---
title: "Nordhaus argues the learning-curve coefficient is statistically unidentified and biased upward — a zero-learning process still fits a coefficient of 0.2"
type: "claim"
status: "seedling"
writer_model: "claude-opus-4-8"
source_url: "https://www.nber.org/system/files/working_papers/w14638/w14638.pdf"
source_author: "William D. Nordhaus"
source_date: 2009
source_quote: "the empirical learning coefficient is 0.2 even though the actual learning coefficient is zero"
source_tier: 1
audit_status: "capture-verified (Nordhaus NBER w14638 quoted by the hop-batch worker at capture time; the promoting engine's independent re-fetch not run); 2026-07-19 cross-model audit (claude-fable-5): independent re-fetch of the full w14638 PDF run — all quoted figures verified verbatim ('biased upwards'; zero-learning case fitting 0.2; 34 industries with only 4 in the 0–0.5 range; specification correlation 0.009; costs 'underestimated by a factor of two' with solar and wind named); corrected source_date 2008 → 2009 (the working paper is dated January 2009) and expanded the truncated paper title in the body"
provenance: "Promotion from 10-inbox/raw/2026-07-11-hop-learning-curve-unidentified.md, 2026-07-18"
origin: "batch"
derived_from: "10-inbox/raw/2026-07-11-hop-learning-curve-unidentified.md"
date_created: "2026-07-18T00:00:00.000Z"
tags: ["economics","learning-curve","experience-curve","wrights-law","econometrics","identification-problem","technology-forecasting"]
drafted_in: ["the-line-no-one-walks"]
audits: ["2026-07-19 claude-fable-5"]
---


In "The Perils of the Learning Model for Modeling Endogenous Technological Change" (NBER w14638, January 2009), William D. Nordhaus argues that the coefficient at the heart of [[claim-wrights-law-cost-falls-per-cumulative-production-doubling|Wright's law / the experience curve]] cannot, in the standard specification, be separated from ordinary exogenous technological progress. Because cumulative production and time both trend upward together, a regression attributing cost decline to learning-by-doing absorbs unrelated progress into the learning term: "the estimated learning coefficient will generally be biased upwards." His numerical demonstration is the sharp edge — in a constructed case with *zero* true learning, the fitted parameter still comes out positive: "the empirical learning coefficient is 0.2 even though the actual learning coefficient is zero."

The empirical instability is of the same order. Across 34 industries, "only 4 have estimated empirical learning coefficients in the plausible range between 0 and 0.5," and the correlation between two reasonable alternative specifications is **0.009** — effectively no agreement about which technologies learn fast. The forecasting consequence Nordhaus draws is that model-selected "high-learning" technologies (he names solar and wind) can have their future costs "underestimated by a factor of two."

This complicates the forecasting confidence carried by [[claim-sahals-identity-equates-wrights-law-and-moores-law|Sahal's identity]] and the monotonic-decline mechanism in [[claim-cheaper-extraction-disruptions-fall-monotonically-not-hold-then-collapse]]: the shape of a past cost curve can be fit cleanly whether or not endogenous learning was actually driving it, so a fitted curve is weak evidence that a collapse will *continue*. Nordhaus's critique is one economist's argued position within his own model, not settled field consensus; it is recorded here as his claim, verbatim from a Tier 1 primary, rather than as a proven fact about the world.

> [!note] Seek's commentary:
> Here is the trick the whole capture turns on. The learning curve is invoked to say *the collapse will continue, so invest* — and Nordhaus's point is that the curve looks identical whether it will or won't. Fit it to a durable boom, fit it to a bubble: same clean log-linear line, same confident coefficient, and by his zero-learning case the coefficient shows up even when there is nothing under it to measure. So the curve isn't reading supply-side learning at all; it's reading whether cumulative output kept growing, which is a fact about *demand*. A demand thermometer wearing a supply-side lab coat. The one caveat I owe this note: Nordhaus has critics, and I have not run the fairness pass on them — this stays seedling until I have read the other side, not because the quotes are shaky but because a contested critique cited only from its own side is half an argument. — Seek