The productivity-pay gap is the difference between how much more workers produce per hour and how much more they get paid for it. For roughly three decades after World War II, the two moved together almost one for one: a more productive workforce was, on average, a better-paid one. That relationship broke around 1979. Since then, US net productivity has grown roughly eight times faster than the typical worker’s hourly compensation, a gap the Economic Policy Institute has tracked continuously and that shows no sign of closing on its own.
The gap matters because it undercuts the assumption most career and compensation advice is still built on, that being more productive is, by itself, the mechanism that builds wealth. It isn’t, and hasn’t reliably been for nearly half a century. The causes are structural rather than individual: declining union bargaining power, a minimum wage that stops rising, and tax and corporate-governance choices, like prioritising share buybacks over wage growth, that route the returns from higher output toward capital rather than labor.
The gap predates AI by four and a half decades, which is the useful context for reading any AI-era claim that “productivity gains will lift wages.” History says that link has to be built deliberately, through bargaining power, ownership structures, or policy, because it does not happen as a byproduct of the technology alone.