Is the Industrial Revolution a good precedent for explosive economic growth today?
A statistical analysis showing that the Industrial Revolution's "10x growth acceleration" meant good years becoming more common—something pre-IR Britain experienced 46% of the time—while today's proposed AI-driven 20%+ growth would be unprecedented, occurring in less than 1% of modern country-years.
Read Original Summary used for search
TLDR
• Pre-Industrial Revolution UK regularly hit 10x its average growth rate (~46% of years), making the IR acceleration feel like "more good harvests" rather than a new economic regime
• Modern economies almost never experience 10x their baseline growth—only 0.76% of country-years since 1950 exceeded this threshold, mostly from oil windfalls or post-conflict recovery
• Using standard deviation analysis, an IR-style acceleration today would mean ~2.8% annual growth for frontier economies, not the 20%+ often claimed for AI scenarios
• The key insight: 10x acceleration in 1700 meant experiencing familiar growth rates more often; 10x today would require entering a completely unprecedented economic state
• Countries with slower baseline growth (0-1%/yr) commonly experience 10x years (94% had at least one), but this drops to near-zero for countries growing >2%/yr
In Detail
The piece dismantles a common argument for AI-driven explosive growth by showing that the Industrial Revolution is a misleading precedent. During 1252-1652, the UK's compound growth rate was 0.07%/year, but annual growth exceeding 10x this rate (0.66%/year) occurred in roughly 46% of years. The IR acceleration to 1.02%/year didn't require unprecedented short-run growth rates—it just meant good years became more common and bad years less frequent. For someone living in 1700, a "10x acceleration" would have felt like experiencing growth rates they'd already seen many times.
The modern situation is fundamentally different. For the US, 10x the compound growth rate implies 19%/year—a level never experienced even once. Across all countries from 1950-2022, only 0.76% of country-years exceeded this threshold, and most were oil windfalls (Kuwait post-Gulf War, Libya's oil boom) or post-conflict recovery (Rwanda 1995, Iraq 2004), not genuine productivity accelerations. The author validates this using modern analogue countries (Benin, Burundi, Chad, etc.) with similar GDP/population characteristics to pre-IR UK, finding they also commonly experience 10x growth years.
Using standard deviation analysis provides a more calibrated comparison: the IR represented a 0.14-0.17 standard deviation increase in growth rates. Applying the same magnitude increase to today's US baseline of 1.9%/year yields approximately 2.8%/year—meaningfully faster than current growth but nowhere near the 20%+ often claimed for AI scenarios. A true 10x acceleration would be 3.1 to 7.5 standard deviations depending on the reference distribution, placing it far outside historical experience. The analysis suggests that skepticism toward explosive AI-driven growth claims is well-calibrated to historical precedent, unlike the skepticism that would have been misplaced in 1700.