Productivity: the boring number behind every real raise

Of all the economic statistics, productivity — output produced per hour worked — gets the least airtime and arguably matters the most for your standard of living. Over generations, it is essentially the ONLY thing that makes a society richer: more value created per hour is the fund from which all real wage growth, all shorter workweeks, and all cheaper goods are ultimately paid. When productivity grows 2%+ a year, raises can beat inflation without squeezing anyone. When it grows 0.5%, the economy becomes a zero-sum knife fight over a nearly fixed pie.

The mechanism, without the math

A business can only sustainably pay workers more if each hour of work produces more value — through better tools, better processes, or better skills. If wages rise faster than productivity for long, businesses raise prices to cover it, and inflation eats the raise: that's the wage-price spiral central bankers fear. If productivity rises and wages DON'T follow, profits swell instead — which is roughly the story of several recent decades, and why 'productivity–pay gap' is a phrase worth knowing. Either way, productivity growth sets the ceiling on how fast real pay CAN grow; politics and bargaining power decide how much of it workers actually capture.

How the number is measured (and why it's noisy)

Productivity is calculated by dividing total output by total hours worked — which makes it a ratio of two estimates, and therefore one of the noisiest major statistics. Quarterly readings swing wildly for mechanical reasons: in a recession's first months, output falls faster than firms cut staff, so measured productivity dives; in early recoveries, output rebounds before hiring does, so productivity 'surges.' Neither swing says anything about technology or skill. The signal lives in multi-year averages, which is why economists talk in eras rather than quarters. Measurement also struggles with services and free goods — a search engine that saves you an hour shows up nowhere in output, and quality improvements in healthcare or education are notoriously hard to count. Some economists suspect true productivity growth is chronically understated for exactly these reasons.

Why some eras feel rich and others feel stuck

• 1948–1973: productivity grew nearly 3% a year and median pay tracked it closely — the era when one income bought a house, and the source of most nostalgia about the economy.
• 1973–1995: productivity slowed to roughly 1.5%; real median wages stagnated. The 'stuck' feeling of that era was arithmetic, not attitude.
• 1995–2005: the computer and internet buildout pushed productivity back near 3%; real wages grew across most of the distribution.
• 2005–2020: back below 1.5% for long stretches — slow raises, cheap-money asset booms, and widespread 'working harder for the same life' sentiment.

Why productivity speeds up and slows down remains one of economics' genuinely open questions. The postwar boom had obvious fuel — electrification maturing, interstate highways, mass education, factories rebuilt around new methods. The 1970s slowdown is still debated fifty years later: oil shocks, measurement problems, and the exhaustion of earlier technologies all get blamed. The IT boom showed up decades after computers arrived, only once businesses reorganized around them — the lag wasn't about inventing the tool but rebuilding the workflows. That history is the strongest argument for humility about the current moment: general-purpose technologies pay off on reorganization timelines, not announcement timelines, and the statistics will settle the AI debate years after the arguments started.

The AI question, honestly

Every productivity wave came from a general-purpose technology — steam, electricity, computing — and each took decades longer than expected to show up in the statistics (economist Robert Solow quipped in 1987 that computers were 'everywhere except the productivity statistics'). AI is the current candidate, and early studies show real task-level gains. Whether that becomes economy-wide productivity — and whether workers capture it as wages — is genuinely unknown. The historical pattern: the gains are real, arrive late, and flow first to whoever wields the new tools well.

What to do with all this

1. Judge economic eras by productivity trends, not vibes — it explains 'stuck' decades better than any villain narrative.
2. Invest in your own output-per-hour every year: tools, skills, and leverage compound like money does.
3. Benchmark raises against inflation AND your rising output; if your value grew 10% and your pay grew 2%, that gap is your negotiation.
4. Own broad stock index funds — it's the simplest way to be paid when productivity gains skip paychecks and land in profits.
5. Treat confident predictions about AI productivity (utopian or doom) as entertainment; position for either by staying skilled and diversified.

The bottom line

Productivity is the quiet arithmetic underneath every raise, every 'golden age,' and every stuck decade: output per hour sets what an economy can afford to pay its people. Nationally, you're along for the ride. Personally, you're not — grow the value of your own hour relentlessly, negotiate to capture it, and hold stocks so the gains you can't capture as wages reach you as an owner instead.