Rewarding the Entrepreneur

I want to take seriously an argument that deserves to be taken seriously.

When I wrote "The Arithmetic of Cheap," I traced where the money goes in a supply chain — and the numbers were damning. The response, from some quarters, was predictable: "But entrepreneurs deserve to be rewarded for risk. They bet everything. They built something from nothing. The market rewards that."

Fine. Let's do the arithmetic on risk.

***20:1***

In 1965, according to the Economic Policy Institute's analysis of executive compensation, CEOs of America's largest companies earned about twenty times the pay of a typical worker.¹ The EPI's methodology tracks CEOs at the top 350 firms against average worker compensation in those industries — a consistent benchmark across decades, though not a firm-by-firm comparison.

Peter Drucker, the management theorist who advised some of the most successful corporations of the twentieth century, argued that anything beyond 20 or 25 to 1 was actively harmful — that it bred resentment, eroded teamwork, and ultimately damaged the enterprise itself. In a 2011 SEC comment letter, the Drucker Institute cited his position that excessive pay ratios lead to resentment and falling morale.²

Twenty to one. The ratio of the post-war era.

***399:1***

By 2021, that ratio had become 399 to 1.

The Economic Policy Institute found that the average CEO of a top-350 firm made $27.8 million in 2021 — approximately 399 times the annual compensation of a typical worker in the same industries (using their "realised compensation" measure, which counts stock options when exercised). From 1978 to 2021, CEO compensation rose 1,460 percent in inflation-adjusted terms. Typical worker compensation, over the same period, rose 18 percent.¹

Let me say that again: 1,460 percent versus 18 percent.

The question is not whether this happened. The question is why. And here is where the "reward for risk" narrative begins to collapse under the weight of its own arithmetic.

***The Policy Ledger***

The shift from 20:1 to 399:1 was not organic. It was not the invisible hand of the market recognising that executives had become twenty times more valuable. It was the result of specific policy changes, made by specific people, in specific decades.

Start with taxes. According to the Tax Policy Center, the top marginal federal income tax rate was 70 percent in 1980;³ according to IRS inflation adjustments for tax year 2025, it remains 37 percent.⁴ Research by Piketty, Saez, and Stantcheva, published in the American Economic Journal: Economic Policy in 2014, found a negative correlation between top tax rates and executive pay across countries — though, as with any correlation, this does not establish causation.⁵

Then there's a rule that was supposed to fix the problem. In 1993, Congress passed IRC §162(m), which limited the tax deductibility of compensation above $1 million for covered executives at publicly held companies. The idea was to discourage excessive pay.

But the rule exempted "performance-based" compensation — a category that included stock options and certain bonuses.

Stay with me. This next part matters.

***The Warning***

In 1991, shortly after launching his presidential campaign, Bill Clinton read an influential book on corporate overcompensation by compensation consultant Graef "Bud" Crystal. Crystal was not an activist; he was an industry insider who had spent decades advising corporations on executive pay.

Clinton called Crystal for his thoughts on using the tax code to curb excessive pay.

Crystal told him, according to multiple accounts: "Utterly stupid."

He warned Clinton, as ProPublica later documented, that the proposal "not only wouldn't hold down executive pay, but would hurt shareholders by increasing the after-tax cost of CEO pay packages."⁶ He knew what would happen: companies would simply restructure compensation to take advantage of whatever exemptions existed.

Crystal was not alone. As ProPublica documented, Representative Martin Sabo of Minnesota had proposed something far more ambitious in 1991: the Income Disparities Act, which would have denied corporations a tax deduction if executive compensation exceeded twenty-five times the pay of the lowest-paid employee.⁶ Not five executives. All employees. A structural limit.

The proposal never gained legislative traction. The proposal died.

What passed in 1993 was something far narrower: a $1 million cap affecting only five named executives at publicly traded companies. And it included the performance-based exemption.

The consultants knew. The critics warned. The stronger version was killed. And then Congress passed a law with a loophole wide enough to drive a Brinks truck through.

***The Predictable Result***

What happened next was not an accident. It was the inevitable consequence of a system designed to appear restrictive while preserving unlimited upside.

A ProPublica and Washington Post investigation compared compensation at forty large companies in 1992 — the year before the law took effect — with 2014.⁶ The results are worth presenting in full, because the scale is difficult to grasp otherwise.

Compensation subject to the deductibility limits — the kind of pay the law was supposed to constrain — rose 650 percent, from $1.1 million to $8.2 million per executive.

Compensation exempt from limits — the performance-based kind — rose from $970,000 to $4.4 million, an increase of 350 percent.

In 1992, thirty-five percent of executives at these companies earned more than $1 million. By 2014, ninety-five percent did.

Let me give you some specific names, because abstraction is the enemy of understanding.

Oracle: CEO compensation went from $4.9 million in 1992 to $119.4 million in 2014. Allergan: from $378,000 to $77.4 million. Cisco: from $1.1 million to $75.2 million. Apple: from $5.7 million to $114.6 million.⁶

By 2021, stock awards and exercised options made up over 80 percent of top CEO pay.¹ The performance-based exemption had become the compensation structure.

As ProPublica reported, Steven Seelig, executive compensation counsel for Willis Towers Watson, described how the deductibility cap shaped corporate decisions: "Decisions on the pay mix are not guided by the deductibility factor...the decision on the pay mix that's appropriate is guided by their companies' unique circumstances."⁶

The effect was straightforward: the deductibility cap became one factor among many in compensation design, and the performance-based exemption provided a well-worn path around it.

***The Geometry of the Trap***

Here is what makes this Master Key evidence rather than merely damning data: the flaw was not discovered after the fact. It was warned about before passage. It was designed in through political pressure. And it worked exactly as the critics predicted it would.

The law was not an earnest attempt that failed. It was a political compromise that preserved the appearance of action while protecting the substance of extraction.

Consider the structure: Congress could not ignore rising executive pay — it was politically untenable in 1993. So they passed a law. But the law had to be acceptable to the corporations who would be affected. So it was narrowed (only five executives) and neutered (performance-based exemptions).

The result was a machine with three gears that interlock perfectly:

First: a visible limit ($1 million) that creates the appearance of constraint.

Second: an exemption (performance-based pay) that allows unlimited compensation through a different door.

Third: a consulting industry that immediately began restructuring compensation to flow through that door.

The trap is elegant because each actor behaves rationally within it. The politician gets to claim they addressed executive pay. The corporation shifts to stock-based compensation and calls it "pay for performance." The consultant collects fees for restructuring packages. The executive gets richer. The shareholder — who actually bears the cost when compensation isn't deductible — loses, but is diffuse and disorganised.

This is not a bug. This is the system working as designed by people who understood exactly what they were designing.

***The Risk Ledger***

But perhaps, you might argue, the people at the top really do bear more risk. They stake their fortunes; workers merely collect wages.

Let me introduce you to the garment workers of Bangladesh, Cambodia, India, Indonesia, Sri Lanka, Myanmar, and Pakistan.

When the COVID-19 pandemic hit in 2020, fashion brands cancelled or refused to pay for orders from their suppliers. Many of these orders were already in production.

The Clean Clothes Campaign calculated that from March 2020 to March 2021, garment workers in these seven major Asian production countries were owed $11.85 billion in unpaid wages and severance.⁷ An estimated 1.6 million workers in those countries were dismissed, many without their legally mandated severance pay. During lockdowns, workers were often paid only a small percentage of their usual wages, leading to widespread hunger and debt.

Who bore the risk? Not the brands — as the Clean Clothes Campaign documented, "many brands have returned to profitability and some have even raked in record level earnings" by 2021, while workers remained unpaid.⁸ The risk was borne by the people who could least afford it — the workers who had already done the work, whose labour had already created the value, and who were simply not paid.

This pattern is not unique to fashion. When startups fail — and failure rates vary widely by definition and time horizon, but are substantial — employees typically lose their jobs, often with no severance in jurisdictions like the United States where it is not legally required. Their unvested stock options are forfeited. Their vested shares may become worthless if the company liquidates. The founder's equity is also worthless in such cases, but founders with strong networks may find new opportunities — getting hired by acquirers or raising funds for new ventures. Institutional investors spread risk across portfolios designed to absorb individual failures.

In both global supply chains and startup ecosystems, the people with the least power often absorb the greatest losses.

***The Motivation Threshold***

Perhaps, though, the promise of extreme wealth is necessary to motivate entrepreneurs in the first place. Without the prospect of billions, why would anyone start a company?

The research on this is surprisingly clear — and it does not support the intuition.

A study by Amit, MacCrimmon, Zietsma, and Oesch, published in the Journal of Business Venturing in 2001, surveyed technology entrepreneurs and found that "wealth attainment was significantly less important to entrepreneurs relative to other decision dimensions" — dimensions like autonomy, innovation, and challenge.⁹ The entrepreneurs in the study did not rate personal wealth goals as more important than non-entrepreneurs did. Many of them started companies even while holding lower expectations about the probability of attaining wealth through entrepreneurship compared to traditional careers.

The researchers concluded: "We have evidence of one group of high-tech entrepreneurs who achieved success without placing much decision weight on attainment of personal wealth."⁹

What motivates founders, it turns out, is not the difference between $100 million and $10 billion. It is the work itself — the creative challenge, the independence, the chance to build something.

The marginal utility of extreme wealth, as a motivational tool, appears to diminish sharply. The prospect of your ten-thousandth million does not make you work harder than the prospect of your hundredth.

***The Threshold***

Here is where "reward" becomes something else entirely.

In the 2012 U.S. federal election cycle, according to the Sunlight Foundation, just 31,385 people — the top 0.01 percent of the population — contributed 28 percent of all disclosed individual political contributions.¹⁰ Every winning member of Congress received funding from this donor class. Among winning candidates, 84 percent took more money from this elite donor class than from all small donors (those giving $200 or less) combined.

At a certain level of wealth — roughly the high tens of millions and above — money stops being compensation and starts being structural power. It buys access. It shapes which candidates are viable. It funds the lobbyists who write the rules.

The pharmaceutical industry offers a case study. In 2003, Congress added a "noninterference" provision to Medicare Part D that barred the government from negotiating drug prices directly (though Part D plans themselves negotiate with manufacturers; the 2022 Inflation Reduction Act partially reversed this, allowing Medicare to negotiate prices for select drugs starting in 2026). According to a 2009 ProPublica investigation, at least 25 congressional staffers, members, and government officials who helped craft the Medicare drug legislation later became pharmaceutical industry lobbyists.¹¹ Representative Billy Tauzin, who helped pass the law, left Congress to become president of PhRMA, the drug industry's lobbying arm, at a salary reportedly more than ten times his congressional pay. According to a 2020 study in JAMA Internal Medicine, the pharmaceutical and health products sector spent more than $6 billion on federal lobbying and campaign contributions combined from 1999 to 2018 — more than any other industry.¹²

This is not reward for risk. This is the conversion of wealth into the power to set the rules of the game — rules that, conveniently, tend to generate more wealth for those who already have it.

The Levers

***Your Ratio***

Let's make this personal. I want you to do some arithmetic.

Take your annual income. Divide it by 399. That's what you would earn if you were the "typical worker" to someone earning what you earn as if they were a CEO.

Now multiply your income by 399. That's what someone would earn if you were their typical worker and they captured the same ratio.

If you earn $50,000, your 399:1 CEO equivalent earns $19.95 million. If you earn $100,000, they earn $39.9 million.

Here is the question the arithmetic forces: where do you sit in the chain? Because there is almost certainly someone below you — a person who made your clothes, picked your food, assembled your phone — whose labour you benefit from at a ratio you have never calculated. And there is almost certainly someone above you capturing value from your work at a ratio you have never been shown.

This is not a system that exists over there, in the boardrooms, among people who are not you. You are in it. You have a ratio. You are someone's 1, and you are someone else's 399.

The question is not whether entrepreneurs deserve reward. The question is whether the word "reward" can survive contact with the arithmetic of who actually bears the cost.

***The Arithmetic***

So let's return to where we started. "Entrepreneurs deserve to be rewarded for risk."

The arithmetic says something different.

It says the ratio of reward shifted from 20:1 to 399:1 not because entrepreneurs became twenty times more valuable, but because policy changed. Tax rates fell. Stock-based compensation exploded through a loophole that was warned about before passage, watered down from a stronger proposal, and then worked exactly as predicted. Union membership collapsed.¹³ The rules were rewritten, and the rewriting had authors.

It says those authors were warned. Compensation consultant Graef Crystal told Clinton the plan was "utterly stupid" and would fail. Representative Sabo proposed a structural limit that could have worked. It was killed by political pressure. What passed instead was a law designed to appear restrictive while preserving unlimited extraction.

It says the people who bear the most risk — the garment workers owed $11.85 billion, the employees whose unvested options vanish when startups fail — are not the people who capture the reward. Risk flows downward. Reward flows up.

It says the research on what actually motivates entrepreneurs points away from extreme wealth and toward autonomy, challenge, and the work itself. The difference between $100 million and $10 billion does not make founders work harder. It makes them politically powerful.

And it says that beyond a certain threshold, "reward" stops being compensation and becomes something else: the ability to shape the rules that determine future rewards. Lobbying. Campaign contributions. The revolving door between regulator and regulated. The game is not played on a level field because the winners of previous rounds get to tilt the next one.

I am not arguing that entrepreneurs should not be compensated. I am arguing that the phrase "reward for risk" has become a story we tell to avoid examining the arithmetic. It flatters. It reassures. It asks no follow-up questions.

But the numbers are not flattering. The numbers say we built a system that socialises risk and privatises reward — that asks workers and taxpayers and the environment to absorb the downside while funnelling the upside to those who already have the most.

And the numbers say the system works this way because it was designed this way. By people who were warned. Who chose this. Who knew.

The next time someone tells you entrepreneurs deserve their billions because they took risks, ask them: whose risk? And show them the ledger.

The arithmetic is patient. It will wait.

Economic Analyst — YAN Consumer Intelligence

What Would Change This Analysis

Three categories of evidence would modify or overturn the core argument of this report.

First, the policy-causation link. This analysis maps the correlation between specific policy changes — the 1993 IRC §162(m) performance-based exemption, top marginal tax rate reductions, declining union density — and the expansion of the CEO-to-worker pay ratio from 20:1 to 399:1. If longitudinal research demonstrated that the ratio expansion occurred independently of these policy shifts — for instance, if countries that maintained high top marginal rates and strong union membership experienced comparable ratio growth over the same period — the policy-design argument would weaken substantially. The cross-country data in Piketty, Saez, and Stantcheva (2014) suggests this did not happen, but a larger dataset covering more economies and longer time horizons could complicate the picture.

Second, the motivation research. The Amit et al. (2001) finding — that technology entrepreneurs did not rate wealth attainment as a primary motivator — is drawn from a specific population (high-tech founders) in a specific period. If comparable studies across different sectors and decades showed that extreme wealth incentives were in fact a decisive factor in entrepreneurial entry — that founders demonstrably would not start companies without the prospect of compensation ratios above, say, 100:1 — the "motivation threshold" section of this report would need revision. The current evidence base is narrower than the claim it supports.

Third, the risk distribution. The report argues that risk flows downward while reward flows upward. If major employers published audited data showing that executive compensation clawback provisions were routinely enforced — that executives at failing firms reliably absorbed losses proportionate to their compensation premiums — the asymmetry documented here would narrow. Similarly, if the garment industry's post-pandemic wage recovery data, once fully compiled, showed that the $11.85 billion wage gap identified by the Clean Clothes Campaign was substantially closed through brand-funded restitution, the "risk ledger" section would require significant qualification.

None of these counter-scenarios is implausible. Each names evidence that is obtainable and would genuinely change the analysis if it materialised.