如何用真实统计数据构建 RE 指标
Constructing RE Indicators From Real-World Statistical Data
本站数据库中的数值为示意性估计。若研究者希望构建可发表、可复制的实证数据集，可参考下表将四个可观察维度与既有的权威国际数据源建立对应关系。本页不提供已下载的原始数据，仅提供方法论层面的映射指引与检索路径。
The values in this site's database are illustrative estimates. Researchers wishing to build a publishable, replicable empirical dataset may use the table below to map the four observable dimensions onto established authoritative international data sources. This page does not host downloaded raw data — it provides methodological mapping guidance and retrieval pathways only.
四个可观察维度对应的候选权威数据源
Candidate Authoritative Data Sources for the Four Observable Dimensions
| 维度 | Dimension | 候选数据源 | Candidate Source | 相关指标示例 | Example Related Indicators | 发布机构 | Publishing Body |
|---|---|---|---|---|---|---|---|
| 财政响应能力 | Fiscal Responsiveness | IMF《政府财政统计年鉴》（GFS）；OECD《政府概览》（Government at a Glance） | IMF Government Finance Statistics (GFS) Yearbook; OECD Government at a Glance | 追加预算占比、财政支出调整弹性、应急拨款响应周期 | Supplementary budget share, fiscal expenditure adjustment elasticity, emergency appropriation response time | IMF / OECD | |
| 技术转换速度 | Technological Conversion Speed | 世界知识产权组织《全球创新指数》（GII）；世界银行《研发支出数据》 | WIPO Global Innovation Index (GII); World Bank R&D Expenditure Data | 技术采用速度分项得分、研发资本形成、产业结构转型速度 | Technology adoption sub-scores, R&D capital formation, speed of industrial structural transition | WIPO / World Bank | |
| 监管适应性 | Regulatory Adaptability | 世界银行《营商环境报告》历史数据；OECD《监管政策展望》 | World Bank Doing Business (historical archive); OECD Regulatory Policy Outlook | 监管质量指数、规则修订周期、监管影响评估（RIA）覆盖率 | Regulatory Quality Index, rule-revision cycle length, Regulatory Impact Assessment (RIA) coverage | World Bank / OECD | |
| 官僚摩擦 | Bureaucratic Friction | 世界银行《世界治理指标》（WGI）；全球竞争力报告（GCI）制度分项 | World Bank Worldwide Governance Indicators (WGI); Global Competitiveness Index (GCI) institutional pillar | 政府效能指数、腐败控制指数、行政审批平均耗时 | Government Effectiveness Index, Control of Corruption Index, average administrative approval time | World Bank / WEF |
从原始数据到 RE 指数的五步流程
A Five-Step Workflow From Raw Data to the RE Index
① 数据采集
① Data Collection
从上表候选数据源下载覆盖目标国家与年份的原始指标面板数据。
Download raw panel data covering target countries and years from the candidate sources above.
② 标准化
② Standardization
对量纲不一致的原始指标进行 Min-Max 或 Z-score 标准化，统一映射至 0–10 标度。
Apply Min-Max or Z-score standardization to indicators with inconsistent units, mapping them onto a common 0–10 scale.
③ 权重确定
③ Weight Elicitation
通过主成分分析（PCA）、专家德尔菲法或结构方程模型确定各维度到 V/C/F 的聚合权重。
Use Principal Component Analysis (PCA), expert Delphi methods, or Structural Equation Modeling (SEM) to determine dimension-to-V/C/F aggregation weights.
④ 指数合成
④ Index Synthesis
代入 RE = VC/(1+F) 计算复合指数，并进行跨国、跨年份的稳健性检验。
Substitute into RE = VC/(1+F) to compute the composite index, then conduct cross-country and cross-year robustness checks.
⑤ 效度与信度检验
⑤ Validity & Reliability Testing
通过 Cronbach's α 检验内部一致性，通过与既有相关构念（如国家能力指数、政府效能指数）的相关分析检验构念效度，并使用 6 节所述策略排除反向因果与遗漏变量偏差。
Test internal consistency via Cronbach's α, assess construct validity through correlation analysis with related established constructs (e.g., State Capacity Index, Government Effectiveness Index), and apply the strategies described in Section 6 to rule out reverse causality and omitted variable bias.