Country Posture — Methodology
Aavistus · methodology v0.2 · last updated 2026-05-03
This page documents how the Aavistus Country Posture engine computes its scores, what data it uses, what it shows accurately, and — critically — every known bias that could push the headline numbers in either direction. We publish this audit because the framework is honest only when the biases are surfaced explicitly.
Direction discipline. Some biases push our headline numbers to look MORE pessimistic than reality; others push them to look MORE optimistic. The audit below labels each. The framework’s directional claim — that younger cohorts hold less wealth at the same age across most developed economies — is robust to all known biases. The MAGNITUDE of the gap is uncertain in either direction; we report numbers to one decimal but acknowledge they should be read as approximate.
Why our fiscal numbers may differ from your national source
Cross-country rankings need a single, harmonised data convention. We use IMF World Economic Outlook as the canonical source for fiscal pillar metrics (general government gross debt, primary balance, GDP growth, inflation). This is the same dataset cited by IMF Article IV reports, OECD Economic Outlook, and most academic cross-country comparison papers — chosen because it’s defined the same way for every country in our coverage.
What this means in practice. Our debt-to-GDP figures will run a few percentage points higher than the equivalent national-accounts numbers most people see in domestic news coverage. For example, the IMF measure of Finland’s general government gross debt for 2024 is about 82% of GDP. If you read Tilastokeskus’ Maastricht-definition figure for the same year, you’ll see something closer to 78%. Both are real numbers. They differ for definitional reasons — IMF’s GGXWDG includes some liability classes that Maastricht consolidation excludes. The gap is structural, not error.
The same applies for primary balance and other fiscal metrics. We anchor on IMF because it’s the only source that lets us put Finland’s number next to Germany’s number next to Japan’s number without comparing apples to oranges.
For demographic data (TFR, working-age growth, net migration) we use a similar harmonisation discipline — Eurostat and OECD Population Statistics where available, with national-source backstops only where the harmonised series doesn’t cover a country.
Single-country deep dives can legitimately use national sources directly. A Finnish reader reading a Finnish-only piece may find Tilastokeskus a more appropriate citation than IMF. The framework’s rankings, however, stay anchored on the harmonised data so the rankings actually mean something.
How the Lifetime Cohort Wealth Ratio (LCWR) is computed
For a given country, cohort, and age:
LCWR(country, cohort, age) = real net wealth held by cohort AT age ÷ real net wealth held by the 1955 cohort at the same age.
Wealth in numerator and denominator are deflated by the country’s national CPI to a common base year. A ratio of 1.0 means the younger cohort is on track to match their grandparents’ generation at the same point in life. Below 1.0 means the deal is breaking. Above 1.0 means the cohort is doing better.
Sources. Resolution Foundation (UK), IFS (UK), Tilastokeskus + ETLA + PT (Finland), Bundesbank Panel on Household Finances (Germany), INSEE patrimoine (France), Banca d’Italia SHIW (Italy), Banco de España EFF (Spain), Statistics Sweden + Riksbank, DNB Household Survey (Netherlands), Cabinet Office (Japan), Korea Statistics Office, DGBAS (Taiwan).
Bias audit
Each bias below is labeled with the direction it pushes the headline LCWR (overstates / understates / variable / qualitative), its severity, where it bites, and how the framework addresses it.
🟡 CPI deflator choice
Direction: ↔ variable direction
Severity: medium
Where it bites: Each country’s national CPI deflates wealth in both numerator (younger cohort) and denominator (1955 cohort). National CPI baskets differ from harmonized HICP and from each other. Housing weight in particular varies by ±10 percentage points across covered countries.
Framework response: We use national CPI for within-country cohort comparisons (the headline number). Cross-country bars use each country’s internal experience, not PPP-converted absolute purchasing power. Future versions will also publish HICP-deflated parallel series for EU countries to support cross-country sensitivity checks.
🟡 Cross-country PPP
Direction: ↔ variable direction
Severity: medium
Scope: cross-country comparisons specifically
Where it bites: The comparative bar chart implies bars are comparable in absolute purchasing-power terms; they are not. Each country’s deflator is its own national CPI. The cross-country gap reflects each country’s internal cohort experience, not absolute lifestyle differences between countries.
Framework response: Footer disclosure on the comparative chart explicitly states ‘cross-country bars not PPP-converted.’ Future PPP-adjusted variant available for analytical use; default chart preserves internal-experience interpretation.
🔴 Basket-composition shift toward inelastic categories
Direction: ↓ understates the gap (pushes our headline less pessimistic)
Severity: high
Where it bites: The 1985 cohort spends 35-50% of income on housing in most covered countries; the 1955 cohort at 35 spent ~25%. Housing prices have risen 2-4x faster than general CPI since 1990 in most of these countries. Wealth deflated by general CPI is more wealth than wealth deflated by housing-weighted CPI. The headline LCWR therefore probably UNDERSTATES the deal-breaking.
Framework response: We publish housing-CPI vs general-CPI delta as a supplementary chart. The framework’s headline is robust to this bias because it pushes in our favor — actual welfare gap is larger than shown. We still flag it for transparency.
🟡 Hedonic / goods-quality adjustment
Direction: ↑ overstates the gap
Severity: medium
Where it bites: 2020 cars, phones, healthcare > 1990 equivalents in quality, safety, capability. CPI hedonic adjustment is contested but broadly accepted; some economists argue it understates actual quality improvement. The 1985 cohort consumes higher-quality goods even with lower nominal real wealth.
Framework response: We acknowledge the welfare interpretation is contested. The LCWR is a measure of accumulated wealth, not consumed-welfare. Both metrics matter. We do not aggregate them into a single score — we publish wealth alongside life-expectancy and goods-quality indicators for the same cohorts in supplementary briefs.
🟠 Public services net of taxes
Direction: ↑ overstates the gap
Severity: medium high
Where it bites: 1985 cohort lives in a more developed welfare state in most covered countries. Healthcare access, education subsidies, future pension claims, childcare provision — these are implicit wealth not in net-wealth surveys. Especially material for Nordics + Continental Europe.
Framework response: We publish a parallel ‘extended net wealth’ figure that adds estimated NPV of public services received per cohort, where data permits. The headline LCWR remains private-net-wealth-only to stay comparable across countries, but the extended series is available in country reports.
🔴 Inheritance flow not captured at age-of-measurement
Direction: ↑ overstates the gap
Severity: high
Where it bites: 1985 cohort will inherit more (in absolute and relative terms) than the 1955 cohort did. The chart says wealth-at-35 is lower; it does NOT say lifetime wealth will be lower. For inheritance-heavy countries (Italy, France) the gap effectively narrows considerably once expected inheritance is included.
Framework response: The framing ‘has X% less wealth at age 35’ is true and important in itself — it speaks to financial autonomy in young adulthood, household-formation decisions, and the lived experience of the cohort. Future inheritance does not erase current constraints. We publish ‘inheritance flow as % of household wealth formation’ as a separate Work-Society Contract sub-metric so readers can adjust their interpretation.
🟢 Cohort-size effect on per-capita inheritance
Direction: ↑ overstates the gap
Severity: low medium
Where it bites: Smaller 1985 cohort vs larger 1955 cohort means per-capita inheritance from boomer wealth will be larger. Concentration of inheritance among smaller cohorts mitigates the apparent gap on a lifetime-wealth basis.
Framework response: Captured by the inheritance flow metric. The within-cohort DISTRIBUTION of inheritance is highly unequal — concentrated in already-wealthy families — so per-capita averaging masks huge variance. We publish inherited-wealth Gini per cohort where available.
🟢 Survivor bias in older cohorts
Direction: ↑ overstates the gap
Severity: low
Where it bites: 1955-cohort wealth at age 65 only includes survivors. Lower-wealth peers more likely to have died. 1955 baseline at 65+ is therefore biased upward. This effect is small but real.
Framework response: We use 1955-cohort-at-35 (1990) and 1955-cohort-at-45 (2000) as the strongest baseline anchors — survivor bias is negligible at these ages. The 65+ comparison is reported but down-weighted in the composite.
🔴 Women’s economic emergence
Direction: ≈ changes interpretation, not magnitude
Severity: high qualitative
Where it bites: The 1955 cohort women had little independent wealth. Their wealth at 35 in 1990 reflects spousal / household wealth. The 1985 cohort women are individually measured. Part of the apparent ‘cohort wealth decline’ reflects ‘measurement era now includes economically-independent women whose mothers weren’t measured separately’ rather than literal welfare deterioration of comparable individuals.
Framework response: We publish gender-disaggregated LCWR for cohorts where data permits. The headline remains household-wealth-per-individual because that’s what wealth surveys measure consistently across cohorts. The women’s-emergence dimension is genuinely important and we surface it in the Work-Society narrative; it does not invalidate the directional finding.
🟢 Geographic mobility / emigration
Direction: ↔ variable direction
Severity: low
Where it bites: Korean and Taiwanese emigrants get measured where they live. Country-resident wealth surveys may understate the cohort’s actual position if high-earners have left. Especially relevant for KR / TW / IT in the leading-indicator set.
Framework response: Acknowledged in country-specific footnotes. Where emigration is large enough to matter (>5% of cohort), we publish a ‘born here’ variant alongside the ‘lives here’ variant.
Net read
Of the 10 known biases catalogued above:
- 5 push the headline number toward MORE pessimistic than reality
- 1 push the headline toward LESS pessimistic than reality
- 3 could push either way depending on country and cohort specifics
- 1 change the interpretation rather than the magnitude
These do not cancel cleanly. The directional claim — younger cohorts hold less wealth at the same age — is robust to every audit we can run. The magnitude of the gap is genuinely uncertain. Stating an LCWR of 0.72 with one decimal of confidence is overconfident; stating ‘the cohort wealth ratio is meaningfully below 1.0 across most developed economies, in the 0.5–0.85 range’ is honest.
Common critiques and responses
These are the rebuttals critics most predictably raise. We document them here so future readers find the framework’s response before they finish forming the critique. None of the critiques below are dismissed — each has a real basis. The responses are empirical, not rhetorical.
“People are objectively wealthier in 2020 than in 1990”
Where this is raised: general macro discourse, social media, financial press
The critique: Real GDP per capita roughly doubled in most covered economies since 1985. Goods quality, healthcare access, life expectancy, consumed welfare — all materially higher today. The chart implying younger cohorts are ‘worse off’ contradicts that.
The response: The critique conflates two distinct measurements. Real GDP per capita and consumed welfare have indeed risen — that’s consumption FLOW. The Lifetime Cohort Wealth Ratio measures balance-sheet net worth at age 35 — accumulated STOCK. These are different things and the chart specifically targets the second.
Consumption flow has improved, true. Median household wealth has not improved at the same rate — concentration effects captured GDP gains unevenly, and the COMPOSITION of wealth shifted (less housing equity, more pension-claim uncertainty, more debt netted against assets). ‘Society is richer’ and ‘the average 35-year-old owns more capital’ are both answerable separately, and the data say only the first is true. Wealth at age 35 predicts financial autonomy, household-formation choices, and crisis-resilience in a way consumption flow does not. A high-consumption / low-wealth cohort experiences security differently than a moderate-consumption / high-wealth cohort, even at identical income.
One-line rebuttal: “You’re describing consumption. We’re measuring wealth. They’re different. A 35-year-old today has a better phone and better healthcare than a 35-year-old in 1990, AND owns less productive capital. Both true.”
“Life expectancy rose, so the deferral compensates”
Where this is raised: thoughtful counter-thesis from informed readers
The critique: Life expectancy at birth rose ~7 years cohort-to-cohort. If the 1985 cohort lives 5 extra years, accumulating wealth 5 years later in life produces the same lifetime wealth — the same-age comparison at 35 is unfairly capturing only the early phase.
The response: The argument has a real basis but the data refute it on five independent grounds.
(1) Working-life span hasn’t extended. Education extension absorbed 2-4 years at the front (1985 cohort enters labor market at 22-24 vs 18-20 for 1955 cohort). Retirement age rose ~2 years. Effective working span is roughly identical at ~45 years.
(2) Compound interest math doesn’t close the gap. $50 at age 30 vs $100 at age 30, both growing at 5% real, over 35 years: $276 vs $552. Adding 5 extra years for the deferred cohort: $352 — still ~36% behind.
(3) Multi-age cohort data already disconfirms it. The 1965 cohort at 55 is BELOW the 1955 cohort at 55. The 1975 cohort at 45 is BELOW the 1965 cohort at 45. If the deferral hypothesis held, we would see catch-up at later ages. We do not.
(4) Healthy life expectancy lags total life expectancy. The extra years are mostly years of declining health — care-needing years, not productive accumulation years.
(5) Inheritance timing moved LATER, not earlier. Longer parental lives mean inheritance arrives at the heir’s age 65+ rather than 50, AFTER it could compound, AFTER it could fund household formation, AFTER it could buy housing at affordable prices. The longer-life mechanism the critic identifies actually deepens the cohort gap rather than compensating it.
One-line rebuttal: “The extra years arrived after working life ended, not extending it. Education absorbed the front-end gain; retirement age barely moved at the back. Inheritance timing actually moved later in heir’s life because parents live longer. The longer-life mechanism deepens the gap; it doesn’t fix it.”
“People consume more BECAUSE of higher debt — that’s the mechanism”
Where this is raised: supportive observation rather than counter-critique
The critique: Younger cohorts consume more (better cars, phones, services, housing-quality) but it’s debt-financed. Household debt-to-GDP roughly doubled across developed economies since 1990. Higher gross consumption and lower net wealth coexist because the consumption is paid for on credit.
The response: Correct, and this is exactly what the LCWR captures. The Lifetime Cohort Wealth Ratio is NET wealth — assets minus liabilities. The 1985 cohort’s consumption gain was substantially financed by debt expansion. Strip out the debt and the cohort wealth gap is visible directly.
Japan is the natural experiment. Japan didn’t allow the household-debt expansion that masked the wealth gap elsewhere — household debt/GDP is roughly flat since 1990 while every other developed economy doubled. And Japan shows the steepest LCWR decline in our dataset. Japan isn’t an outlier; it’s a leading indicator stripped of the lifestyle subsidy that easy credit provided everywhere else.
When critics say ‘people consume more,’ they’re correctly describing the symptom. What they’re missing is what financed it. The chart’s point is that the consumption is rented, not owned.
One-line rebuttal: “Yes — consumption rose AND so did debt. The chart measures net wealth (assets minus liabilities). Japan didn’t allow the debt expansion and shows the steepest wealth decline. Japan is the natural experiment with the mask removed.”
“China shows the framework is biased toward decline”
Where this is raised: criticism that the methodology pre-determines pessimistic findings
The critique: Aavistus shows broken-deal trajectories everywhere. China, the world’s #2 economy, has had the largest cohort wealth gain in human history. The framework’s findings would be biased if it always produced ‘deal-breaking’ results.
The response: China is exactly the case proving the framework is honest, not biased. Run the same methodology on China and the LCWR for the 1985 cohort comes in around 4-5× the 1955 cohort. The framework would report that, not hide from it. Direct deal-strengthening case.
What’s analytically interesting about China is that the cross-cutting Future Confidence dimension shows the withdrawal patterns appearing on schedule despite the wealth gain. Tang ping (2021), bai lan (2022), youth unemployment series suspended (2023), TFR ~1.0 (below Japan), marriage rate -47% since 2013, suspended-mortgage protests against unbuilt apartments (2022), property values down 15-25% from 2021 peak in tier-1 cities.
China is the cleanest evidence that wealth growth and confidence in the future are separable. The framework would correctly report rapid Work-Society Contract index gains AND simultaneous Future Confidence deterioration. That’s the analytical claim the dimensional structure is designed to surface, and China is the live case showing it works.
One-line rebuttal: “Run the methodology on China and it shows the opposite. The framework is honest — it reports whatever the data show. What’s interesting is that even with massive wealth growth, China’s cultural-withdrawal signals are appearing on schedule, exactly as the framework predicts.”
If you can articulate a critique not in this list — or you think we’re handling one of these badly — write to [email protected]. Critiques that hold up are added to this catalog with attribution.
How a verdict is produced
Every published thesis goes through a structured analytical method designed to surface dissent, score evidence quality, and preserve counter-arguments alongside the conclusion. The method is multi-perspective by construction — competing inquiry frames are run independently against the same evidence base and then synthesized into a structured report with explicit dissent preservation, calibrated confidence bands, and pivot data points that would flip the verdict if proven.
The method’s outputs are open: every published verdict includes the synthesized report, the dissent that was raised against it, the evidence-quality grade, the falsifiable predictions, and the calibration record at 12-month and 24-month resolution audits. The method’s implementation — the specific systems used to produce a verdict, the inquiry frames the analytical layer applies, and the synthesis discipline — is the proprietary analytical capital we have built. We document what the method achieves and audit how well it does so; we do not publish the recipe.
Calibration discipline
Every Aavistus Country Posture brief logs falsifiable predictions with explicit horizons (12 or 24 months), confidence levels, and disconfirmation conditions. At the horizon, predictions are scored as confirmed / disconfirmed / partial / inconclusive, and failure modes are categorized (data drift, narrative drift, structural surprise, analytical error). The full prediction history and calibration record is published on the Track Record page and is never gated behind a paywall — the calibration record IS the credibility moat.
What we welcome from critics
If you spot a bias not on this list — or think we’re misjudging the direction or severity of one we’ve listed — write to [email protected]. Methodology critiques that hold up are added to this page with attribution and the engine is updated. The framework is improved by good-faith critique, not by deflection.