The math is not subtle. While the lowest-paid workers in America saw their real wages fall in 2025, the companies employing them spent more than $1 trillion buying back their own stock at the fastest pace ever recorded.
That is not a coincidence. It is a choice.
A Record That Nobody Celebrated on the Factory Floor
By August 20, 2025, U.S. companies had crossed the $1 trillion threshold in stock buybacks faster than at any point in history. S&P 500 firms set a quarterly record in Q1 alone, $293.5 billion in share repurchases, a 20.6% jump from the prior quarter. Goldman Sachs tracked authorized buybacks surpassing $1.2 trillion through October, a 15% increase over the same period in 2024. When the full-year accounting was done, the number was on course to exceed $1.1 trillion, an all-time high.
Apple announced a $100 billion buyback program in May. Alphabet committed $70 billion. Nvidia, sitting on $57 billion in cash, announced $60 billion in repurchases. Six of America’s largest banks, JPMorgan, Goldman Sachs, Wells Fargo, Bank of America, Morgan Stanley, and Citigroup, all ranked among the top 10 repurchasers. Goldman alone bought back shares equal to 18.1% of its own market value.
Meanwhile, at the bottom of the wage ladder, real earnings were moving in the opposite direction.
The Economic Policy Institute found that the real, inflation-adjusted wages of low-wage workers declined 0.3% in 2025, a sharp reversal from five consecutive years of meaningful gains. It was not a market inevitability. The EPI was explicit: the reversal “was caused by policy decisions that weakened the labor market.” The money existed. The question was always where it went.
The Mechanism Nobody Wants to Name Out Loud
To understand the buyback economy, you need to understand a single regulatory decision made in November 1982. That month, the SEC under Reagan-appointed chairman John Shad adopted Rule 10b-18, which gave corporations a legal safe harbor to repurchase their own shares on the open market without being charged with stock price manipulation. Before that rule, such activity was effectively prohibited. After it, the floodgates opened.
The economist William Lazonick, who has studied this machinery for decades, has called the decades of buybacks that followed nothing less than “the legalized looting of the U.S. business corporation.” By 1997, buybacks had surpassed dividends as the dominant form of returning capital to shareholders. The logic of the modern public company had been fundamentally rewritten away from long-term investment and workforce development, toward quarterly earnings management and stock price maintenance.
The mechanism works like this. When a company reduces the number of shares outstanding, each remaining share represents a larger slice of earnings. Earnings per share rise even if the underlying business hasn’t grown at all. That number drives CEO bonus calculations. It influences analyst ratings. It moves stock prices. And stock prices determine the value of the equity compensation that constitutes the majority of pay for senior executives.
In other words, buybacks are not just capital allocation decisions. They are compensation decisions. Just not for the workers.
The Gap Has a Face — Several of Them
The Institute for Policy Studies analyzed the 100 largest low-wage corporations in the S&P 500 in its 2025 Executive Excess report and found that the CEO-to-worker pay ratio had ballooned from 560-to-1 in 2019 to 632-to-1 in 2024. That widening happened not because workers were paid less in nominal terms; their wages did technically rise, but because CEO compensation, overwhelmingly tied to stock performance, rose far faster. Average CEO pay at these firms climbed 34.7% between 2019 and 2024. Average median worker pay rose 16.3%. Inflation over that period was 22.6%.
The workers lost ground. The CEOs did not.
Starbucks set a record that should require no further comment. CEO Brian Niccol’s total compensation package came to $95.8 million in 2024. The company’s median employee earned $14,674. That is a ratio of 6,666 to 1. Niccol’s pay was not a salary; it was almost entirely stock options and performance awards, the very instruments that buybacks inflate. As the IPS noted, the company’s median pay rose just 4.2% between 2019 and 2024, during a period when more than 570 U.S. Starbucks locations unionized, in part to protest exactly this dynamic.
At Lowe’s, CEO Marvin Ellison took home $20.2 million in 2024, 659 times the company’s median worker pay of $30,606. The company spent $46.6 billion on stock buybacks between 2019 and 2024. The IPS calculated that the same sum could have funded an annual bonus of $28,456 for every one of Lowe’s 273,000 global employees for six years.
Those bonuses were not paid.
They Have a Point, Partially
Before dismissing every defender of the buyback model as a Wall Street apologist, consider the strongest version of their argument. It deserves an honest hearing.
Buybacks are not inherently a theft from workers. In a company where genuine investment opportunities are limited, expanding capacity would simply destroy value. Returning cash to shareholders is arguably the rational thing to do. Capital that flows back to investors can, in theory, be redeployed toward faster-growing sectors of the economy, creating jobs and opportunity elsewhere. The economist’s term for this is allocative efficiency.
There is also something to the argument about flexibility. Unlike dividends, which create an expectation of recurring payments, buybacks are discretionary. A company that commits to a buyback can slow or pause it. A company that commits to higher wages cannot easily cut them without triggering legal exposure, worker revolt, or reputational damage. The optionality, proponents argue, is genuinely valuable.
And it is true, at the level of macro-data, that the U.S. economy added jobs in 2025 even as buybacks surged. Low unemployment coexisted with record repurchases. The world did not end.
But here is what the efficient-markets version of this argument quietly ignores. The companies leading the buyback surge were not companies with limited investment opportunities. Apple, which spent $100 billion buying back its own stock, posted $94.9 billion in net income in fiscal 2024. It was simultaneously lobbying against regulations that would compel it to open its ecosystem to competition. Alphabet, which authorized $70 billion in repurchases, was cutting thousands of jobs. Goldman Sachs bought back 18% of its own market cap while presiding over a financial system that, as EPI data shows, channeled nearly all productivity gains to the top.
When the most profitable companies on earth cannot find a better use for $1 trillion than purchasing their own shares, the question is not whether the market is being efficient. The question is whose efficiency is being served.
What Productivity Actually Bought
The long-run numbers are damning in a quiet, bureaucratic way that tends to get ignored. According to the Economic Policy Institute, from 1979 to 2025, worker productivity in America increased 92.4%. Over that same period, average hourly compensation adjusted for inflation rose just 33.6%.
That gap does not represent market failure. It represents a sustained, systematic policy of wage suppression that predates the 2025 buyback surge by decades. The deregulation of corporate finance in the 1980s, the decline of union density, the erosion of the real minimum wage, and the adoption of at-will employment as the American norm were not economic accidents. They were choices, each of which shifted the balance of power between capital and labor in one direction.
The buyback surge of 2025 is not the cause of that imbalance. It is the expression of it. The trillion dollars is not new money extracted from workers. It is the annual dividend on four decades of policy architecture built to ensure that when productivity grows, shareholders collect the rent.
And now, with AI-driven automation beginning to hollow out middle-income roles, the same automation is reshaping how economies allocate labor across sectors. The structural conditions for wage suppression are deepening, not reversing. Productivity gains from AI will accrue first to the companies deploying the technology, then to their shareholders, and last, if at all, to the workers displaced or reskilled in the process.
The 1% Tax That Changed Nothing
Congress did try. The Inflation Reduction Act of 2022 introduced a 1% excise tax on net corporate stock buybacks, effective January 2023. The stated goal was to discourage excessive repurchases and nudge companies toward reinvesting in their operations and workers. The actual result: S&P 500 Q1 2025 buybacks hit a quarterly record. The 1% tax reduced operating earnings by 0.50% that quarter. It was a rounding error.
The political economy of buyback reform runs into a structural problem. The shareholders who benefit most from buybacks are institutional investors, pension funds, mutual funds, and index funds, which manage the retirement savings of tens of millions of ordinary Americans. When Apple buys back $100 billion in stock, and the price rises, the people who benefit include not just Tim Cook but every American with an S&P 500 index fund in their 401(k). The diffuse benefits make buyback reform harder to sell than its concentrated costs might suggest.
What that framing obscures is that the top 10% of Americans own roughly 93% of all stocks, according to Federal Reserve data. The 401(k) argument is real but mathematically small. For most working Americans, particularly those in the bottom half of the wage distribution who live paycheck to paycheck and have no meaningful investment portfolio, the rise in share prices generates no benefit. What they needed was a raise. What their employers chose was a buyback.
The Velocity of Inequality
There is a compounding logic to all of this that standard wage data tends to miss. CEO compensation tied to stock performance does not just grow, it accelerates. Between 2019 and 2024, the value of CEOs’ personal stock holdings at Low-Wage 100 firms grew more than three times as fast as their median worker pay. A bonus paid in restricted stock units in January is worth substantially more by December if the buyback program has been running in the interim. The executive knows this. The program is designed this way.
The workers at those companies are not passive observers. Over half of U.S. workers reported in 2025 surveys that they felt underpaid. Labor force participation dipped to 62.2% by mid-year, its lowest level since 2022. Layoffs were rising at twice the prior year’s rate, with AI adoption cited as a primary driver. The Bureau of Labor Statistics reported that annual wage growth in November 2025 was just 3.5%, the slowest since May 2021, while inflation continued eating into purchasing power.
The questions about how wealth concentrates in parallel economies when capital allocation systematically favors shareholders over labor are no longer abstract. They are being answered, in real time, by balance sheets.
What Has to Change — and Why It Probably Won’t Yet
The reform agenda is not obscure. Raise the excise tax on buybacks from 1% to something that actually creates friction, 4%, which the Biden administration once proposed, would at least register. Restrict buybacks at companies accepting federal contracts or subsidies. Require meaningful employee representation on corporate boards. Tie CEO pay ratios to procurement eligibility. These are not radical ideas. Germany, Japan, and most of northern Europe have long imposed structures that moderate the shareholder-first extremism that U.S. corporate law has cultivated since 1982.
But reform requires a political coalition that does not currently exist in its necessary form. The business lobby remains formidable. The ideological consensus that equates shareholder returns with economic health is embedded in business school curricula, earnings call language, and the daily grammar of financial journalism. The concept of demographic and economic decline accelerating when workers lose purchasing power is gaining traction in academic circles, but it has not yet translated into legislation with teeth.
What is shifting is the social temperature. The numbers are now large enough to be viscerally legible. A trillion dollars is not an abstraction. It is the number that your employer chose instead of your raise.
The First Sentence, Revisited
Corporate America spent over a trillion dollars in 2025 buying its own stock. The lowest-paid workers in that same economy saw their real wages decline. These two facts do not require a conspiracy to explain. They require only a system that, for four decades, has been built to produce exactly this outcome and a regulatory environment content to watch it happen, one quarterly earnings call at a time.
The silence is the message. A trillion dollars was spent. Nobody called it theft. Nobody called it anything. The press release said it was “returning value to shareholders.”
The workers got the memo.
