美国非农就业数据下修，是统计困局还是政治阴谋？

Group 1 - The core viewpoint of the article revolves around the unexpected adjustments in U.S. non-farm employment data, which initially suggested economic growth but were later revised down significantly, raising concerns about the reliability of economic indicators [3][5][10] - The initial non-farm employment numbers for May and June were revised down from 144,000 to 19,000 and from 147,000 to 14,000 respectively, indicating a downward adjustment of 258,000 jobs over two months [5][16] - Economists expressed relief that economic logic remained intact, but the significant revisions led to doubts about the reliability of data used for decision-making [5][16] Group 2 - The U.S. Bureau of Labor Statistics (BLS) collects employment data from approximately 121,000 businesses and government agencies, which is subject to statistical errors due to response rates, seasonal adjustments, and estimation models [6][8] - The BLS employs a "Current Employment Statistics" (CES) survey and a "Net Birth-Death" (NBD) model to estimate employment changes, which can lead to discrepancies in initial data [6][8] - The adjustments in employment data are considered a normal statistical practice, and significant downward revisions have occurred in the past during economic crises, such as the 2008 financial crisis and the COVID-19 pandemic [11][14][15] Group 3 - Following the release of the revised employment data, there was a notable market reaction, with the Nasdaq experiencing a single-day drop of 2.24%, the largest since May [17] - The Federal Reserve's decision-making regarding interest rates is complicated by the mixed signals from employment data and inflation rates, with experts suggesting that other economic indicators should also be considered [18][22] - Despite concerns about a potential recession, the article suggests that the impact of artificial intelligence on productivity and demand may mitigate severe economic downturns [23]