (“Cobb and Douglas were influenced by statistical evidence that appeared to show that labor and capital shares of total output were constant over time in developed countries; they explained this by statistical fitting least-squares regression of their production function. There is now doubt over whether constancy over time exists.”)

5 prior_approval September 5, 2013 at 5:39 am

I do wonder how one judges the labor/capital ratio of something like Linux.

And to what fraction software such as Linux (aka Android for those interested in smartphones/tablets), and the world it has created (anyone used the Internet recently?) is simply not covered by an economics model which still talks about labor/capital share.

Sadly, it seems like in an age where being above average is the goal, the implications of having a model of explicit sharing (the GPL being a wionderful example of enforcing a vision of freedom using constitutionally granted monopoly power) is simply not worthy of discussion.

6 Brendan Perrine September 5, 2013 at 10:36 pm

Really good questions. I would totally like to have classes studying that. Another thing that could be studied is its effect on knowledge spillovers.

7 collin September 5, 2013 at 9:05 am

Because there are more assets and capital income long term, is this the main reason why we live in the “bubble” generation?

8 Frederic Mari September 5, 2013 at 9:10 am

Stability of the labour share of GDP is an important component of macro models? When? Where? I don’t think it’s part of the DSGE modelling tool-kit…

It’s also interesting b/c it seems to me that this is an interesting updating of Marxist dogma. If you recall, Marx identified the tendency to substitute labour with capital as THE factor leading to the collapse of rates of return (since only labour is susceptible to be ‘exploited’).

I never liked this Marxist description since the idea you couldn’t ‘exploit’ machines made no sense to me. And, certainly, adjusted for risks, returns certainly haven’t collapsed… SO FAR.

Because one thing capitalists do need is customers. And if Labour share of GDP keeps getting slashed/Labour keeps being substituted, how will this consumption be funded? We’ve just seen the limits of debt-based consumption. Robots and software don’t get paid/don’t consume either.

Bottom line: Rates of return need not collapse (imho) even if all employees were automated away. But I do predict that rates of return will collapse if consumers disappear.

So… what do we do?

9 JWatts September 5, 2013 at 10:20 am

“Bottom line: Rates of return need not collapse (imho) even if all employees were automated away.”

If we are seeing employees being automated away in macro, it’s a very slow rate.

“So… what do we do?”

Well the utopian answer is a universal income, but that can’t happen without:
a) much stricter border control than has ever been managed or
b) a much greater equalization of income between third world and first world countries (developed vs developing, or whatever you prefer)

On the other hand, at the rate the worlds median income is increasing, this will probably all be academic in another 100 years barring catastrophe. When 90% of the world population is in a developed country, the world will be far richer and less volatile.

10 Z September 5, 2013 at 9:24 am

Hmmm… This seems to me like a failure to see the obvious. Financial services have become a greater share of GDP for post-reality economies. Much of it is automated, conducted by robots. It is no coincidence that financial services exploded with the Louvre Accords in the 1980’s. One begat the other. It seems to me that a set of macro polices intended to blow up a heavily automated segment of the economy will inevitably lead to a lower labor share. That does not mean, however, that labor share is declining across the economy. It is just fun with numbers.

In order to tease out anything useful here, I’d want to compare across sectors. Is labor share falling in retail, for example. That could tell me something important about the economy. The old saying, “Wall Street is not Main Street,” has never been more true. Finance is just about divorced from the main economy now. Treating it as just another part of the aggregate economy is a mistake. That said, I’ll read the paper and maybe find out I am mistaken 😉

11 Axa September 5, 2013 at 11:18 am

You can find that in Figure 5 in the paper. Mining is bleeding employees.

12 JWatts September 5, 2013 at 10:24 am

“Is labor share falling in retail, for example.”

Yes, almost certainly. Automatic tellers for banking, automated attendants for gas pumps, self service checkouts at major stores, laser bar code scanners for ringing up sales, robotic DVD stores (Redbox,etc.), streaming music, video, software, etc.

13 mbutu o'malley September 5, 2013 at 11:09 am

Go ahead and toss in RFID for inventory and order management as well as returns processing, automated warehouses, and better inventory control management systems have all reduced the headcount needed for a retail operation.

14 fwiw September 5, 2013 at 12:32 pm

Yep. Look at the growing market share of Amazon.

15 Merijn Knibbe September 5, 2013 at 3:56 pm

The ILO looks at a larger set of explaining variables (including ‘financialisation’) than these authors do and finds:

“We have found that globalisation, i.e. increased international trade, has negative effects on the wage share in advanced as well as in developing economies, which is in contradiction to the Stolper-Samuelson Theorem. Overall, the results are similar for advanced and developing economies, with the possible exception of low-income countries. Financialisation has had the largest negative effect on wage shares. Technological progress (including structural change) has had substantial effects on the wage share, but these have been positive since 1980 and can therefore not explain the decline in the wage share. Globalisation and welfare state retrenchment have had moderate negative effects on the wage share.”

16 Steve Sailer September 5, 2013 at 5:46 pm

Clearly, this demonstrates that the U.S. needs more immigration so that owners of capital can get an even larger share of the pie.

17 Eliezer Yudkowsky September 5, 2013 at 7:19 pm

Obvious question: Is the labor share of income lower now than in 1850? Certainly *vast* quantities of capital equipment are priced lower now than then, and it was harder to access capital markets.

(Looks it up…)

This says *no*:

18 Sanjeev Sabhlok September 6, 2013 at 1:29 am

Just like inequality is irrelevant, so also labour share is irrelevant. These are SOCIALIST CONCERNS and we should not waste time worrying about them.

It doesn’t matter what share of the “pie” is split between capital and labour. What matters (in the broad, philosophical sense) is that (a) people are free (b) there is justice and good governance, (c) as a result the pie is increasing and (d) no one is worse off. Let labour and capital bargain in freedom. The main thing is there should be incentives to and freedom to produce and create wealth through new technology and innovation.

Robotics and IT is entirely positive, the moment people remove their socialist blinkers. I invite economists to provide input/comment on the initial compilation for a book project: The Glorious Abundance and Creativity of the Robotic Age.

19 Larry September 6, 2013 at 3:23 am

Consider a thought experiment in which the capital/labor split is regulated/fixed. That would protect labor’s share, but how big would the resulting pie be?

Now consider the opposite. That’s the factory that hires one worker and one dog. The worker’s task is to feed the dog. The dog’s job is to keep the worker from touching the machines. Ha ha. What is the appropriate labor share of that factory’s output?

20 Thomas Sewell September 6, 2013 at 1:56 pm

Is it possible that the additional labor-related regulations have been pricing labor up relative to capital-substitutions over the last 80 years or so and thus contributing to the decline of the labor share?

Compare paying someone $50K/year to do a job as your total cost to paying that same person to do the same job $100K due to the additional regulatory, benefit and tax burden. Suddenly capital-intensive alternatives that can run 24×7, replacing 3 workers, start to look much more appealing.

21 Hernando Zuleta September 9, 2013 at 1:57 pm

The natural explanation is factor saving innovations.
For the last 10 years I have been writing about that:

and the story becomes more clear once you sparate raw labor share from human capital share:

Comments on this entry are closed.

Previous post:

Next post: