The Winners and Losers: Who Pays the Price of Divergence?

Across eleven episodes and seventy-nine years of Bureau of Labor Statistics data, we have traced the price of nearly everything an American consumer buys — from a loaf of bread to a year of college, from a gallon of gasoline to a night in a hospital, from a television to a pack of cigarettes. The pattern that emerged was not random. It was structural, persistent, and accelerating.

Things produced by machines got cheaper. Things produced by people got more expensive. Televisions fell 99%. College tuition rose 1,338%. That single sentence captures the central finding of this series. But what it does not capture is the distributional consequence: who benefits from the deflation, and who bears the burden of the inflation.

This final episode connects the threads. It does not introduce new data — every number here was documented in Episodes 1 through 11. Instead, it asks the question the data cannot answer on its own: who wins, who loses, and what — if anything — can be done about it?

Before we ask who wins and who loses, let us see the final scoreboard. Across twelve episodes, we tracked sixteen major categories of American consumer spending. Their price changes since the late 1970s and early 1980s range from a 1,338% increase to a 99% decline. The chart below ranks every category from most inflated to most deflated.

The amber line marks the All Items CPI — the official yardstick of overall inflation at +319%. Everything above it inflated faster than the economy as a whole. Everything below it inflated slower, or outright deflated. The gap between the top and the bottom of this chart is 1,437 percentage points. That is not a gap between different countries or different centuries. It is the gap within a single American household’s budget.

The American economy’s great price divergence has clear beneficiaries. They are not always the people you would expect — the advantages are structural, not intentional, and they compound over decades in ways that widen inequality without anyone making a conscious decision to do so.

Consumers of manufactured goods. Anyone buying televisions, computers, clothing, toys, or electronics has been the beneficiary of relentless deflation. As we documented in Episode 6, televisions have fallen 99% in CPI-adjusted terms since 1996. Computers have fallen 78%. Toys have fallen 66%. Apparel has risen just 51% over four decades — meaning it has fallen substantially in real terms. A five-hundred-dollar budget today buys incomparably more than it did in 1980: a flat-screen television, a smartphone more powerful than a 1990s supercomputer, and a wardrobe that would have cost thousands. The technological revolution has delivered a bonanza to anyone whose consumption tilts toward goods.

Homeowners. Episode 3 documented that shelter costs have risen 456% since 1980. But that figure falls almost entirely on renters. Homeowners with fixed-rate mortgages lock in their largest monthly expense at the time of purchase. As inflation erodes the real value of that payment, their housing costs effectively decline over time. A mortgage payment of $800 per month in 1990 is still $800 per month in 2026 — while the same home might now rent for $2,500. Homeownership is, in effect, an inflation hedge that transfers the burden of shelter inflation entirely to those who cannot access it.

Workers in high-productivity industries. Technology, manufacturing, and logistics workers benefit from the same forces that drive goods prices down. Their industries produce more output per worker each year, which supports higher wages. A software engineer’s productivity in 2026 is orders of magnitude higher than in 1990, and compensation has followed. Workers in low-productivity service sectors — healthcare aides, teachers, retail clerks — have seen nothing comparable.

Those with employer-provided health insurance. The 726% increase in medical care costs documented in Episode 4 is real — but for the roughly 155 million Americans with employer-sponsored insurance, much of it is invisible. The employer absorbs the premium increases, and the employee sees only copays and deductibles. This is not free — economists have shown that employer health costs suppress wage growth — but the pain is diffuse and indirect, not the raw shock of a hospital bill at full price.

The wealthy. Higher-income households spend a larger share of their budgets on goods — electronics, automobiles, luxury items, travel — which have inflated less than services. Meanwhile, the services that have inflated most dramatically — tuition, healthcare, housing — consume a smaller share of wealthy households’ budgets. A family earning $500,000 per year spending $30,000 on tuition feels the pain differently than a family earning $60,000. The price divergence is, in effect, a regressive tax: it takes a larger share from those with less.

The losing side of the divergence is populated by the Americans who can least afford it. The categories that inflated fastest — shelter, healthcare, education — are not discretionary. You cannot choose not to have a roof, not to see a doctor, not to educate your children. The deflation in goods is real but irrelevant when the rent is due.

Renters. Full exposure to the 456% increase in shelter costs. No equity buildup, no inflation hedge, no mortgage interest deduction. As we showed in Episode 3, renters spend a larger share of their income on housing than at any point in the BLS survey’s history. For the roughly 44 million renter households in the United States, the shelter component of the CPI is not an abstraction — it is the check they write every month. And unlike a mortgage, rent resets upward every year.

Students and their families. College tuition has risen 1,338% since 1978, as Episode 5 documented. The result: student loan debt now exceeds $1.7 trillion, second only to mortgage debt among American household obligations. An entire generation has financed its education with debt that cannot be discharged in bankruptcy, that accrues interest during graduate school, and that follows borrowers for decades. The Americans who invested most heavily in their own human capital are, paradoxically, among the most burdened by the price divergence.

The uninsured and underinsured. Medical care costs have risen 726% since 1980. For the roughly 27 million Americans without health insurance — and the tens of millions more with high-deductible plans that leave them exposed to catastrophic costs — the hospital bill documented in Episode 4 lands at its full, unmitigated weight. Medical debt is the leading cause of personal bankruptcy in the United States, a distinction that exists in no other developed nation.

Retirees on fixed income. Episode 11 showed that a 1947 dollar is worth 6.6 cents today. Social Security benefits are adjusted annually for CPI inflation — but the CPI measures the average consumer’s basket, not the retiree’s basket. Retirees spend disproportionately on healthcare and housing, the two categories that have outpaced headline CPI most dramatically. The result is a slow, invisible erosion of purchasing power that accelerates as retirees age and healthcare needs intensify.

Smokers. Tobacco prices have risen 1,326% since 1986 — the second-highest increase in our entire scoreboard. As Episode 7 documented, this is largely deliberate: federal and state taxes have been raised repeatedly as a public health measure. But smoking rates are highest among low-income populations. A pack-a-day habit at $10 per pack costs $3,650 per year — a meaningful fraction of a minimum-wage income. What is designed as a public health tool functions, in practice, as a regressive tax on addiction.

The young. Perhaps the most consequential losers of the price divergence are young Americans entering adulthood in the 2020s. They face a triple burden that no previous generation confronted simultaneously: student debt from tuition that rose 1,338%, housing costs with shelter up 456%, and healthcare costs up 726% — all landing at career stages when income is at its lowest. The deflation in goods is cold comfort when you cannot afford rent, cannot afford insurance, and are already carrying five or six figures of educational debt. A cheaper television does not offset a more expensive life.

Episode 10 introduced the concept that the BLS data, taken as a whole, reveals most clearly: America does not have one economy. It has two. A goods economy, where technology and globalization drive relentless deflation. And a services economy, where human labor, regulation, and institutional complexity drive relentless inflation. The two have been diverging since at least 1980 — and the gap is now 305 index points wide.

The chart below shows this divergence in its simplest form. Both lines are indexed to 1980 = 100. The Services CPI has risen to approximately 582 by early 2026. The Commodities (goods) CPI has risen to approximately 277. The amber-shaded area between them represents the gap — the space where winners and losers are made.

Government policy has, by some measures, succeeded in making certain things more expensive. Tobacco taxation is a public health triumph — smoking rates have plummeted from 33% of adults in 1980 to under 12% today, and the 1,326% price increase documented in Episode 7 was instrumental. The policy worked. People smoke less because cigarettes cost more.

But in the categories where Americans most need prices to fall — healthcare, education, housing — government has been unable or unwilling to achieve anything comparable. Indeed, the items that have risen fastest are all in sectors where government involvement is deepest. Healthcare is the most regulated industry in the economy. Higher education runs on federal student loans and grants. Housing is shaped by zoning laws, building codes, tax policy, and monetary policy. This is not to say that government caused the inflation — the structural forces documented in Episode 10 would operate with or without regulation. But government is the only actor with the scale and authority to potentially contain it, and the record across seven decades of data is, at best, mixed.

The irony is sharp: we have demonstrated the political will to make tobacco unaffordable through taxation, but we have not demonstrated the institutional capacity to make healthcare, education, or housing affordable through any mechanism — subsidy, regulation, or reform. The market alone will not solve the problem. Technology does not disrupt hospitals the way it disrupts electronics factories. Baumol’s Cost Disease, the subject of Episode 10, is not a disease with a market cure.

At the heart of every episode in this series sits a single economic mechanism, first described by William Baumol and William Bowen in 1966: the cost disease of the service sector. A factory that made 100 televisions per worker in 1990 makes 10,000 per worker today. The cost per unit plummets. A surgeon who performed one knee replacement per day in 1990 still performs one knee replacement per day in 2026. The cost per procedure rises with the surgeon’s wages — which must keep pace with wages in the productive sectors, even though the surgeon’s output has not increased.

This is the common thread that connects every episode. Groceries have risen because food production, while more mechanized than in 1980, still requires human labor at every stage from farm to shelf (Episode 1). Gasoline has been volatile because it sits at the intersection of global commodity markets and domestic refining labor (Episode 2). Rent has soared because building and maintaining housing requires carpenters, plumbers, and electricians whose productivity has barely changed (Episode 3). Hospital bills have exploded because healthcare is the most labor-intensive sector in the economy (Episode 4). Tuition has risen because a classroom still requires a professor in front of students (Episode 5).

And the things that got cheaper? They got cheaper because machines replaced people. Television manufacturing is almost entirely automated. Computer chip fabrication is performed by robots in clean rooms. Clothing is sewn in factories where the ratio of machines to workers has risen every decade. Wherever technology could substitute for human labor, prices fell. Wherever it could not, prices rose.

This is not going to reverse. Artificial intelligence may eventually transform some service sectors — medical diagnosis, legal research, educational delivery. But the physical, hands-on services that dominate the CPI — surgery, construction, caregiving, teaching — remain resistant to automation. The divergence documented across these twelve episodes is structural, not cyclical. It will continue.

The table below summarizes every episode of The Price of Everything. Twelve episodes, sixteen categories, seventy-nine years of data, and one unmistakable pattern.