The establishment survey suffers from over-hype. In a given quarter, more than 7.5 million jobs are created and roughly that number lost, with the difference -- 179,000 in the fourth quarter, 513,000 more in the first quarter -- being the job growth shown in the establishment survey. It is based on a statistical sample seeking to measure small changes in a large number, the 131 million U.S. non-farm labor force. It's like taking the temperature in one city per state, then asking people to guess whether the average temperature for the whole nation is 51.3 degrees or 51.4 degrees.Here we have a statistic which is inherently wildly volitile - the small difference of two large statistical results - but it is all the rage for a day or two. I'm glad it's up; don't get me wrong - but seen in historical perspective that number may mean far less than we hope ( in this case). We just don't know. But, we have a number to talk about!Jobs, Jobs, Jobs
Wall Street Journal | April 6, 2004 | DAVID MALPASS
Why do journalists talk about statistically insignificant results? Because we are curious, and willing to get sucked in to speculating on what that number means for Bush/Kerry. The reason to suspect that Kerry is in trouble on the economy (meaning, we-the-people are not) has more to do with the timing of the last recession than most anything. The "Jobs, Jobs, Jobs" statistic is really just a straw in the wind. But it gives journalists something to talk about when their deadline is approaching - as it always is. The show must go on.