Human Capitalists

That is the title of a new and important paper by Andrea L. Eisfeldt, Antonio Falato, and Mindy Z. Xiaolan.  It seems that perhaps the share of labor in gdp has not fallen much after all:

The widespread and growing practice of equity-based compensation has transformed high-skilled labor from a pure labor input into a class of “human capitalists”. We show that high-skilled labor income in the form of equity claims to firms’ future dividends and capital gains has dramatically increased since the 1980s. Indeed, in recent years, equity-based compensation represents almost 45% of total compensation to high-skilled labor. Ignoring such income results in incorrect measurement of the returns to high-skilled labor, with important implications for macroeconomics. Including equity-based compensation to high-skilled labor cuts the total decline in the labor share since the 1980’s by over 60%, and completely reverses the decline in the high skilled labor share to an increase of almost 1%. Correctly measuring the return to high-skilled labor can thus resolve the puzzling lack of a skill premium in recent data, as well as the corresponding lack of evidence of complementarity between high-skilled labor and new-economy physical capital. Moreover, tackling the capital structure question of who owns firms’ profits is necessary to provide a link between changing factor shares and changing income and wealth shares. We use an estimated model to understand the rise of human capitalists in an economy with declining capital goods prices. Finally, we present corroborating cross section and time series evidence for complementarity between high-skilled labor and physical capital using our corrected measure of the total return to human capitalists.

Since smart people are bearing more and more risk, this may be another reason why income inequality is rising.

Via the excellent Kevin Lewis.


Labor that owns equity is the right mix that both capitalists and socialists can agree on. Yet worker ownership of the means of production is still so rare.

Well when we're talking about "equity-based compensation" being paid to labor, we might see a few shares of stock or options being earned by a high-skilled programmer. But that's surely dwarfed by the 7- and 8-figure stock payouts to top business executives that we constantly read about. That's compensation to high-skilled labor, but I don't think a multi-million stock option package to a CEO is what advocates for labor have in mind when they say that labor ought to get a greater share of national income.

More common than programmers are skilled professionals that form firms like they do in accounting, engineering, law, or medicine for the white collar set or electricians, plumbers, and carpenters for the blue collar set. These arrangements may not trade on a public stock exchange but it can pay dividends or pass-throughs and certainly qualifies as equity.

How "skilled" must a skilled professional be in order to receive compensation that rivals or perhaps surpasses the total economy of a small Caribbean island nation or Central American backwater? What is this individual's particular skill that makes him so valuable?

Making the business successful and keeping socialism and liberalism from destroying it.

"What is this individual's particular skill that makes him so valuable?"

He found someone willing to pay him that much.

While ESOPs theoretically improved worker performance, they have performed dismally at doing so. The free rider problem is just too powerful to overcome. Bonuses tied to individual performance does much better, to the extent performance can be adequately measured.

Workers owning shares in their own company violate a several financial investment guidelines. First, it puts their eggs in one basket such that they risk both their jobs and their capital if the firm fails. Second, owning one firm's stock is seldom good diversification, and workers in ESOPs seldom own anything outside their firm's stock. Third, trading restrictions are economically suboptimal. Fourth, share ownership is unstable; workers generally sell the shares when they can.

But it does make wage compensation a lot more flexible. One of the biggest issues in downturns is that wages are sticky. Bonuses can be cut much easier. So this makes the firm much more stable, instead of cutting jobs and losing valuable skills, they can cut bonuses and be better able to ramp up once the economy starts to recover.

I agree with performance based compensation (bonuses). I've worked that way for decades.

I'm skeptical of stock options as compensation. They have advantages and disadvantages for the workers and don't provide firms with all the benefits they expect. They shift risk onto workers, as does bonuses. But workers paid mostly by bonuses usually can't survive a bad downturn so it ends up reducing staffing anyway.

Stocks are relatively less liquid than bonus income. Workers earning bonuses need to increase their savings rates to withstand downturns and plain bad luck.

Isn't the big point of stock options to get people to stick around til they vest and not go chasing a 2% raise at a crosstown rival firm?

That's possibly one aspect of the incentive.

I'm considering not just the employer's motivation for using them but the economic effects on the employees.

Depends on how rapacious the competition is. That can be considerable in financial advising and tech.

Getting back to the original article, the issue is accurately measuring compensation. One problem is that making an adjustment for stock compensation would over count the gains and fail to count the risk associated with the gains.

"and workers in ESOPs seldom own anything outside their firm's stock. "

Yes, but how many of those workers would own any stock absent their ESOP?

Isn't "Labor that owns equity" the same thing as "worker ownership of the means of production"? Are you implying that total ownership by labor is the real goal?

How can you possibly lay any claim to ownership to the means of production, if you have not contributed capital to acquire the assets or alternatively you have not assumed the risk in securing loans to acquire the assets?

How did Jobs and Wozniak do it when they were broke 20-somethings working out of a garage? The same answer as the article title. Human Capital.

Employers have at least a titular claim to the skills of their employees. They rely on various kinds of formal and informal commitment mechanisms.

"If human capitalists leave to start a new firm, we assume that they will also receive their flow wage whh"

This is merely assumed to make the participation constraint nice; but it is a very heavy-handed economic assumption and I can't help but suspect that it's largely driving the results. That is, any improvement in labor compensation is essentially assumed to be of the "human capital" component, not the wage component, so no wonder it has a large presence in the model.

Isn't that a reasonable assumption given "If human capitalists leave to start a new firm..."? Or am I misunderstanding the import?

Why would it be reasonable? I don't see any ex-ante reason why someone would receive precisely the same income flow if they were to start their own firm. Overall expected compensation would be higher, otherwise they wouldn't bother starting their own firm in the first place, but how that compensation is divided between income flow and equity could be any linear combination. They simply assume one linear combination in particular because the math works out nicely.

If we aren't accounting for compensation in the form of equity claims, aren't we also understating GDP and productivity gains?

Nah, in practice GDP is typically measured via the expenditure approach, not factor income.

Stock based compensation IS accounted for. Amazon famously paid no taxes recently because of deferred tax assets created by such compensation. The accounting isnt the issue. The timing is.

Relying on rising asset prices for prosperity has all kinds of ramifications, including stock-based compensation. When the key employees total compensation is based in large part on rising stock prices, there's a built-in incentive for the executives to juice-up the stock price. Whether that results in short-termism is debatable (the conventional wisdom is that it does), but it definitely skews decision-making to what will juice-up the stock price. Did the executives at Lyft and Uber juice-up the stock prices? The high initial trading prices created vast wealth even though the holders of the options likely were unable to exercise their stock options. Stock options usually come with a price: you leave you lose. This helps lock-in key employees because they don't want to lose their options. Non-qualified stock options (the typical kind granted only to key employees) also come with a price to the key employees when the option is exercised: they realize ordinary income (not capital gain) when they exercise the option measured by the difference between the value of the stock at the time of exercise and the option price. Ouch! And this is one case where the tax law provides symmetry: the company gets a deduction in the same amount even though the company has no cash outlay. That's partly why companies like Amazon, etc. have such large tax losses and tax loss carryovers.

And then there's the point I have tried make again and again: relying on rising asset prices for prosperity creates the conditions for financial and economic instability. Here today, gone tomorrow. What is "value", what is "economic growth", when it depends ever increasing asset prices rather than earnings. It's an illusion. But much in life in America today is an illusion, isn't it?

Additional reading:

You have a valid point on the "time-honored" corporate finance practice of corporations' stock repurchases pumping share values. As always, caveat emptor.

Make no mistake, ten-plus years of near-zero US Fed economic engineering/management (and sub-zero BoJ and ECB) exacerbate risk asset price inflation.

L. Frank Baum wrote about one "Emerald City" and conman pulling levers behind a green curtain. Now, we actually have (at least four) pulling economic engineering/management "levers."

Sorry, "The Atlantic" isn't my go-to source for economics, or any other topic.

Not to worry, if the next financial crisis strikes in this Congress, the Crapo-Waters Bill will hand over more authority and power to the Fed.

Re stock buybacks: you do realize Rayward they are the same as dividends, arguably even better?

Agreed. If the execs can't figure out something more profitable to do with the cash(!), then returning capital to me, the shareholder, is the correct move. And the impact is only mildly different if it is returned via dividend or repurchase/buyback.

Uber Stock was priced at $45, opened at $42 and has been between $43-$44 almost the entire time it's been traded.. seems like a poor example of a "juiced-up" stock.

For a company that loses over $1 billion per year and no projection of ever turning a profit?

Partying Like It's 1999!

You talked about a high initial trading price and a jumped up stock.. it's basically been flat since they went public, would seem to suggest it was priced appropriately.

People have had plenty of time to look at their books at this point, unless you're suggesting they outright committed fraud, it doesn't seem to be the case the price was greatly inflated in any way.

Universal Basic Equities?

Pretty much what Chile did in privatizing SOEs and their pension system. It distributed shares of SOEs rather broadly in their second wave, and required retirement savings in approved retirement funds with relatively safe investments.

We should have done that here a long time ago.

Back in the 1950s, employee ownership plans where, like Avis (?) rental car, the employees own the company they work for, were considered "communist", wrongly IMO.

As for the OP, the fact that more than 50% of company costs are due to wages and bonuses might be skewed by the rise of the finance sector, from single digits to double digits share of GDP, since it's routine for Wall Street firms to pay something like 75% of their costs in bonuses and salaries.

".. the employees own the company they work for, were considered "communist", wrongly IMO."

You're an idiot. ESOPs were very rare up until Congress passed a law in 1974.

"Senator Russell Long, a Democrat from Louisiana, helped develop tax policy for ESOPs within the Employee Retirement Income Security Act of 1974 (ERISA), calling it one of his most important accomplishments in his career."

Isn't stock based compensation just an artifact of tax laws (deferred taxation)?

Artifact? No.

Deferred tax implications? Yes.

Stock options are great for companies that are initially capital-starved but with optimistic prospects. They are a reward for employees who share the company risks. They provide incentives for hard work, albeit imperfectly.

In Germany stock options are awarded exceedingly rarely. Performance based personal bonuses are way more frequent. But then, they do tax all unrealized capital gains on options every year. A popular worst case - imagine your options are worth $100 y1, then $300 y2, then $50 when you cash in. You get $50. But you did pay $100x25% in taxes first year and $200x25% in the second year, $75.

Perhaps I overstated my case. They certainly have some tax benefits, but I think these would exist in the absence of such benefits. That's why I disagreed they are an artifact of tax laws.

There are other differences between the US and German labor markets besides tax laws. I see options being used more broadly in highly speculative firms like tech startups.

For top executives they are a strong incentive to align management objectives with those of shareholders, overcoming some principal agent problems.

Germany is a lot more conservative in business and finance. I'll admit to being wrong though if you push back from first hand experience.

"A popular worst case - imagine your options are worth $100 y1, then $300 y2, then $50 when you cash in. You get $50. But you did pay $100x25% in taxes first year and $200x25% in the second year, $75."

That seems nutty.

Sure. But it's just as nutty as making long term capital gains (e.g. from options) tax free or nearly so. If you pay 50% tax on current income but zero on deferred pay, you know which you would choose.

Human capitalist is like a shareholder with no rights =)

I thank Mr. Johnson for his leadership.

Another installment of Fun with Imaginary Numbers.

Every single variable quoted (including the parting "income inequality") is a 3rd+ degree derived variable with tons of underlying assumptions, many of which change over time and can't be adequately controlled for.

It's great that people find ways of padding their CVs with this work, but let's not take it for anything other than academic masturbation, lest some innumerate mainstream media types get hold of it and start making unfounded extrapolations to fit their view of the universe.

When the public discusses "diminishing returns to labor's share", these are not the type of people they are talking about. White collar, high-skilled workers who are high up enough to be given stock options are not the poster children for income inequality.

The research appears to be valuable, but the way that this post is presented looks like a thinly veiled attempt to prove that "hey it's not so bad for workers after all!", when that clearly is not the case for the vast majority of the working and middle class.

Don't worry. The numbers are all fake anyway, so you can massage them to say anything you want.

You seem to want them to say that it's pretty hard for workers. There are numbers to show that. Others want the numbers to say something else. There are numbers to show that, too.

Congratulations. We're all correct!!!

I think that income and wealth inequality are a problem. In order to adequately describe the problem, you need a sense of how unequal things are currently, taking the entire workforce into account.

I think it's an issue with proportions. If by moving the top 5% of income earning workers into the "capital" category due to the fact that they receive equity options, and that dramatically increases the returns to capital, all that means is that there is a huge schism between those top income workers and everyone else - which we already knew.

What would be more helpful in a conversation about inequality or the returns to labor, is why aren't more companies incentivizing new employees with equity options? Wage increases have been marginal for some reason, and I'm not entirely convinced by the Baumol effect.

I think most people are aware of this and treat equity returns as returns to capital. I’ve read articles that explicitly say executives and other highly placed employees should be treated as capitalists based on their stock options. And of course many people including myself are annoyed by how certain extremely highly placed workers can reclassify their labor income as capital through the carried interest rule.

Of course that just goes to show you that the labor/capital dichotomy is pretty arbitrary and meaningless today; the dichotomy is correlated between rich and poor, but if you think the problem is the rich have too much and the poor have too little, just say that instead of saying capital has too much and labor too little.

The distinction between capital and labor is a bit contrived. Capital itself is built with labor. Similar laws of supply and demand govern labor and capital. The distinctions appear to be social, not economic.

"Indeed, in recent years, equity-based compensation represents almost 45% of total compensation to high-skilled labor. "

Gee, I thought I was high-skilled, but I'm not getting any equity based compensation. Maybe I'm not so skilled after all. The paper is not clear who counts as high-skilled, but it must not be very many people, so they can have a high figure like 45%.

"Correctly measuring the return to high-skilled labor can thus resolve the puzzling lack of a skill premium in recent data . . ."

Might be nice if smart people would bear more and more risk for getting their measurements right the first time, or maybe the second or third times.

On what approximate scale do measurement errors commonly occur today (by industry or sector, let's say)?

Do the multitude of these errors occur because the overwhelming amounts of data available today cannot be sorted or digested properly?

--or are measurement protocols failing to assess properly what they purport to measure because what ostensibly is being measured fails to conform to whatever "rational" categories control the formulation or construction of the measurement protocols? (That is: is our insistence upon all data's [ESPECIALLY "human data"] conforming to our "rational" measurement protocols and expectations itself far from rational? Has measurement mania become so irrational as to've become pathological for one and all?)


This is completely obvious to anyone working in IT. It is also not restricted to increase in wages. Some companies use stock options to keep employees around as well (by spreading out vesting dates into 3/5 year periods). This is a BIG deal for IT employees, and it can end up impacting your yearly earnings in 30% to 50%.

if labor's share is 55.8% ( first quarter actual) and you add 10 to both the numerator and denominator for stock options, labor's share only goes up to 59.8%, less than half the 10 increase.

So, if I am paid a salary in cash and buy stock in my firm and the stock increases that increase counts as labor income

These comments are missing an important factor, appetite for risk. Each individual bet doesn't need to pay off, but making lots of bets increases the chances that one will. So a person who in yesteryear would have worked up a corporate ladder somewhere, instead builds some expertise then takes a leap to a start-up. The start-up has some salary (if they can make payroll, not guaranteed) and an equity claim that is unlikely to pay off. A year in, the start-up is shaky, wants to cut payroll, and the investors/board members move the talented people to another start-up in their portfolio - another shot at a positive equity event.

Each start-up, failure or not, adds to a person's desirability - particularly if they were not the founder, but "hired help" with a couple percent and actual skills. I've seen this not just with programmers, but finance/accounting, marketing in several subcategories, sales, at least.

After each go-round, the person has the option to exit to a large company at a significantly higher salary. One reason, large companies read blogs and think they need to be like start-ups, so what better strategy than to hire start-up veterans. Second reason, being at a start-up is a massively accelerating skill building opportunity - one has to literally everything in their area, from strategic decisions to the minutiae. Particularly for software, spending too much time in meetings or drawing boxes on whiteboards makes Jack a dull boy, so a tour actually writing code for 90 hrs/wk gets one up to what the cool kids are doing post-haste.

Note that actually being in a startup, even with skill, is a risky proposition. The salary is usually lower than one could get in a corporate gig, paychecks can be missed - so it is tempting, and often not optional, to pause or draw down retirement savings, run up credit cards, etc, aka acquire risk. Not everyone has this appetite. But those that do can reap rewards other the rare actual equity event.

Comments for this post are closed