Tariffs are Reward Hacking Squared: A Potemkin Economy where Statistical Prosperity Masks Industrial Decline
Introduction: The Measurement Problem That Became a Monster
In the aftermath of recognizing that American manufacturing statistics have divorced from reality, policymakers have increasingly turned to tariffs as a solution. The logic appears straightforward: if foreign competition has hollowed out domestic industry while GDP statistics claim otherwise, then protecting domestic producers should restore genuine productive capacity. Yet this seemingly logical response contains a fatal flaw. Rather than fixing the disconnect between economic statistics and economic reality, tariffs amplify it exponentially. They create what can only be described as “reward hacking squared” — a meta-layer of distortion that makes the original measurement problems not just worse, but unfixable.
To understand this perverse outcome, we must first grasp how modern economies have become optimization machines targeted at statistical measures rather than real production, and then examine how tariffs transform from an intended cure into an accelerant of the disease.
The Original Sin: When GDP Became the Target
The story begins with GDP itself, that supposedly objective measure of economic output. Originally designed in the 1930s to track national economic health, GDP has evolved into something far removed from its purpose. Through decades of methodological “improvements” — quality adjustments, hedonic pricing, retroactive base-year modifications — GDP has become a statistical construct that can show “growth” even when actual production declines.
Consider the steel industry. Between 1997 and 2023, raw tonnage of American steel production fell 18%, yet GDP statistics claim “real value-added” in steel manufacturing rose 125%. This isn’t measurement; it’s alchemy. The statistical agencies achieve these miraculous transformations through arcane methodologies that no outside observer can verify or replicate. When semiconductor statistics show 1,698% growth in value-added while actual chip production merely doubled, we’re no longer measuring reality — we’re measuring the echoes of bureaucratic assumptions.
This is classic reward hacking: when a metric becomes a target, it ceases to be a good metric. Organizations and entire economies reorganize themselves not to achieve the original purpose (productive capacity) but to optimize the measurement (GDP). The result is a Potemkin economy where statistical prosperity masks industrial decline.
Enter Tariffs: The Cure That Multiplies the Disease
Into this already distorted landscape, tariffs arrive as a supposed correction. The political appeal is obvious: if American manufacturing has been undermined by unfair foreign competition, then taxing imports should level the playing field and restore domestic production. The rhetoric is compelling — protect American jobs, rebuild American factories, restore American greatness.
But tariffs don’t operate in a vacuum. They operate within the same broken measurement system that created the problem. And within that system, they become a powerful amplifier of distortion rather than a corrective force.
The First Layer: How Protection Inflates Phantom Value
When a 100% tariff doubles the price of imported washing machines from $400 to $800, domestic manufacturers can suddenly charge $750 for products that previously sold for $500. The GDP statistics record this price increase as “value creation.” After all, American companies are now generating 50% more revenue per unit.
But here’s where the distortion begins. The Bureau of Economic Analysis, seeing domestic products maintaining market share despite higher prices, interprets this as evidence of quality improvement. Their models assume consumers are rational actors who wouldn’t pay more unless they were getting more value. Through the lens of hedonic adjustments and quality indexing, that $250 price increase becomes “real” productivity growth.
The washing machine hasn’t improved. No innovation has occurred. No efficiency has been gained. Yet GDP statistics will show a manufacturing boom. This is the first multiplication of the lie — tariffs creating statistical growth from mere price inflation.
The Second Layer: The Potemkin Factory Problem
The distortion deepens when companies realize they can game both the tariff system and the GDP measurements simultaneously. Consider a hypothetical electronics “manufacturer” that imports 95% complete products from Vietnam, adds a plastic case and power cord in Ohio, and sells the result as “Made in USA.”
Without tariffs, this operation might add $100 of value to a $900 import, generating a modest $100 of GDP “value-added.” But with 300% tariffs on finished electronics and 0% on components, the game changes entirely. Now the company can charge $2,000 for the same product. GDP statistics record this as $1,100 of value-added — an eleven-fold increase in “manufacturing output” for the exact same screwdriver assembly operation.
Multiply this across thousands of products and industries, and you have entire sectors showing explosive GDP growth while actual manufacturing capability — the ability to truly make things from raw materials — atrophies further.
The Third Layer: Rent-Seeking Replaces Innovation
Protected markets inevitably breed rent-seeking behavior, but in combination with GDP gaming, they create something worse: the optimization of extraction rather than production. Companies discover that lobbying for higher tariffs generates better returns than investing in research and development. Why improve your product when you can simply improve your protection?
The Jones Act provides the perfect case study. American shipbuilders, protected by laws requiring domestic water transport to use American-built ships, now charge ten times the global price for vessels. GDP statistics record this as exceptional productivity — after all, they’re generating ten times the value per ship! But the reality is an industry so uncompetitive it can barely function, surviving solely through regulatory capture.
This pattern replicates across protected industries. Each sees GDP “growth” proportional to its dysfunction. The more a sector extracts from consumers through protection rather than serving them through innovation, the better it looks in the statistics. The measurement system doesn’t just tolerate parasitism — it rewards it.
The Fourth Layer: The Supply Chain Shell Game
Sophisticated companies don’t just game tariffs; they architect entire supply chain structures to maximize the statistical fiction. A “manufacturer” imports “industrial equipment parts” (5% tariff) that are actually 95% complete cars. A subsidiary imports “rubber products” (2% tariff) that are actually complete wheels, adds valve stems, and sells them to the parent at a 1,000% markup.
The GDP statistics faithfully record each step as value creation. The parent company shows massive manufacturing value-added. The subsidiary shows impressive component production. Together, they contribute billions to manufacturing GDP while performing perhaps 1% of actual manufacturing work.
These aren’t aberrations or fraud — they’re rational responses to a system that rewards statistical gaming over genuine production. The tariff system doesn’t just enable this behavior; it institutionalizes it.
The Fifth Layer: Quality Adjustments Go Quantum
The interaction between tariffs and quality adjustments creates truly absurd mathematical outcomes. When tariff protection doubles prices, and statisticians interpret this as quality improvement, the retroactive adjustments to “real” GDP can show infinite or even undefined growth rates.
Here’s the mathematical perversion in action: A product selling for $100 with $75 in costs shows $25 of value-added. Tariff protection allows the price to double to $200. Statistical agencies, seeing the price increase, retroactively adjust the base year “real” output down to $50, making the original value-added negative $25. The current year’s $125 value-added compared to negative $25 shows infinite growth — from actually making something to making the same thing behind a tariff wall.
This isn’t a bug in the system; it’s the logical outcome of combining protection with quality adjustments. The more dysfunctional and expensive protected industries become, the more “productive” they appear in the statistics.
The Sixth Layer: The Bilateral Gaming Death Spiral
When multiple countries engage in tariff wars, each can show GDP “growth” while actual global production declines. America taxes steel, showing steel GDP growth. China retaliates on soybeans, showing agricultural GDP growth. Both subsidize affected industries, showing government spending growth.
Every retaliatory measure creates new statistical gains for someone while actual economic efficiency — the ability to produce goods at lowest cost for maximum human benefit — collapses. It’s a positive-sum game in the fantasy world of statistics and a negative-sum game in reality.
The Seventh Layer: Financial Engineering as Manufacturing
Protected industries discover they can generate more “manufacturing GDP” through financial engineering than actual production. Create an Irish subsidiary to hold intellectual property. License it to a Mexican assembly plant at inflated prices. Import the nearly-finished goods to America for final touches. Sell at protected prices. Offer financing at usurious rates.
GDP records this Rube Goldberg machine as a triumph of American manufacturing. The value-added is enormous. The financial services component adds even more. Yet no actual manufacturing capability exists — just an elaborate structure for extracting rents while generating statistics.
The Meta-Hack: Making Failure Unmeasurable
The most insidious aspect of tariff-amplified distortion is that it makes its own failure impossible to detect within the measurement framework. Tariffs increase GDP. GDP measures success. Therefore tariffs are successful. Therefore we need more tariffs.
Any attempt to point out the emperor has no clothes gets dismissed. Why count actual units produced when “quality” matters? Why compare to global prices when American products are “premium”? Why worry about consumer costs when GDP is rising? The circular logic becomes impenetrable.
Statistical agencies, under political pressure to show tariffs “working,” adjust their methodologies. Academics who question the numbers lose funding while those who validate them get published. An entire intellectual infrastructure builds around defending the fiction. The measurement system itself becomes corrupted, not through conspiracy but through the inexorable logic of institutional incentives.
The Potemkin Economy End State
The end result is an economy that optimizes for statistical fiction while its real productive capacity collapses. GDP shows manufacturing at all-time highs. Productivity metrics soar. Value-added breaks records. Politicians declare victory.
Meanwhile, consumers pay multiples of world prices for basic goods. Innovation grinds to a halt as protection removes competitive pressure. Real manufacturing knowledge — the ability to actually make things — atrophies and disappears. Screwdriver plants assembling imported components are classified as factories. Financial engineering replaces industrial engineering.
The economy becomes a massive machine for converting consumer wealth into producer rents while generating beautiful statistics showing prosperity. It’s not just that the measurements are wrong — it’s that they’re wrong in ways that make discovering the truth impossible within the system.
The Compound Lie: Why Tariffs Are Reward Hacking Squared
Tariffs don’t fix the reward hacking problem in GDP — they square it. The original hack was bad enough: a measurement system that confused statistical manipulation with real production. But tariffs add a second layer of hacking on top of the first, creating new opportunities for gaming while making detection impossible.
Where GDP reward hacking created fictional growth, tariff-enhanced reward hacking creates fictional growth that costs real money. Where the original distortion could be revealed by comparing statistics to reality, the tariff distortion corrupts reality itself, creating protected markets where dysfunction becomes profitable and profitable dysfunction becomes statistically invisible.
This is reward hacking squared: not just gaming the metric, but gaming the game of gaming the metric. It’s a meta-hack that makes the original problem not just worse but structurally unfixable. The patient was sick from eating junk food, so we prescribed poison that makes junk food look nutritious in the blood tests.
Conclusion: The Impossible Problem
The tragedy is that tariffs emerge from a genuine recognition that something is wrong. People see factories closing while economists celebrate record manufacturing output. Workers lose jobs while statistics show productivity soaring. The lived experience diverges so completely from the official narrative that protective measures seem not just reasonable but necessary.
Yet by operating within the same broken measurement paradigm they’re meant to address, tariffs become accelerants rather than solutions. They don’t restore American manufacturing; they complete its transformation into a statistical fiction. They don’t fix the disconnect between metrics and reality; they institutionalize it.
The real solution requires something far more difficult than imposing taxes on imports. It requires acknowledging that our core economic measurements have become so corrupted by decades of gaming that they no longer measure anything meaningful. It requires developing new metrics that can’t be gamed, that track actual productive capability rather than statistical artifacts. It requires the humility to admit that we’ve been optimizing for lies.
Until we confront this deeper problem, every solution — tariffs included — will only compound the distortion. We’ll continue building elaborate statistical castles while our real productive foundation crumbles. And the metrics we rely on to guide us will lead us deeper into the darkness, their brightness nothing but the glow of burning value.
The choice is stark: continue reward hacking squared until reality itself becomes unmeasurable, or abandon the corrupted metrics and rebuild from first principles. The tariff debate isn’t really about trade policy. It’s about whether we’ll face the truth about what we’ve become — a civilization optimizing for statistical fiction — or double down on the lie until the fiction becomes all we have left.
Further Reading:
Rebuttal to the Palladium essay: https://x.com/sichuan_mala/status/1974935552303776074
