Comment by some_guy_nobel
2 days ago
You can show any number in isolation and it can mean anything.
Now try presenting it the distribution of typical job gains/losses!
2 days ago
You can show any number in isolation and it can mean anything.
Now try presenting it the distribution of typical job gains/losses!
The reason the percentage was included was to take the number out of isolation. What can mean anything is when people tell you to ignore the numbers, because they could mean anything.
> Now try presenting it the distribution of typical job gains/losses!
You first!
> You first!
In any honest conversation, the person presenting the numbers will provide this, but I'll bite:
How many years are cumulative negative in this table?
https://data.bls.gov/timeseries/ces0000000001?output_view=ne...
(I guess the correct way to model this would be as the difference of Poissons between hires and separations, but a student-t should do just fine since we don't have separations)