From CNBC's Steve Liesman
The biggest news in the September jobs report is not the underwhelming 51,000 jobs created last month, or the upward revision of the August number to 188,000 new jobs from 128,000. The big story is that a previously unknown 810,000 new jobs were actually added to US payrolls for the year through March 2006. This is the biggest revision to payrolls in at least the past decade, and possibly ever.
This was a massive whiff by the jobs counters at the Labor Department. In today's hyper-computerized world of advanced statistics, a 45% upward revision to any data, let alone the most important data series the US government prints, is nothing less than a scandal. From an economics perspective, it's as big a mistake as the CIA missing the WMD call in Iraq.
The reasons for the miss are complicated and even poorly understood within the government. But suffice it to say, it should prompt an urgent rethinking about how we measure job growth. If we're not spending enough on data collection, we need to spend more. If we're not allocating resources correctly, we need to reallocate. Getting the jobs numbers right -- or at least close to correct -- is critical. Markets, Fed policymakers, and individual investors all rely on these numbers to make decisions. It's not way to run the most advanced economy on the face of the earth.