OECD rate of profit new source

The same rate-of-profit calculation as the world tool, but from a different data source entirely: OECD's STAN database instead of the Penn World Table. STAN is actively updated (its companion ICIO/TiVA tables now run to 2022, with a new edition roughly every two years) and its industry-level detail is a path to questions PWT's whole-economy aggregates can't touch, like productive-sector profit rates -- not built here yet, but the door PWT doesn't have.

r = s / (C + v)  ·  s = gross operating surplus + mixed income (B2A3G)  ·  v = compensation of employees (D1)  ·  C = net capital stock, current replacement cost (N11NA)

Why a second source matters: every number on the other pages ultimately traces back to one data family (PWT) or one country's statistical agency (BEA). An academic comparison of WIOD against this same OECD system found their value-added estimates disagree by a mean of ~45% at the country level, even over the same years — different institutions building input-output data from similar raw material still diverge substantially. Whatever this page's rate of profit does or doesn't have in common with the PWT-based one is exactly that kind of independent check, not a rounding difference.

No claim of Roberts's in this project's indexed corpus (1,600+ posts checked) uses OECD/STAN or TiVA data — everything he cites is PWT, AMECO, EPWT, BEA, or WIOD. This page extends what this project can compute; it isn't verifying something he's already published.

Static snapshot — this page shows one pre-computed result (G20, aggregate, C+v, 1995–2023). The controls below are for reference; changing them won't recompute anything on this deployment, which has no live backend. The full interactive version, with any country set or date range, runs locally from the marxist-ai project.

Country set

OECD/STAN coverage is narrower than PWT's — mostly OECD members plus major non-OECD economies. Some G20/BRICS members (e.g. Saudi Arabia) may not report.

Method
Years

STAN's flow variables reach back to 1970 for some countries, but coverage is thinnest before ~1995.

Set the controls and choose Run calculation. The first run for any country set is slower — OECD's API is queried live and cached to disk after that.