as an analyst, you are taught that data never lies. however, you eventually learn that data can be used to demonstrate almost ANY point you want to make. it’s not creative analysis but rather, slicing and dicing based on what you want to tell.

this is what the government is doing with the unemployment number. the number is WAY less than it was in the 1930s (although at the time of the crash, it was about 8% or so, which is near where it is today – it didn’t reach 25% until a few years later). however, a big reason for this is that the labor market is shrinking.

the calculation is unemployed / labor force, and unemployed is a person available to work and currently seeking work. that’s the key statement. if someone stops looking, they’re not considered unemployed. as people get discouraged, they’ve stopped seeking work, which ends up lowering our unemployment numbers.

see this post by barry ritholtz (his blog is the big picture, and i highly recommend it). Ritholz explains this phenomenon in some great graphs. in addition, Merrill Lynch says the real unemployment rate is around 14%.

how do you use numbers? are you providing all the scenarios or just the one you want people to know?