Human Capital-weighted population estimates for 185 countries from 1970 to 2100

We provide a novel dataset of human capital-weighted population size (HCWP) for 185 countries from 1970 to 2100. HCWP summarizes a population's productive capacity and human capital heterogeneity in a single metric, enabling comparisons across countries and over time. The weights are derived from Mincerian earnings functions applied to multi-country census data on educational attainment. The model used to compute the returns to schooling accounts for the diminishing positive relative relationship between education and wages as the overall education of populations rises. The population weights are adjusted by a skills assessment factor representing differences in education quality across countries and years. HCWP is calculated by applying these adjusted human capital weights to population estimates and projections disaggregated by age, sex and education, spanning the period 1970-2020 and 2020-2100 for five Shared Socioeconomic Pathway scenarios. Validation analyses demonstrate the utility of the new HCWP data in explaining national income trends. As a more comprehensive population measure than basic size and age-sex indicators, HCWP enhances the power of statistical models aimed at the assessment of socioeconomic change impacts and forecasting.本研究构建了一套覆盖185个国家、时间跨度为1970年至2100年的人力资本加权人口规模（human capital-weighted population size, HCWP）全新数据集。HCWP可在单一指标中整合人口的生产能力与人力资本异质性特征，支持跨国及跨时间维度的比较分析。该权重基于应用于多国教育成就普查数据的明瑟收入函数（Mincerian earnings function）推导而来，用于计算教育回报率的模型可考量随着人口整体受教育水平提升，教育与工资间正向相对关系的边际递减效应。人口权重将通过技能评估因子进行调整，该因子用以表征不同国家及年份间的教育质量差异。HCWP通过将上述调整后的人力资本权重应用于按年龄、性别及受教育水平细分的人口估算与预测数据计算得到，该数据覆盖1970年至2020年的历史人口估算，以及五种共享社会经济路径（Shared Socioeconomic Pathway）情景下2020年至2100年的人口预测。验证分析结果表明，这套全新的HCWP数据集可有效解释各国国民收入的变化趋势。相较于基础人口规模与年龄-性别结构指标，HCWP是更为全面的人口测度方式，可提升用于评估社会经济变化影响及开展预测的统计模型的拟合效能。