竹中平蔵批判をファクトチェック｜怒りはGDPに足せない——大村大次郎・MONEY VOICEの計算と政策責任を検証｜Anger Does Not Add to GDP
Dislike is not evidence Disliking Heizo Takenaka does not make an article attacking him correct. Criticism of a disliked policymaker requires especially careful arit
Dislike is not evidence
Disliking Heizo Takenaka does not make an article attacking him correct. Criticism of a disliked policymaker requires especially careful arithmetic; weak arguments offer the target an escape from legitimate accountability.
My verdict: the wage-distribution question is legitimate, but the article contains a fundamental GDP-accounting error and does not establish that one person was the principal cause of Japan’s prolonged stagnation. I would not adopt it as a verified economic explanation. That judgment does not absolve Takenaka of policy responsibility.
This review covers the free public article, not the entire paid newsletter. The site credits the headline and headings to its editors, so responsibility for framing must be distinguished from responsibility for the author’s argument.
GDP and consumption
GDP’s production, income and expenditure measures describe the same economy. They are not amounts to add together.
In an illustrative company whose production and sales remain unchanged, an additional ¥1 million of wages offset by ¥1 million less profit changes income distribution, not current value added. If households subsequently spend an additional ¥600,000, the original wage payment cannot be added again to obtain ¥1.6 million of GDP growth.
Additional domestic production can respond to demand, but imports, inventories, prices and offsetting changes in investment or other consumption matter. Gross wages also differ from disposable income.
Average propensity to consume means consumption divided by disposable income. It is not marginal propensity to consume, the share of additional income spent. In a simplified closed-economy model, an assumed marginal propensity of 0.6 gives an autonomous-expenditure multiplier of 2.5. That is neither an empirical estimate for Japan nor an automatic multiplier for a transfer from profits to wages.
Wage-led demand is a serious argument. It does not need incorrect national accounting.
Wages and currency
The labour ministry’s 2026 committee materials report that nominal wages rose 2.3% in 2025 while real wages fell 1.3%. Higher pay and lower purchasing power can coexist.
Long-term comparisons must distinguish annual pay from hourly wages, working time, employment composition and inflation. I did not verify the original 2019 Nikkei wage article and its underlying series, so I do not certify or reproduce that international index.
Dollar GDP also changes with exchange rates. Illustratively, unchanged yen GDP converted at ¥165 rather than ¥150 per dollar falls about 9.1% in dollar terms. This is arithmetic, not a forecast. Living standards require additional measures, including per-capita and real indicators. The specified IMF vintage and 2027 country values could not be reproduced here; the ranking prediction remains unverified, not disproved.
Profits, investment and retained earnings
The finance ministry’s FY2025 corporate survey reports ordinary profits excluding finance and insurance of ¥124.8313 trillion, up 8.8%. This supports the approximate ¥125 trillion scale, but the period is a fiscal year.
The same table reports investment including software of ¥57.8999 trillion, up 4.3%. Investment can grow while remaining inadequate relative to profits or replacement needs.
The Cabinet Office’s historical CSV gives 2021 investment indices, with 2000 equal to 100: Japan 108.7, United States 177.5, United Kingdom 116.0, Germany 124.7 and France 151.4. These match the article’s table. The chart concerns private business investment in real national-currency terms. It compares growth from each country’s own baseline, not investment levels across countries, and ends during the pandemic.
The accompanying official analysis distinguishes weak domestic investment from rising overseas investment and acquisitions. Weak domestic investment does not establish that companies invested nowhere.
Retained earnings are accumulated profits on the financing side of the balance sheet. Corresponding assets may include cash, equipment, intangibles or investments. An empirical argument about inadequate investment cannot overturn that accounting definition. Nor does aggregate profitability prove that every business has identical wage-paying capacity. Ordinary corporate profits are not the same as GDP operating surplus; overseas dividends and intercompany transactions complicate direct comparisons. I did not fully reproduce the retained-earnings time series or its claimed international ranking.
Takenaka’s record deserves scrutiny
The national weighted-average minimum wage was ¥663 in FY2001, ¥663 in 2002, ¥664 in 2003, ¥665 in 2004, ¥668 in 2005 and ¥673 in 2006: an increase of ¥10, about 1.5%, from FY2001. Annual revisions do not precisely coincide with cabinet terms. The increases were nevertheless small.
Minimum wages pass through central and regional deliberations; the economic-policy minister did not set them alone. Official cabinet records establish Takenaka’s economic-policy role from 2001.
Dispatch-work liberalization also has a timeline. General liberalization occurred in 1999; manufacturing dispatch was expanded through the 2003 amendment implemented in 2004. Agency workers are a subset of nonregular employment, not its entirety.
Pasona Group’s official biography records his appointment as chairman in August 2009. In July 2022 the company announced his planned departure. Policy involvement followed by a role in a related industry warrants conflict-of-interest and revolving-door scrutiny. It does not independently establish that earlier policies were designed to enrich that company.
My criticism is that advocates of labour flexibility and corporate efficiency owe an evidence-based account of protection for bargaining power, training, transitions into regular employment and income stability. A later industry role increases the need to explain separation of policy judgment from private interests. This is an accountability standard, not a criminal finding.
Counterevidence belongs in the record too: the Financial Services Agency reported major banks’ nonperforming-loan ratio falling from 8.4% in March 2002 to 2.9% in March 2005, meeting the financial rehabilitation programme’s target. That is a financial-health outcome, not proof of one person’s sole achievement and not an excuse for labour-policy shortcomings.
The writer and editors also bear responsibility
The Cabinet Office dates a 73-month expansion from the January 2002 trough to the February 2008 peak. Counting expansion from the following February is consistent with that duration. Duration does not establish strong growth or broad household prosperity.
Assigning primary responsibility for decades of stagnation requires comparison with demographics, working hours, productivity, deflation, financial crises, overseas demand, investment and subsequent administrations. Temporal overlap does not identify a causal contribution.
A former tax-official credential is not proof of sound macroeconomic inference. Dismissive treatment of disagreement can encourage identification with a camp rather than examination of evidence. Selling a newsletter is not itself improper, but does not excuse factual or analytical errors. The author should answer for the calculation; editors should answer for framing they produced.
The article mixes a real distributional problem with mistaken accounting and unsupported causal concentration. That undermines reader understanding. One article does not establish that every item on the platform is false, nor does it reveal the writer’s private intent.
A crowd-psychology hypothesis, not a measured effect
My hypothesis is that shared grievance, a single personalized culprit and dismissal of dissent can create a feeling of understanding. This is analysis of the article’s structure, not evidence that its audience actually experienced that effect.
Testing it would require comparison with neutral wording using the same data, measurement of causal understanding and observation of responses to correction. Feeling anger alone proves neither manipulation nor irrationality.
For investing, political criticism needs a specific transmission mechanism into wages, prices, productivity, contracts, labour supply or compliance costs. Dislike of a politician does not establish a share-price direction.
Demand distributional results and disclosure of interests from Takenaka. Demand correct arithmetic, definitions and causal evidence from the writer and editors. The evidentiary threshold should not fall because the target is unpopular.
All source links and verification limits appear in the common bibliography below.
