First of all, the US unemployment report is largely flawed in many respects. Number one, it is actually a survey of employers, not actual hires the past month. Secondly, the unemployment rate is a single number and doesn't take into account the split between full-time and part-time workers (full-time usually make more $). Also, it doesn't take into account the Labour Force Participation Rate, which right now, is near an all-time low (not a good sign).
All it's failings aside, the unemployment rate can be a good predictor of a recession.
Scott shows this chart most months when he blogs about the US unemployment data (which he will likely do tomorrow). I am only showing it today as an add-on to my blog yesterday:
- Bars - the last three US recessionary periods
- White Line - the actual US unemployment rate (4.4%)
- Orange Line - the 36 month moving average of the US unemployment rate (5.1%)
Notice that every time the US unemployment rate (white line) crosses up through the downward-moving 36 month moving average, a recession begins right at that time.
As you can see in the bottom-right in the white box, today, we are still a little ways out from a possible recession. For this historical metric to work again, we need to start seeing the unemployment rate bottoming and starting to tick up while the 36 month moving average will keep going down. When they cross...beware of history.
Tomorrow 8:30am will give us another data point to watch.