The 1973 Break: What Happened and Why the Lines Diverged

In Episode 1, we established the central fact: since 1947, nonfarm business productivity in the United States has risen roughly 420%, while real hourly compensation has risen roughly 150%. The two lines, which once climbed in near-lockstep, began separating sometime in the early 1970s and have drifted apart ever since. The gap between them is among the most consequential economic phenomena of the past half-century.

But why did it happen? And why 1973?

The answer is not a single event but a cluster of them — an oil embargo, the collapse of the postwar monetary order, the beginning of a decade of stagflation, the early stages of deindustrialization, and a set of policy choices whose consequences would take decades to fully materialize. The year 1973 was not the cause. It was the hinge.

This episode examines what the data looked like before the break, what happened during it, and what the three eras since tell us about the structural transformation of the American economy. The numbers are drawn from the Bureau of Labor Statistics nonfarm business sector data, indexed to 2017 = 100, using Q4 readings to smooth seasonal variation.

To understand what broke, you first need to understand what worked. The quarter-century from 1947 to 1973 was the most remarkable period of shared prosperity in American economic history. It was so extraordinary that economists have given it multiple names: the Golden Age, the Great Compression, the postwar boom. The numbers justify the superlatives.

In 1947, the BLS productivity index for the nonfarm business sector stood at 23.0 (on a scale where 2017 = 100). By Q4 1973, it had reached 45.6. That is a 98% increase in 26 years — a near-doubling of output per hour worked. The annual compound growth rate was 2.7%.

What made this era remarkable was not just that productivity grew fast. It was that compensation grew almost exactly as fast. Real hourly compensation — wages and benefits adjusted for inflation — rose at approximately 2.6% per year over the same period. The gap between the two growth rates was a negligible 0.1 percentage points per year.

Think about what that means in practice. When a factory worker in 1955 helped produce 3% more output than the year before, that worker’s paycheck grew by roughly 3% in real terms. The gains from rising productivity were shared almost dollar for dollar between capital and labor. An assembly line worker in Detroit, a clerk at Sears, a longshoreman in Oakland — all participated in a rising tide that genuinely lifted most boats.

Several structural factors sustained this equilibrium. Unions were powerful — roughly a third of private-sector workers were unionized in the 1950s and 1960s, giving labor substantial bargaining power. The Bretton Woods system anchored the dollar to gold and other currencies to the dollar, creating a stable international monetary framework. Global trade existed but was modest by modern standards; American manufacturers competed primarily with each other, not with factories in Shenzhen or Bangalore. Corporate tax rates were high, top marginal income tax rates were very high, and the political consensus favored broadly shared prosperity.

The economy was also structurally simpler. Manufacturing employed roughly a third of workers. The service sector was growing but had not yet come to dominate. The financial sector was small, tightly regulated, and boring — a utility, not an industry. Healthcare consumed roughly 5% of GDP, not 18%. Higher education was affordable. Housing was cheap relative to income.

None of this was utopian. Racial discrimination was pervasive and legally sanctioned. Women were largely excluded from high-paying occupations. Poverty rates, though declining, remained high by any reasonable standard. But for the workers who were inside the system — predominantly white men in unionized manufacturing — the compact between productivity and compensation held firm. You produced more, you earned more. It was that simple.

The year 1973 sits at the intersection of several tectonic shifts in the global economy. No single event explains the productivity-compensation break, but taken together, the events of 1971–1974 constitute the most concentrated period of structural change in postwar economic history.

The oil shock was not just an energy crisis. It was a profitability crisis. American industry had been built on cheap energy. When the price of that energy quadrupled, the capital stock — the factories, the machines, the transportation networks — became partially obsolete overnight. Equipment designed to operate with $3 oil was suddenly uneconomic at $12 oil. The investment needed to retool was enormous, and much of it was defensive — restoring productivity levels rather than advancing them.

At the same time, the regulatory environment was shifting. The early 1970s saw the creation of the Environmental Protection Agency, the Occupational Safety and Health Administration, and a wave of new environmental and workplace safety regulations. These were valuable on their own terms — cleaner air, safer workplaces — but they represented real costs that showed up as reduced measured productivity growth. A factory that spends 5% of its capital budget on pollution control equipment produces cleaner air but not more widgets per hour.

And beneath these visible shocks, a deeper transformation was underway. The low-hanging fruit of postwar industrialization had been picked. The productivity gains from electrification, from the interstate highway system, from the mechanization of agriculture, from the expansion of universal education — these one-time structural shifts had largely run their course by the early 1970s. The next wave of transformative technology — computing and the internet — was decades away.

The chart below zooms in on the critical fifteen-year period from 1968 to 1982 — the window that spans from the late Golden Age through the break and into the stagflation era. Two lines are plotted: nonfarm business productivity and real hourly compensation, both indexed to 2017 = 100.

Before 1973, the two lines move in tandem. They are not perfectly correlated — compensation dips slightly relative to productivity in some years, catches up in others — but the general trajectory is shared. A worker in 1968 looking at these two lines would see no reason to expect them to diverge.

Then watch what happens after 1973. Productivity stalls, dips, and then resumes growing — slowly, fitfully, but upward. Compensation, however, flatlines. From 1973 to 1982, real hourly compensation barely moves. The index goes from 66.5 in Q4 1973 to 70.9 in Q4 1982 — a gain of 6.6% over nine years, or less than 0.7% per year. Over the same period, productivity rose from 45.6 to 50.2 — a gain of 10.1%, or about 1.1% per year.

The gap had opened. Not dramatically — not in a way that would make headlines — but measurably, persistently, and as it turned out, permanently.

Several features of this chart deserve attention.

First, the pre-1973 synchronization. From 1968 to 1972, productivity rose from 40.9 to 45.2, and compensation rose from 62.0 to 66.8. Both lines were moving up at roughly the same pace. The compact was still holding.

Second, the 1973 productivity peak at 45.6. This was the last year that belonged unambiguously to the old regime. Productivity in 1974 actually fell — to 45.3 — the first outright decline in the postwar period. It would not recover to the 1973 level until 1976.

Third, the compensation freeze. Real hourly compensation peaked at 69.5 in 1977, dipped to 70.0 in 1979, and actually fell to 70.0 in 1981 before recovering slightly. For nearly a decade, the average American worker saw essentially no improvement in real compensation. Meanwhile, the cost of housing, healthcare, and education was beginning its long structural ascent.

Fourth, the asymmetry of the recovery. After the 1973–75 recession ended, productivity did resume growing — slowly, but it grew. Compensation barely participated. This was the first clear evidence that the postwar compact was dissolving. The economy was becoming more productive, but workers were not capturing the gains.

With the benefit of half a century of hindsight, we can now divide the postwar period into three distinct eras defined by the relationship between productivity growth and compensation growth. Each era has its own economic logic, its own structural forces, and its own implications for workers and investors.

Productivity growth: 2.7% per year. Real compensation growth: 2.6% per year. Gap: 0.1 percentage points.

This was the era of shared prosperity. The American economy was growing rapidly from the low base of the Depression and wartime economy. Massive infrastructure investment — the interstate highway system, suburban expansion, rural electrification — created productivity gains that were broadly distributed. Unions ensured that wage growth tracked productivity growth. International competition was limited. Corporate governance norms favored stakeholder capitalism over shareholder primacy.

The productivity index rose from 23.0 to 45.6 — a near-doubling. Real compensation rose from roughly 29 to roughly 67 on the same indexed scale — a more-than-doubling, reflecting the fact that the starting base for compensation was even lower. A factory worker in 1973 was roughly twice as productive as a factory worker in 1947 and was paid roughly twice as much in real terms. The system worked as advertised.

Productivity growth: 1.8% per year. Real compensation growth: 1.1% per year. Gap: 0.7 percentage points.

This era encompasses the stagflation of the 1970s, the Volcker recession of the early 1980s, the Reagan expansion, the savings-and-loan crisis, and the long boom of the 1990s. Productivity growth slowed markedly from the Golden Age — from 2.7% to 1.8% per year — but it continued at a respectable pace. The productivity index rose from 45.6 to 73.4, a 61% increase over 27 years.

Compensation growth, however, did not just slow. It collapsed. From 2.6% per year to 1.1% per year — less than half its Golden Age pace. Over 27 years, that meant the productivity index outpaced the compensation index by a growing margin. The gap of 0.7 percentage points per year may sound modest, but compound interest is relentless. Over 27 years, a 0.7-point annual gap means productivity grew 61% while compensation grew only about 34% — a divergence of 27 percentage points.

What drove this? Several mutually reinforcing forces.

Globalization. The opening of global trade — accelerating through the 1980s and 1990s with NAFTA, the Uruguay Round, and China’s integration into the world economy — put downward pressure on wages in tradeable sectors. An American autoworker competing with a Mexican or Japanese counterpart had less bargaining leverage than one competing only with other Americans.

Declining union power. Union membership in the private sector fell from roughly 34% in 1955 to about 20% by 1983 and continued declining thereafter. The PATCO strike of 1981, when President Reagan fired 11,345 striking air traffic controllers, was widely interpreted as a signal that the political balance had shifted against organized labor. Without unions to enforce wage-productivity linkage, the link dissolved.

The rise of shareholder capitalism. The 1980s saw a philosophical shift in corporate governance. Milton Friedman’s doctrine that a corporation’s only social responsibility is to maximize shareholder value moved from academic provocation to operational mandate. Hostile takeovers, leveraged buyouts, and the growing power of institutional investors all increased pressure on management to prioritize returns to capital over returns to labor.

Technology. The early stages of computerization began automating routine tasks — first in manufacturing, then in office work. Each wave of automation increased output per hour while reducing the number of hours needed. The productivity gains were real, but they accrued disproportionately to the owners of the technology and to the skilled workers who operated it, not to the workers whose tasks were automated away.

Productivity growth: 2.0% per year. Real compensation growth: 0.9% per year. Gap: 1.1 percentage points.

This is, on the surface, a strange era. Productivity growth actually recovered slightly from the post-1973 slowdown — from 1.8% to 2.0% per year — finally benefiting from the information technology revolution that had been building since the 1980s. The productivity index rose from 73.4 to 119.6, a 63% increase in 25 years. The internet, smartphones, cloud computing, logistics software, and eventually artificial intelligence all contributed to genuine gains in output per hour.

But compensation growth decelerated further. From 1.1% per year to 0.9% per year. The gap widened from 0.7 points to 1.1 points. In percentage terms, productivity grew 63% while compensation grew roughly 25% — a divergence of 38 percentage points in a single generation.

The forces driving this widening were the same as in Era II, but more intense. Globalization reached its apex with China’s entry into the WTO in 2001, the resulting “China shock” that hollowed out American manufacturing communities. Union membership in the private sector fell below 7%. The financial sector — which captures productivity gains through fees, interest, and capital appreciation rather than wages — grew to consume roughly 8% of GDP. Healthcare costs continued their relentless rise, consuming an increasing share of total compensation in the form of employer-provided insurance premiums rather than take-home wages.

And technology, which had always been a double-edged sword for labor, became sharper. The automation of routine cognitive tasks — bookkeeping, data entry, basic legal research, customer service — extended the displacement that manufacturing automation had begun. The winners were highly skilled knowledge workers, executives, and capital owners. The losers were everyone whose work could be codified and replicated by software.

The table below summarizes the three eras in a single view. The critical column is the last one: the annual gap between productivity growth and compensation growth. What appears as a small number — 0.1, 0.7, 1.1 — compounds into an enormous divergence over decades.

The cumulative divergence row tells the deeper story. In the Golden Age, the total divergence over 26 years was roughly 3 percentage points — negligible. In Era II, it was approximately 19 points. In Era III, 28 points. Add those together and you get the 50-plus points of divergence that define the productivity-compensation gap today.

To put it another way: if the pre-1973 compact had held — if compensation had continued to track productivity growth at the 2.6-2.7% annual rate — the average American worker today would be earning roughly twice what they actually earn in real terms. That is the magnitude of the structural shift. It is not a rounding error. It is a wholesale redistribution of economic output from labor to capital, from workers to owners, from the many to the few.

It is tempting to treat 1973 as the year the compact broke and leave it at that. But the data show something more troubling: the gap did not just open in 1973 — it has widened in every subsequent era. It went from 0.1 points per year to 0.7 to 1.1. The divergence is accelerating.

This pattern suggests that the break was not primarily caused by the oil shock, the end of Bretton Woods, or any other discrete event. Those events may have triggered the initial divergence, but the widening gap reflects deeper structural forces that have intensified over time.

The declining labor share of income. In national income accounting, GDP is divided between labor compensation and capital income (profits, interest, rent). From 1947 to 1973, labor’s share of nonfarm business income was remarkably stable at roughly 63-65%. After 1973, it began a slow decline. After 2000, the decline accelerated. By 2025, labor’s share had fallen to roughly 56-58% — a transfer of roughly 6-7 percentage points of national income from labor to capital. On an economy producing roughly $28 trillion in GDP, that transfer amounts to nearly $2 trillion per year.

The superstar firm effect. The economy has become more concentrated. In sector after sector — technology, retail, finance, healthcare — a handful of dominant firms capture an outsized share of industry profits. These firms tend to be highly productive, highly profitable, and relatively labor-light. Apple, with roughly $400 billion in revenue and 164,000 employees, generates roughly $2.4 million in revenue per employee. When productivity statistics capture the output of these firms, the aggregate numbers rise. But the gains accrue to shareholders (through stock buybacks and dividends) and to a relatively small number of highly compensated employees, not to the median worker.

The measurement wedge. There is a technical issue that accounts for a portion of the gap. Productivity is calculated using an output price deflator (which measures the prices of goods and services produced), while real compensation is deflated by a consumer price index (which measures the prices of goods and services consumed). Because healthcare, housing, and education costs have risen faster than the general price level, the consumer price deflator has grown faster than the output price deflator. This means some of the measured gap reflects the fact that workers are paying more for what they consume than producers are charging for what they produce. Economists estimate this measurement wedge accounts for roughly a third to a half of the total gap — significant, but still leaving a substantial portion unexplained.

The benefits shift. Total compensation includes both wages and benefits, primarily employer-provided health insurance and retirement contributions. Healthcare costs have risen far faster than inflation for decades. A worker whose total compensation has risen 20% may have seen most or all of that increase consumed by rising insurance premiums, leaving take-home pay essentially flat. The compensation data captures total compensation, which understates the squeeze on wages. The wage data alone would show an even starker divergence.

Numbers on a chart are abstractions. What did these three eras feel like for the workers living through them?

Era I (1947–1973): The escalator. A young man in 1950 could leave high school, walk into a factory or a union construction site, and begin climbing an escalator that would carry him upward for the rest of his career. Each year, his productivity rose and his paycheck rose with it. He could buy a house on a single income, support a family, and expect his children to do even better. The American Dream was not an abstraction; it was a statistical regularity.

Era II (1973–2000): The treadmill. The escalator slowed down. Real wages for non-supervisory workers peaked in 1973 and did not recover their pre-recession levels until the late 1990s — a quarter-century of stagnation. Families compensated by sending a second earner into the workforce. The two-income household became not a choice but a necessity. The economy was still growing, productivity was still rising, but the gains were increasingly captured by the top of the income distribution. The median worker ran faster just to stay in place.

Era III (2000–2025): The split screen. For the top quintile of earners — college-educated professionals in finance, technology, law, and medicine — the period from 2000 to 2025 was an era of extraordinary prosperity. Asset prices soared. Stock options vested. Real estate in desirable cities appreciated beyond anything the previous generation could have imagined. For the bottom three quintiles, the experience was different. Real wages grew slowly if at all. The cost of the necessities — housing, healthcare, childcare, education — rose far faster than income. The social contract that had linked productivity to compensation was, for most workers, a memory.

Economists have attempted to decompose the productivity-compensation gap into its component causes. The exercise is imprecise — the factors overlap and interact — but the broad outlines are widely accepted.

The measurement wedge is the single largest factor, accounting for roughly a third to half of the total gap. This is important because it means that a significant portion of the gap is not, strictly speaking, about employers paying workers less for their productivity. It is about the prices of what workers consume rising faster than the prices of what they produce. In practical terms, however, this distinction matters less than it might seem — workers still experience the gap as stagnant purchasing power regardless of its accounting origins.

After the measurement wedge, the decline in union bargaining power and the effects of globalization are the next-largest contributors. Together, they account for roughly a quarter to a third of the gap. Both operate through the same mechanism: reducing the bargaining leverage of workers, which allows a larger share of productivity gains to be captured by capital. A union autoworker in 1960 could credibly threaten to strike if wages did not keep pace with productivity. A non-union warehouse worker in 2020 has no such leverage, especially when the alternative is a robot.

Technology operates in a subtler way. It does not reduce wages directly (except through automation-driven displacement). Rather, it creates a winner-take-more dynamic where the most productive and highest-skilled workers capture disproportionate gains. When productivity statistics average across all workers, they capture the gains of the superstars. When compensation statistics average across all workers, they are dragged down by the stagnation at the median and below. The average rises; the median does not.

One of the most important features of the 1973 break is that it was essentially invisible at the time. There was no announcement, no policy change, no single day when the compact dissolved. The break was gradual — visible only in retrospect, only when the data had accumulated enough years to make the divergence statistically clear.

In 1973, the immediate crises were the oil embargo, Watergate, and the Vietnam War. These dominated public attention. The notion that something fundamental had changed in the relationship between productivity and compensation would not become a topic of serious economic debate until the 1980s, and would not fully enter the public consciousness until the 2010s.

This is characteristic of structural breaks. They rarely announce themselves. The stock market crash of 1929 was dramatic, visible, unmissable. The productivity-compensation break of 1973 was none of those things. It was a slow leak, not an explosion. The pipe cracked, the water seeped, and by the time anyone noticed the damage, the foundation had shifted irreversibly.

Consider a worker who entered the labor force in 1968 at age 20. By 1973, he was 25 — five years into a career in which productivity and compensation had risen in lockstep. He had every reason to expect this to continue. By 1978, he would have been 30, and his real wages would have been roughly flat for five years. By 1983, he would have been 35, and the gap between what he produced and what he earned would have been widening for a decade. But at each point, the year-to-year change was small enough that it could be attributed to the business cycle, to inflation, to bad luck. The structural nature of the shift was not evident until the 1980s at the earliest.

This matters because it explains why the policy response was so slow and so inadequate. You cannot fix a problem you do not see. And the productivity-compensation gap was, for its first two decades, effectively invisible to the people living through it. By the time it became widely recognized as a structural phenomenon rather than a cyclical one, the institutional changes that had caused it — union decline, globalization, financialization, tax reform — were deeply entrenched and politically defended.

Let us make the compounding concrete. Suppose that in 1973, a worker earned $20 per hour in real terms (roughly the actual figure for production and nonsupervisory workers at the time, in 2017 dollars). And suppose we trace two paths forward from that point: one where compensation tracks productivity (the pre-1973 compact), and one where it tracks the actual post-1973 experience.

The hypothetical is striking. If the pre-1973 compact had held — if workers had continued to receive compensation growth matching productivity growth — the typical production worker in 2025 would be earning roughly $74 per hour in real terms instead of roughly $33 per hour. That is more than double. The gap of $41 per hour, multiplied across 2,000 working hours per year, is $82,000 per year in lost real compensation.

This is, of course, a simplified illustration. The pre-1973 growth rate of 2.7% may not have been sustainable indefinitely. The structural changes that reduced compensation growth also changed the composition of the workforce, the nature of the work itself, and the prices of goods and services. The counterfactual is not a prediction of what would have happened; it is a measure of the magnitude of what did happen.

But the magnitude is real. The roughly $41 per hour difference between the compact path and the actual path is not an abstraction. It shows up in the housing affordability crisis. It shows up in the student debt burden. It shows up in the decline of the single-income household. It shows up in the rising age of first home purchase, first marriage, first child. It shows up in the anxiety and frustration that pervade the economic experience of working Americans who feel, correctly, that the economy is growing but their lives are not getting proportionally better.

If 1973 was the first crack, 1979 was the hammer blow. The second oil shock — triggered by the Iranian Revolution and the subsequent Iran-Iraq War — sent crude oil prices from roughly $15 to $40 per barrel. Inflation spiked again, reaching 13.5% in 1980. The Federal Reserve, under Paul Volcker, responded with the most aggressive monetary tightening in American history, raising the federal funds rate above 20%.

The resulting recession of 1981–82 was devastating. Unemployment peaked at 10.8% in December 1982 — the highest since the Great Depression. Manufacturing employment fell by 2.8 million. The industrial Midwest, already weakened by the first oil shock and growing foreign competition, was shattered. The “Rust Belt” was born.

Volcker succeeded in breaking inflation — it fell from 13.5% in 1980 to 3.2% by 1983. But the cost to workers was enormous. Real wages fell sharply during the recession and recovered only slowly afterward. The industries that had been the backbone of middle-class prosperity — steel, auto, rubber, machinery — entered a secular decline from which they would never fully recover.

The 1979–82 period amplified the 1973 break in two ways. First, it destroyed the wage bargaining power of industrial unions. Workers who were grateful to have any job at all after two years of double-digit unemployment were not in a position to demand wages that tracked productivity growth. Second, it shifted the policy consensus toward prioritizing inflation control over full employment. For the next four decades, the Federal Reserve would lean hawkish, and the working class would bear a disproportionate share of the cost of price stability.

The political response to the crises of the 1970s and early 1980s cemented the structural break into something more permanent. The Reagan administration, which took office in January 1981, pursued a package of policies that, intentionally or not, accelerated the divergence between productivity and compensation.

Tax reform. The Economic Recovery Tax Act of 1981 slashed the top marginal income tax rate from 70% to 50%, and subsequent reforms reduced it further to 28% by 1988. Capital gains taxes were also reduced. The distributional effect was to increase the after-tax returns to capital and high-income labor while reducing the progressivity of the tax code.

Deregulation. The deregulation of airlines, trucking, telecommunications, and finance — which had begun under Carter but accelerated under Reagan — intensified competition in these sectors. The benefits to consumers were real (lower airfares, cheaper phone calls). The costs fell disproportionately on workers in the deregulated industries, whose wages and job security declined.

The PATCO strike. In August 1981, Reagan fired 11,345 striking air traffic controllers and banned them from federal service for life. The action was legal — federal employees were prohibited from striking — but its symbolic impact was enormous. It signaled to private-sector employers that aggressive anti-union action would not be punished by the government. Strike activity plummeted. In 1970, there had been 381 major work stoppages in the United States. By 1990, there were 44. By 2000, there were 39.

By the mid-1980s, a new equilibrium had taken shape. Productivity would continue to rise. Corporate profits would continue to grow. The stock market would enter one of the great bull markets in history. But the gains from all of this growth would be distributed very differently from the postwar norm. The link between what workers produced and what they earned — the compact that had defined the American economy for a generation — was broken, and the institutional changes of the early 1980s ensured that it would stay broken.

The productivity and compensation data from the BLS are among the most carefully constructed economic statistics in the world. But they have important limitations that must be acknowledged.

Composition effects. The workforce of 2025 is radically different from the workforce of 1973. It is more educated, more female, more diverse, more service-oriented. Comparing aggregate productivity and compensation across these very different workforces is inherently imprecise. Some of the measured divergence reflects changes in who is working and what kind of work they are doing, not just changes in how the gains from that work are distributed.

Quality adjustments. The goods and services produced today are vastly different from those produced in 1973. A 2025 automobile is incomparably safer, more efficient, and more capable than a 1973 automobile. A 2025 medical procedure may cure what a 1973 procedure could only palliate. The productivity statistics attempt to account for quality improvements through hedonic adjustments, but these are inherently judgmental. It is possible that actual productivity growth has been either understated or overstated by these adjustments.

The non-market economy. A growing share of what people consume is produced outside the market economy — open-source software, user-generated content, unpaid domestic labor, volunteer work. None of this shows up in productivity statistics. To the extent that the non-market economy has grown faster than the market economy, measured productivity growth understates the actual improvement in human welfare. But this non-market output also generates no compensation, which means it contributes to the measured gap even though it represents genuine value creation.

Distribution within labor. The productivity-compensation data describe averages. They do not capture the distribution within the labor force. The experience of a software engineer at Google (whose compensation has far outpaced productivity growth) is fundamentally different from the experience of a retail worker at Walmart (whose compensation has stagnated). Aggregate statistics smooth over these differences, creating a misleading appearance of uniformity.

The 1973 break is now more than fifty years old. The structural forces that caused it — globalization, declining union power, skill-biased technology, financialization — are mature. Some are beginning to reverse: nearshoring is partially replacing offshoring, tight labor markets in the early 2020s gave workers more bargaining power, union organizing is experiencing a modest revival, and political attitudes toward labor have shifted in both parties.

But new forces are emerging. Artificial intelligence represents the most significant productivity-enhancing technology since the computer itself, and possibly since electrification. If AI delivers even a fraction of the productivity gains that its proponents predict, the question of how those gains are distributed will become the central economic issue of the next generation. History suggests that without deliberate institutional choices to share productivity gains broadly, the default outcome is further concentration at the top.

The next episode in this series will examine the measurement question in greater depth — what exactly do we mean by “productivity,” how is it measured, and what are the alternative measures that tell a somewhat different story. Understanding the measurement is essential before we can address the question of what, if anything, can be done to restore the compact.