方法说明
本页解释「我到底有多穷?」背后的数据来源、模型与假设。目标是完全透明 — 工具中出现的每个数字都可以追溯到来源与计算方法。
财富分布数据
财富份额数据来自 World Inequality Database,该数据库为数十个国家发布分布式国民账户(DINA)。这些账户把国家财富划分为以下群体:
- 底部 50% — 按净资产排序处于下半部分的人群
- 中间 40% — 第 50 到第 90 百分位,常被称作「中产阶层」
- 前 10% — 最富的十分之一,多数国家这部分人占总财富的 60–80%
- 前 1% — 前 10% 的子集,通常占总财富的 25–40%
这些份额构成了把你定位到分布中的边界点。某国边界数据越细,最终百分位就越精确。
由收入估算财富
大多数人知道自己的收入,但不知道净资产。为弥合这一缺口,工具使用 包含 18 个因子的估算模型,依据人口与财务特征调整收入对财富的比例:
- 年龄段(年轻人通常财富/收入比更低)
- 学历(高学历与终身收入及储蓄正相关)
- 就业形态(个体经营 vs. 受雇,公部门 vs. 私部门)
- 储蓄率与投资行为
- 是否拥有房产及房贷状态
- 未偿债务(学贷、消费贷)
每个因子都会缩小不确定区间。未填任何因子时,模型大约有 ±70% 的散布;填完 18 个因子时,不确定性降至约 ±10%。工具会与估算的百分位一并展示置信区间。
百分位的计算
在估算出净资产后,工具采用 分段线性插值,在已知的财富份额边界之间为你定位。例如,若底部 50% 占总财富的 5%、中间 40% 占 35%,则你处于这两个边界之间的位置按线性插值得出。这是一种近似 — 真实分布并不是严格线性的 — 但在已有数据下能给出合理估计。
亿万富豪对比
「要多久?」模式使用 Forbes 实时亿万富豪榜 列表,构建时为各国最富的人打包好净资产数值。
「需要多少年」是有意为之的简单算法:把亿万富豪的净资产除以你的年收入,不考虑利息、复利、税收或通货膨胀。这是有意的 — 重点不是理财规划,而是让差距「可感」。当答案是「400 万年」时,是否考虑 7% 的年化收益其实无关紧要。
税率数据来源
按财富层级划分的有效税率来自学术研究与政府统计。和财富分布数据不同,它没有统一的 API,需根据下面的公开来源逐一手工维护。
| Country | Source | Year |
|---|---|---|
| 🇦🇺Australia | ATO Taxation Statistics; Leigh (2009, updated); Grattan Institute distributional analysis | 2022 |
| 🇦🇹Austria | Statistik Austria, Integrierte Lohn- und Einkommensteuerstatistik (integrated wage and income tax statistics). Social contributions ~18% employee share, capped. Flat 27.5% KESt on capital income reduces effective rates for top brackets. Sub-percentile rates estimated from capital income concentration data. | 2023 |
| 🇧🇪Belgium | Statbel fiscal income statistics (decile distribution of net taxable income, total tax, and average tax rate, income year 2023). Very high social contributions (~13% employee). No general capital gains tax; 30% withholding on dividends. Sub-percentile rates estimated from OECD/WID capital income concentration. | 2023 |
| 🇧🇷Brazil | Receita Federal; Morgan (2017), WID.world; Gobetti & Orair (2017) | 2021 |
| 🇨🇦Canada | Statistics Canada; PBO Distributional Analysis 2023; OECD Tax Database 2024 | 2022 |
| 🇨🇱Chile | SII (Servicio de Impuestos Internos) bracket data; WID.world Chile distributional accounts; López & Figueroa (Harvard CID). Highly regressive due to 19% IVA consumption tax burden on lower incomes and favourable capital income treatment. Top 1% earn ~25% of pre-tax income (WID). Sub-percentile effective rates estimated. | 2024 |
| 🇩🇰Denmark | Danish Ministry of Taxation; OECD Tax Database 2024; WID.world Denmark series | 2022 |
| 🇫🇮Finland | Statistics Finland income distribution statistics; Verohallinto public tax data. Finland's dual income tax system: progressive earned income tax up to ~51.4%, flat 30–34% capital income tax. Sub-percentile rates estimated from capital vs labour income shares. OECD Taxing Wages 2025: average net tax rate 30.3%. | 2023 |
| 🇫🇷France | Landais, Saez & Zucman (2020); WID.world France series; EU Tax Observatory (2024) | 2022 |
| 🇩🇪Germany | Bach, Beznoska & Steiner (2020), DIW Berlin; Bundesfinanzministerium Datensammlung; OECD 2024 | 2021 |
| 🇮🇪Ireland | Revenue Commissioners income tax distribution tables; CSO Ireland's Tax Statistics 2024; Social Justice Ireland effective rate analysis (10.3% at €25K, 39.0% at €120K). Top rate = 40% income tax + 8% USC. Top 1% (>€203K) pay ~19% of personal tax. Entrepreneur relief (10% CGT) benefits top brackets. | 2024 |
| 🇮🇹Italy | Ministero dell'Economia; Acciari & Morelli (2023); EU Tax Observatory (2024) | 2022 |
| 🇯🇵Japan | National Tax Agency statistics; Moriguchi & Saez (2008, updated); OECD Tax Database 2024 | 2022 |
| 🇳🇿New Zealand | NZ Inland Revenue, High-Wealth Individuals Research Project (April 2023): median effective rate of 8.9% for 311 high-wealth individuals on economic income incl. unrealised gains, vs 20.2% for middle NZ. Lower brackets from NZ Treasury distributional analysis. No general capital gains tax. | 2023 |
| 🇳🇴Norway | Alstadsæter, Johannesen & Zucman (2019); SSB tax statistics; WID.world Norway series | 2021 |
| 🇿🇦South Africa | SARS Tax Statistics; Chatterjee, Czajka & Gethin (2022), WID.world; National Treasury | 2021 |
| 🇪🇸Spain | Agencia Tributaria; Alvaredo & Saez (2009, updated); EU Tax Observatory (2024) | 2022 |
| 🇸🇪Sweden | Waldenström (2020), IFN Stockholm; SCB tax statistics; WID.world Sweden series | 2021 |
| 🇨🇭Switzerland | Swiss Federal Tax Administration; Brülhart et al. (2022); OECD Tax Database 2024 | 2022 |
| 🇳🇱The Netherlands | CPB Netherlands Bureau; CBS income statistics; EU Tax Observatory (2024) | 2022 |
| 🇬🇧United Kingdom | Advani, Chamberlain & Summers (2023); HMRC Survey of Personal Incomes; ONS household data | 2022 |
| 🇺🇸United States | Saez & Zucman (2019), The Triumph of Injustice; updated with IRS microdata through 2018 | 2018 |
Effective tax rates include all taxes: income, payroll, corporate (allocated to shareholders), property, estate, and consumption taxes, divided by total pre-tax economic income. Sources combine academic research, government tax statistics, and the EU Tax Observatory Global Tax Evasion Report (2024). Tax rate data is not API-fetchable and is maintained manually from published government statistics and academic papers.
局限性
- 顶端财富低估 — 基于调查的财富数据系统性低估超富裕人群,他们在家庭调查中代表性不足。WID 用税务数据部分修正,但仍存在缺口。
- 自报收入偏差 — 用户自填收入,可能未涵盖全部薪酬(奖金、股权、未实现收益)。
- 国别注意事项 — 各国数据质量不同:一些有详尽税务型财富数据,另一些只能依赖误差更大的调查估计。
- 模型为近似 — 18 因子收入-财富模型是统计近似,不是个人化财务建议。个人情况可能与人口平均值差异显著。
数据时效性
所有数据在构建时打包并以静态形式提供 — 你使用工具时不会有任何外部 API 调用。一个抓取脚本(scripts/fetch-all-data.mjs)从以下来源拉取数据:
- WID.world API — 财富份额、收入份额、平均/中位数财富、基尼系数
- 世界银行 API — 人口(SP.POP.TOTL)
- ECB / Frankfurter API — 用于货币换算的汇率
- Forbes RTB API — 亿万富豪的净资产数据
- OECD / FRED — 工资、CPI、房价指数
税率数据是例外 — 它来自学术论文,由人工维护并附上完整引用(见上方表格)。