Group 1: Employment Data Revision Insights - The July non-farm payroll report revised the employment data for May and June down by 258,000, raising concerns about the accuracy of U.S. employment data[4] - Since 1979, the probability of downward revisions in non-farm payrolls has increased during economic slowdowns, with a current downward revision probability of 72.4% since 2023[10] - The average downward revision magnitude has widened to -12.8% in 2023, compared to a historical average of -8.4%[10] Group 2: Factors Contributing to Data Volatility - Four main factors have contributed to increased volatility in U.S. employment data post-pandemic: decreased response rates in non-farm surveys, increased immigration complicating data collection, seasonal disruptions, and larger errors in the Birth-Death Model[6] - The response rate for non-farm surveys has dropped from 59% pre-pandemic to 42.9% in May 2025, indicating a significant reduction in sample size[19] - The Birth-Death Model has led to an estimated overestimation of employment levels by approximately 440,000 jobs due to structural changes in new business formations[21] Group 3: Upcoming Employment Data Revisions - Attention is drawn to the upcoming benchmark revision of non-farm employment data scheduled for September 9, 2025, which may lead to further downward adjustments[7] - The QCEW data, which covers over 95% of employment, is expected to provide a more accurate reflection of employment levels compared to the non-farm payrolls, which cover only about one-third[44] - Historical trends suggest that the upcoming benchmark revision may result in a downward adjustment similar to the previous year's initial revision of 818,000 jobs[52]

非农就业数据历来是美联储观察经济冷热的核心指标之一。它直接影响利率政策预期、美元走势乃至股 债市场情绪。 但这个数据并非一锤定音，往往存在"事后回头修"的现象。每月非农报告发布后，通常会在随后两个月 进行初步修正，而每年还会在次年年初发布所谓"基准修订"（BenchmarkRevision），通过对涵盖更多样 本与更全面税务数据的分析，对过去一年甚至更久的数据进行再校准。 在市场密切关注美联储政策走向之际，就业数据这一关键指标正在释放不同于表面的信号。高盛最新研 报指出，美国劳工部对5月与6月的非农就业新增人数合计下调了25.8万，创下1968年以来最大规模的两 个月修正纪录。 不仅如此，高盛还预测，今年9月即将发布的年度非农就业基准修订初步估算将进一步大幅下调55万至 95万人，这一数字远超市场预期，可能重塑市场对当前劳动力市场"强劲"的认知。 非农数据的"修正门"：为何重要？ Vatee万腾据悉，高盛预计9月发布的修订将对去年全年数据下修55万至95万人之间，若兑现，将成为近 十年来最大规模的基准调整。此前2023年1月的修订下调仅5万人，影响相对温和，而此次可能多达十 倍。 其他同步指标已现疲态 与非农 ...

金十回顾：美国劳工部将修正4月非农数据 美国劳工统计局（BLS）在本周二宣布，将于周五发布5月非农就业数据时修正4月数据中的"轻微误 差"，并强调此次修正不会影响失业率等关键指标。它在声明中称："由于引入重新设计的当前人口调查 （Current Population Survey，CPS）样本时涉及权重的轻微误差，2025年4月的部分数据估计值将于2025 年6月6日修正。不过，劳动力市场主要指标，如失业率、劳动参与率和就业人口比率均未受影响。"非 农数据即将公布，美国劳工部一般修正前两个月的非农数据，修正幅度大小也会引发市场关注。 ...