经济数据的修正(revision)不是造假

For example, the nonform payrolls number. At the time it is initially announced, it was based upon so-called Established survey, interviewing about 119,000 business and government agency. This is a sampling process. And it will be adjusted by seasonality and the closing and opening of business.  In short it is an estimate. 

Then BLS will compare the estimate vs. the actually tax number in the later date when it become available. The difference is reported as revision.

During Trump presidency, there were a lot of revision as well. 

It is part of the process. 

 

所有跟帖: 

这个revision有点生猛 -酱油缸- 给 酱油缸 发送悄悄话 (0 bytes) () 11/01/2024 postreply 07:11:12

如果错误总是一个方向,那多半是造假 -枪迷球迷- 给 枪迷球迷 发送悄悄话 枪迷球迷 的博客首页 (51 bytes) () 11/01/2024 postreply 07:13:55

单向出错,数据则是近乎随机,误差可以超过100%. 这是科学。欺负我不读书? -TalkToMi- 给 TalkToMi 发送悄悄话 (0 bytes) () 11/01/2024 postreply 07:25:53

估计是所用的数学模型参数调整太随意 -richard_hz- 给 richard_hz 发送悄悄话 richard_hz 的博客首页 (95 bytes) () 11/01/2024 postreply 07:17:06

以科学的名义造假。 -TalkToMi- 给 TalkToMi 发送悄悄话 (0 bytes) () 11/01/2024 postreply 07:22:28

Tere were upward revisions, and there were downward revision -未知就是风险- 给 未知就是风险 发送悄悄话 (159 bytes) () 11/01/2024 postreply 07:22:38

请您先登陆,再发跟帖!