THE BLS KEEPS SIX DIFFERENT measures of unemployment, of which the official unemployment rate, U-3, is only one. While the official measure covers just those who have looked for work over the previous four weeks, U-6 builds on that category and adds two others: the “marginally attached,” jobless folks who have not looked over the previous four weeks, but have looked at some point over the past year; and the “involuntary part-time,” part-timers who are seeking full-time work. In November, with U-3 at 7%, U-6 ran 13.2%.
Now, I would not join the chorus of critics who argue that 13.2% is somehow the more “accurate” figure. Let’s cast ourselves back to the year 2000, when U-3 was at 4%. No one doubted that, at that rate of joblessness, the full-employment economy had been achieved and that, in fact, many jobs were going begging (for which the obvious solution would have been to offer a higher wage). But naysayers could have pointed out that, through the year 2000, U-6 was at 7%, or 75% higher.
Some say that wherever U-3 is, U-6 typically should be no more than three percentage points higher. I take the far more liberal view that the U-6 figure should be about three-quarters higher. Yet even by that standard, today’s U-6 number still looks too high.
From 1994 (the first year the BLS began keeping these data), through 2012, U-6 has averaged 76% higher than U-3, or virtually the same ratio that prevailed in 2000. On that basis, U-6 is about a full percentage point higher than it should be. When U-3 was 7.3% in October, U-6 should have been 12.8% (7.3 times 1.76); instead, it was 13.8%. At an official unemployment rate of 7% in November, U-6 should have been 12.3%. Instead, it was 13.2%.
Put another way, at a U-6 of 13.2%, U-3 should be 7.5% because 1.76 times 7.5 equals 13.2. That’s why, as a measure of unemployment broadly defined, the official unemployment rate is currently misleading.