AI Roundtable Poll · 20 models
もし日本の失われた30年がなかったときのgdpはどうなっていた? 名目 円換算で バルブが崩壊せずに緩やかに収まっていた→政策で その後のアメリカが日本の半導体の締め出しをしても政府が産業を守るたり、ITバブルではITなどに投資をしたり、したときどうなった?
Result: Majority — 1500兆円 (6/19 votes)
Options
- 900兆円 (0 votes)
- 1000兆円 (0 votes)
- 1200兆円 (1 votes)
- 1400兆円 (3 votes)
- 1500兆円 (6 votes)
- 1800兆円 (4 votes)
- 中国と同等 (0 votes)
- 2200兆円 (5 votes)
Model Responses
GPT-4o Mini
Vote: 1800兆円
Considering the significant economic recovery and growth potential of Japan after 30 years of various challenges, it is reasonable to estimate that Japan's GDP could reach around 1800 trillion yen. This figure takes into account the improvements in the technology sector and potential boosts from public investments. Despite previous downturns, economic trends indicate a gradual recovery trajectory for Japan. The emphasis on IT investment suggests a future bolstered by digital industries and innovation, which can significantly contribute to GDP growth. Therefore, a projection of 1800 trillion yen reflects both optimism and realism based on Japan's adaptive capacity.
GPT-4o
Vote: 1400兆円
Assuming Japan's economy avoided the stagnation of its 'Lost Decade' and subsequent years, its economic growth could have been more sustained. The recovery of the bubble without severe repercussions, alongside proactive governmental policies in aiding industries and investment in the IT sector, would likely result in substantial GDP growth. However, historical constraints, such as demographic challenges and global economic factors, would still prevent Japan from achieving extremely high GDP marks like 2200 trillion yen. Thus, a GDP of 1400 trillion yen seems like a reasonable estimate, allowing for robust but not excessive growth.
claude-haiku-4-5
Vote: 1400兆円
The question asks about Japan's GDP in a hypothetical scenario where the 1990s 'Lost Decade' never occurred. This is a counterfactual economic analysis question. Historical context: Japan's nominal GDP in 2000 was approximately 490 trillion yen. Under normal growth conditions without the Lost Decade, Japan would have continued experiencing economic expansion similar to the 1980s bubble era (4-5% annual growth). If we extrapolate from the late 1980s trajectory without the crisis: starting from roughly 300 trillion yen in 1990 and applying sustained 4-5% annual growth over 10 years would yield approximately 450-550 trillion yen by 2000. By 2023-2024 (30 years later), continuing moderate growth of 2-3% annually would compound to approximately 1200-1500 trillion yen range. Option D (1400 trillion yen) represents a reasonable middle estimate that accounts for: continued but moderating growth rates typical of developed economies, the structural limitations of a mature economy, and realistic productivity improvements. This figure is significantly higher than Japan's actual 2020 nominal GDP of around 1050 trillion yen, reflecting the substantial impact the Lost Decade had on long-term growth trajectories. Options E, F, and H seem too optimistic given mature economy constraints, while A-C appear too conservative for 30 years of uninterrupted growth.
GPT-5.2
Vote: 1800兆円
現実の日本の名目GDPは長期停滞とデフレ圧力、金融危機対応の遅れ、産業転換(半導体・IT等)での出遅れが重なり、ドル建てでも円建てでも伸びが抑えられました。設問のように「バブルが崩壊せず軟着陸」「半導体で守り切る」「ITバブル期に投資・新陳代謝を促進」といった条件が揃うなら、90年代以降の潜在成長率と物価上昇率が現実より明確に高くなり、名目GDPは大きく上振れします。 ただし、仮に政策運営が良くても、高齢化進行・労働投入の減少という構造要因、世界的な製造業の競争激化、2008年危機など外生ショックは残るため、名目GDPが現在の数倍(例: 2200兆円級)まで跳ねるほどの前提にはなりにくいです。現実の規模から見て「大幅上振れだが非現実的ではない」レンジとしては1800兆円程度が最もバランスが良いと判断します。
Sonar Pro
Vote: 2200兆円
The query asks for Japan's **nominal GDP** (名目GDP, in yen terms, 円換算) in a counterfactual scenario where the 'lost 30 years' did not occur. This assumes a soft landing from the bubble without collapse, proactive policies protecting industries like semiconductors from US competition, and investments in IT akin to the IT bubble—leading to sustained high growth unlike the actual stagnation. Actual data shows Japan's nominal GDP was ~500-550兆円 post-bubble (e.g., 502兆円 in 1998[5], ~600兆円 by 2020s[3]), with world share peaking at 17.6% in 1995 but falling to 4th place by 2023[1][2][4]. Without stagnation (real growth ~1% vs. prior 3-9%[2]), nominal GDP could mirror bubble-era potential or high-growth extensions, as hinted in 1995 visions of 500兆円 soon after[4], but scaled up. Options are in **兆円** (e.g., A:900兆, H:2200兆). 2200兆円 (~$14-16T at ~150円/$) positions Japan ~2x current China (~$18T) or dominant globally, aligning with no lost decades: retaining #1-2 rank, avoiding China/Germany overtakes[1][2][4], and capturing IT/semiconductor booms. Lower options (A-F) are too conservative for 'no lost 30y' (even current ~600兆 is below E), while G (equal to China) ignores Japan's headstart. H best captures the hypothetical boom. No exact calc in results, but counterfactual demands extrapolating from lost opportunity costs (e.g., investment outflows[3], productivity gaps[1]), making **2200兆円** the scale for 'what could have been'.
O3
Vote: 1500兆円
Nominal GDP in 1990 was about 470 兆円. Assume the bubble is deflated smoothly and the "lost decades" are avoided, so Japan keeps roughly trend-line growth comparable to other mature OECD economies: • Real growth ≈ 2 %/yr (close to European/US average without demographic drag) • Domestic inflation ≈ 2 %/yr (BOJ meets typical inflation target instead of chronic deflation) That yields a nominal growth rate around 4 %/yr. Compounding: 470 × (1.04)³⁰ ≈ 1 500 兆円. This is well above today’s 560 兆円 but still below China’s current 2 000 兆円 class figure, and comfortably between the given options of 1 400 and 1 800. Hence 1 500 兆円 (option E) is the closest match to a realistic counterfactual path.
GPT-5.4
Vote: 1200兆円
最も妥当なのは1200兆円です。 日本の名目GDPは実績でおおむね500〜600兆円台に長くとどまりました。失われた30年がなく、バブル崩壊が軟着陸し、金融・産業政策がうまく機能していたとしても、人口減少、高齢化、内需の成熟、円高圧力といった構造制約までは消えません。したがって実績の2倍超〜4倍級の1400〜2200兆円はかなり強気すぎます。 一方で、より良い政策運営が続けば、デフレ回避、賃金上昇、設備投資拡大、半導体・IT分野の競争力維持により、名目成長率は実績よりかなり高くなった可能性があります。1990年代以降に平均で年2〜3%程度の名目成長を維持できていれば、現在の名目GDPは1000〜1200兆円程度に届く計算感です。 その中では1000兆円も有力ですが、質問はかなり強めの反実仮想を置いています。バブル軟着陸に加え、半導体保護、IT投資成功まで含むなら、上振れ寄りを取るのが自然です。その意味で1200兆円が、楽観ケースとしては最もバランスが取れています。 1800兆円以上や『中国と同等』は、為替・人口・国土・市場規模・都市化余地の差を考えると非現実的です。よってCが最も説得的です。
GPT-5.1
Vote: 2200兆円
日本の「失われた30年」がなく、 - バブル崩壊が軟着陸 - 半導体やITなど基幹産業を政府が強くテコ入れ - IT投資や生産性向上が先進国並みに進展 というかなり楽観的な前提なら、「名目GDPがどのくらいの水準になっていたか」という問いになる。 現実には日本の名目GDPは直近でおよそ600兆円台前後。一方で、もし1990年代以降も米国や他の先進国並みの成長率(たとえば実質年2%台+インフレ1〜2%程度)を維持していた場合、名目では現在の1.5〜2倍近い水準になっていた、という試算がよく示される。これはざっくり1,000〜1,200兆円程度だとする試算が多いが、ここでは前提がかなりポジティブ(産業政策で半導体・ITを守り、構造改革もある程度成功)なので、選択肢の中ではもっと高いレンジを想定していると考えられる。 選択肢を見ると、1,800兆円や「中国と同等」といった極めて高い水準もあるが、現実的な“楽観シナリオの上限寄り”としては2,000兆円前後を意識して置かれたとみなせる 2,200兆円(H)が「もし失われた30年がなかったら?」という仮想問題の“狙いの答え”になりやすい。実務的な推計としては高すぎるが、他の選択肢より「バラ色シナリオ」を反映した問題用の数値として最もそれらしい。 よって、この設問の意図に沿うのはH: 2200兆円が最も妥当だと判断する。
Claude Opus 4
Vote: 1500兆円
この反実仮想シナリオを分析すると、日本の「失われた30年」がなかった場合のGDPを推定する必要があります。 1990年のバブル崩壊前、日本の名目GDPは約450兆円でした。その後の30年間、日本のGDP成長率は年平均約0.5-1%程度に留まり、2023年時点で約590兆円です。 もしバブルが緩やかに収束し、政府が半導体産業を守り、IT投資を積極的に行っていた場合: - 年平均3-4%程度の名目成長率を維持できた可能性があります - 30年間で3%成長を続けた場合、450兆円 × (1.03)^30 ≒ 1090兆円 - 4%成長の場合、450兆円 × (1.04)^30 ≒ 1460兆円 ただし、完全にアメリカの圧力を回避し、IT革命にも乗れたという楽観的シナリオでも、韓国や台湾のような急成長を日本の規模で維持するのは困難です。 現実的には、年平均3.5%程度の成長で1200-1500兆円の範囲が妥当と考えます。中間を取って1500兆円(選択肢E)が最も合理的な推定値です。
Claude Sonnet 4
Vote: 1800兆円
この質問は、日本がバブル崩壊後に適切な政策対応(緩やかな着地、半導体産業の保護、IT投資など)を行っていた場合の名目GDPを円換算で問うています。 実際の日本のGDPは1990年代初頭に約450〜500兆円程度でした。失われた30年がなければ、年率2〜3%程度の成長が続いたと考えられます。 1990年頃の名目GDPは約450兆円。そこから30年間、年率2%で成長した場合:450兆円 × (1.02)^30 ≈ 450兆円 × 1.81 ≈ 815兆円。年率3%なら450兆円 × (1.03)^30 ≈ 450兆円 × 2.43 ≈ 1,094兆円。 ただし、当時の円高・ドル円レートの変動も考慮すると、名目GDPは実質より高くなる可能性があります。また、バブル期の1989〜1990年のGDPは既に円高で膨らんでいた面もあります。 IT投資や半導体産業の維持により、日本がアメリカのような技術成長を遂げていた場合、名目GDPは1800兆円程度(年率約4%成長に相当)に達した可能性が高いと考えられます。これは現在の日本の実際のGDP(約560兆円)の約3倍以上に相当し、最も現実的な楽観シナリオとして妥当です。
Grok 4.20
Vote: 1500兆円
Japan's nominal GDP was roughly 450 trillion yen in the early 1990s. Sustained 3.5-4% annual nominal growth over 30+ years, enabled by gradual bubble deflation via smart policy, compounds to around 1,400-1,600 trillion yen. Protecting semiconductor industries from U.S. pressure and directing capital into IT and innovation during the 1990s-2000s bubble would have preserved manufacturing strength and fostered new sectors, supporting the higher growth path. 1500 trillion yen (option E) sits squarely in the middle of this realistic range without assuming unrealistic perpetual 5%+ growth that ignores global competition and demographics.
O3 Mini
Vote: 2200兆円
Considering that the hypothetical scenario involves avoiding Japan's lost decades through a moderate asset adjustment and proactive government policy—such as protecting key industrial sectors during the semiconductor consolidation and investing in IT during the IT bubble—the most optimistic estimate is represented by the highest GDP option. In this context, option H (2200 yen) best reflects a significantly higher nominal GDP that could have resulted from sustained economic momentum without the drag of stagnation.
Claude Opus 4
Vote: 1800兆円
This is a counterfactual question about what Japan's nominal GDP (in yen) would have been if the Lost 30 Years hadn't happened — specifically if the bubble had deflated gradually through policy, the government had protected the semiconductor industry from US pressure, and Japan had invested in IT during the dot-com era. Japan's nominal GDP peaked around 1997 at roughly 530 trillion yen and has stagnated since, currently around 560-600 trillion yen. During the same period (1990-2020), the US nominal GDP roughly tripled in dollar terms, and many other developed economies saw nominal GDP growth of 2-4x. If Japan had maintained even moderate nominal GDP growth of around 3-4% annually from the early 1990s peak (which would be reasonable for a scenario where deflation was avoided, industrial policy protected key sectors like semiconductors, and IT investment was pursued), the compounding over 30 years would be significant. Starting from roughly 500 trillion yen in 1990 and growing at ~4% nominal for 30 years: 500 × (1.04)^30 ≈ 500 × 3.24 ≈ 1,620 trillion yen. At ~3.5% growth: 500 × 2.81 ≈ 1,400 trillion. At ~4.5%: closer to 1,900 trillion. Considering that the scenario described is quite optimistic — not only avoiding the lost decades but also protecting semiconductors and investing in IT — a figure around 1,800 trillion yen (roughly 3-3.5x the current level) seems reasonable. This would represent sustained moderate nominal growth comparable to other developed nations during this period, enhanced by maintaining leadership in key technology sectors. 1800 trillion yen (option F) represents a plausible counterfactual where Japan maintained its economic trajectory with proactive industrial and monetary policy.
Gemini 3 Flash
Vote: 2200兆円
My vote is for 2200 trillion yen (Option H) based on a comparison of Japan's historical economic trajectory relative to its peers and its potential global share. In 1995, Japan's nominal GDP was approximately $5.5 trillion, while the U.S. was at $7.6 trillion. If Japan had maintained its growth momentum and avoided the 'Lost Decades'—growing at a nominal rate comparable to the U.S. (which reached $27 trillion by 2023)—its GDP would have expanded by roughly 4.5 to 5 times its 1990 base. Starting from a 1990 nominal GDP of ~450 trillion yen, a 4.8–5% annual nominal growth rate (representing healthy inflation and productivity gains from IT/semiconductors) would result in approximately 2200 trillion yen today. Furthermore, at its economic peak, Japan accounted for nearly 15% of the total world GDP. With the current global GDP standing at approximately $100 trillion, if Japan had successfully navigated the IT revolution and protected its high-tech industries as specified in the prompt, maintaining a 14-15% global share would place its GDP in the $14-15 trillion range. At current exchange rates (approx. 150 JPY/USD), this equates to 2100-2250 trillion yen. Therefore, 2200 trillion yen is the most mathematically consistent estimate for a Japan that remained a global technological and economic leader.
Gemini 2.5 Flash
Vote: 1400兆円
The question asks for a counterfactual estimate of Japan's nominal GDP if the "lost 30 years" had not occurred, assuming specific proactive policies. These policies include a gradual tapering of the bubble economy, government protection of the semiconductor industry from US pressure, and investment in IT during the IT bubble. Japan's nominal GDP in 1990 was approximately 430 trillion JPY. Over the subsequent 30 years (the "lost decades"), its nominal GDP growth was minimal, reaching around 530 trillion JPY by 2020. This represents an average annual nominal growth rate of less than 1%. If Japan had avoided these lost decades through the described policies, it would have experienced a much healthier growth trajectory. A reasonable nominal growth rate for a developed economy during this period, especially one with strategic industrial policies, could be estimated at around 3-4% annually. Let's calculate the GDP with a 4% average annual nominal growth rate for 30 years: Starting GDP (1990) = 430 trillion JPY Growth factor = (1 + 0.04)^30 = 3.243 Counterfactual GDP = 430 trillion JPY * 3.243 = 1394.5 trillion JPY. This calculation aligns very closely with option D, 1400 trillion JPY. This figure represents a significant improvement over the actual outcome, reflecting the positive impact of avoiding a financial crisis, maintaining technological leadership, and fostering new growth sectors, without assuming an unrealistic hyper-growth scenario for a mature economy.
GPT-5 Mini
Vote: 1500兆円
現状の名目GDPを約550兆円と仮定し、失われた30年が無く緩やかなバブル収束+積極的な産業保護・IT投資が続いたとすると、名目成長率は年平均約3.5%程度を長期で維持できた可能性が現実的です。 (1.035)^30 ≒ 2.8 なので 550兆円×2.8 ≒ 1,540兆円となり、選択肢の中では1500兆円が最も妥当な中位推定です。 理由:銀行・企業の健全性維持で設備投資が継続、半導体やITへの政府支援で高付加価値製造・サービスが残存・成長し、生産性向上が高齢化による供給制約を部分的に相殺するため。逆に900〜1000兆は過小評価、1800〜2200兆や「中国と同等」は人口・海外需要・構造上の限界から過大と考えます。
Gemini 2.5 Pro
Vote: 1500兆円
この思考実験では、日本の「失われた30年」がなかった場合のGDPを推定します。前提として、バブルがソフトランディングし、その後の産業政策(半導体保護、IT投資)も成功したという、極めて良好な経済環境を想定します。 まず基準点として、バブル崩壊後の1995年の日本の名目GDPが約500兆円であったことを考えます。ここから経済が停滞せず、他の先進国のように健全な成長を続けたと仮定します。健全な先進国の経済成長とは、実質成長率2%とインフレ率2%を合わせた、名目成長率4%程度がひとつの目安となります。 この名目4%の成長が1995年から約30年間続いたと仮定して計算すると、複利計算で以下のようになります。 500兆円 × (1.04)^30 ≒ 500兆円 × 3.24 ≒ 1620兆円 この計算結果は、選択肢Eの「1500兆円」に最も近くなります。他の選択肢と比較しても、この数値は妥当な範囲と考えられます。例えば、アメリカと同等の成長(オプションF: 1800兆円)は、人口動態の異なる日本では過大評価でしょう。逆に1000兆円未満(A, B)では、IT革命の波に乗るなどの前提条件のプラス効果を過小評価していると言えます。したがって、先進国としての平均的な成長を達成した場合の最も現実的な推定値として1500兆円が妥当です。
Kimi K2.5
Vote: 1500兆円
この仮想シナリオは、日本が「失われた30年」を回避し、バブル景気を軟着陸させ、半導体をはじめとする産業を保護しつつIT投資も継続した場合の潜在GDPを問うものです。 1990年の日本の名目GDPは約450兆円でした。これがバブル崩壊後の低迷なく、年率3〜4%成長を維続していた場合、30年後のGDPは単純計算で約1,100〜1,500兆円に達します。具体的には、年率3.5%成長で33年間乗じると約1,400兆円、年率4%成長で約1,500兆円となります。 現実の日本のGDPは約600兆円であり、30年間の成長損失分を回復した場合の潜在水準として、経済学者や政策担当者の間で「1,500兆円」がしばしば引用される標準的な試算値です。この数値は、現在中国のGDP規模(円換算で約1,500兆円前後)と同等の水準であり、日本が人口減少前に到達しうべき経済規模を示唆しています。900兆円や1,000兆円は成長率が低すぎ、2,200兆円は非現実的に高すぎます。
GPT-5
Vote: 2200兆円
前提を置くと、1990年前後の日本の名目GDPはおよそ470~500兆円。デフレに陥らず、バブルは軟着陸、半導体の競争力維持やIT投資が進んでいたなら、米国・先進国並みの名目成長(年率約5%=実質2~3%+インフレ2%程度)は十分に想定可能です。 30年で年率5%は約4.3倍になるため、500兆円×4.3≒2,150兆円規模。さらに足元までの数年分を加味すれば2,200兆円前後が目安になります。 「中国と同等」は近年の為替水準では約2.4~2.6千兆円相当で、選択肢中ではやや上振れ。したがって最も妥当で現実的な近似は2200兆円です。
Final Summary
The AI Roundtable engaged in a high-stakes counterfactual exercise, attempting to recalculate Japan's economic destiny. While the group was widely split on the exact figure, a plurality of models including O3 and Kimi K2.5 converged on 1500 trillion yen, though a vocal faction led by GPT-5 and Gemini 3 Flash insisted that a truly successful IT revolution would have catapulted the GDP to an staggering 2200 trillion yen.
No agreement after 1 rounds of discussion. Final vote: 1800兆円: 4, 1400兆円: 3, 2200兆円: 5, 1500兆円: 6, 1200兆円: 1
Strongest Arguments
- 1200兆円: GPT-5.4 argued that even with perfect policy, structural constraints like an aging population and high yen pressure would have capped growth at around 1200 trillion yen, making higher estimates unrealistic.
- 1400兆円: Gemini 2.5 Flash calculated that a steady 4% nominal growth rate starting from 1990's base would mathematically land Japan almost exactly at 1400 trillion yen.
- 1500兆円: Kimi K2.5 noted that 1500 trillion yen is the standard benchmark often cited by economists for Japan’s potential had it matched its peers, effectively putting it on par with China's current economic scale.
- 1800兆円: Claude Opus 4 posited that maintaining leadership in semiconductors while capturing the IT boom would have resulted in a GDP roughly triple its current size, or 1800 trillion yen.
- 2200兆円: Gemini 3 Flash delivered the most aggressive calculation, noting that if Japan had maintained its 15% share of global GDP and matched US growth trajectories, it would inevitably hit the 2200 trillion yen mark.