Don't get hung up on "If total factor productivity". https://economics.stackexchange.com/questions/33080/dimensional-analysis-of-gdp https://economics.stackexc

Don't get hung up on "If total factor productivity". https://economics.stackexchange.com/questions/33080/dimensional-analysis-of-gdp https://economics.stackexchange.com/questions/33080/dimensio... sort of clears some stuff up. Economists are not that good at dimensional analysis, despite their worship of physicists, and every physicist knowing dimensional analysis is key to staying sane. GDP, for the purposes of this graph, is stuff measured in dollars, per time, so "value"/time. The estimates are not that different from linear regressions, asymptotically. That's it. "Productivity" is a nebulous concept. Ultimately, we only "see" the value in the market price of things being exchanged, so productivity is thus reasoning backwards with division/differentiation. There is no empirically rigorous notion of per-person productivity to integrate over, and whether the ability to divide/differential makes that platonically-valid concept nonetheless a murky philosophical debate one need not get into to understand this graph.