Pre-registration
Share of value added going to financial-sector profits in the US rose from under 15% in 1970 to over 30% by 2007, consistent with Hilferding-style financialisation hypothesis.
Falsification criterion — what would disprove this
This hypothesis is considered falsified if:
PRIMARY (dispositive): the hypothesis is SUPPORTED if BOTH endpoint conditions hold for the constructed series financial_share = A453RC1Q027SBEA / (A453RC1Q027SBEA + A446RC1Q027SBEA), annual mean of quarterly observations: (a) 1970 mean financial_share <= 0.15 (i.e. ~<15%), with tolerance band [0.10, 0.18] treated as "near threshold"; values strictly above 0.18 falsify the start-condition. (b) 2007 mean financial_share >= 0.30 (i.e. ~>30%), with tolerance band [0.27, 0.33] treated as "near threshold"; values strictly below 0.27 falsify the end-condition. REFUTED if EITHER endpoint is on the wrong side of its tolerance band (1970 > 0.18 OR 2007 < 0.27). PARTIAL if direction is correct (rise from 1970 to 2007 of at least +10pp) but one endpoint sits inside its tolerance band rather than meeting the strict threshold. INFORMATIVE: the change in financial_share between 1970 and 2007 (delta_pp); the implied compound annual change; and the year in which financial_share first crosses 0.20 and 0.30. Reported in diagnostics. METHOD_VALID: both FRED series A453RC1Q027SBEA and A446RC1Q027SBEA must be available on disk with quarterly observations covering 1970Q1 and 2007Q4. If either series is missing or has fewer than 3 quarterly observations in the 1970 or 2007 calendar year, emit `inconclusive (data gap on fred:<series>)` rather than a verdict.
formal test & threshold
test: us_financial_corporate_profit_share_endpoint_levels_1970_vs_2007 threshold: PRIMARY: financial_share_1970 <= 0.15 AND financial_share_2007 >= 0.30, where financial_share = A453 / (A453 + A446) and the annual figure is the mean of within-year quarterly observations. Tolerance band ±0.03 around each threshold; outside-band misses on the wrong side falsify.
Method
- Template
descriptive- Clustering
none- Sample
- 1 countries · 1970 – 2007
- Evidence type
- descriptive
Single-country US time series of financial-sector profit share of domestic non-financial-corporate profit and of total value added, 1970-2007. Reports cumulative change and trajectory; magnitude is dispositive. No causal identification claim.
Data
| Variable | Source | Transform |
|---|---|---|
financial_sector_profit_share outcome | fred:A453RC1Q027SBEAtier 1 fred:A446RC1Q027SBEAtier 1 | — |
● ready · ● pending · ● reconstruct-needed
Detailed result card
US financial-sector profit share, 1970 vs 2007
Verdict: refuted — endpoint outside falsification band: 2007 share = 7.1% < 27% falsify-band. Direction wrong (-6.5pp 1970→2007).
Summary
- 1970 financial_share = 13.6% (target ≤ 15%; falsify-band: > 18%).
- 2007 financial_share = 7.1% (target ≥ 30%; falsify-band: < 27%).
- Δ 1970→2007 = -6.5pp (-0.17pp/yr).
- First year with share ≥ 20%: 1981.
- First year with share ≥ 30%: 1982.
Method
Constructed series: financial_share = A453RC1Q027SBEA / (A453RC1Q027SBEA + A446RC1Q027SBEA). Both inputs are quarterly billions-of-dollars NIPA flows (financial corporate profits before tax, and domestic non-financial corporate profits before tax). The annual figure is the simple mean of within-year quarterly observations of each numerator/denominator, then the share is constructed from the annual means.
Endpoint tests:
- PRIMARY (a): share[1970] ≤ 0.15. Falsify-band: > 0.18.
- PRIMARY (b): share[2007] ≥ 0.30. Falsify-band: < 0.27.
If either endpoint sits inside its tolerance band but the direction is correct (Δ ≥ +10pp), the verdict is partial. If either is on the wrong side of its falsify-band, the verdict is refuted. If the input vintages are missing, the verdict is inconclusive.
Data
- fred:A453RC1Q027SBEA (financial corporate profits before tax, quarterly)
- fred:A446RC1Q027SBEA (domestic non-financial corporate profits before tax, quarterly)
Notes
Seeded from a Marxian financialisation claim that US financial-sector profit share rose from <15% (1970) to >30% (2007). Descriptive single-country time series; magnitude is dispositive but sector definitions require human review.