[PEN-L:5037] Re: Unemployment Duration

[jtreacy <JTREACY@xxxxxxxxxxxxxxxxx>]

Treacy: Using decenial data may allow you to compare the number of people
	in the cenus that have moved.  For example in the 1950's and
	early 1960's we were still experiencing major moves of workers
	out of agriculture.  Search times may be longer for those that
	have to make longer moves.

	Longer moves to find work would increase lose from job lose.
JTREACY@xxxxxxxxxxxxxxxxx COPYRIGHTED

On Wed, 10 May 1995, Eric Nilsson wrote:

> A few hours ago I posted data on the length of time
> an job loser (fired or laid off permanently) could expect
> to be unemployed.  This is an update of those numbers
> (based on the work I've done the past few hours).
>
> These are the numbers:
>
> decade               weeks job losers unemployed
> 1950s                            20.8 weeks
> 1960s                            21.3
> 1970s                            22.2
> 1980s                            24.8
> 1990-1994                     28.2
>
> That is, in the 1990s someone who is fired or laid off permanently
> can expect to be unemployed about 6 1/2 months before they
> find a job or give up looking. Things have become
> much worse for these people in recent years -- notice the
> big jump between the 1980s and the 1990s. (Some of this
> growth is due to the recent recession).
>
> More on what these numbers mean:
> The data above are an estimate of the number of weeks
> a worker who is fired or layoff permanently can expect
> to be unemployed. The spell of unemployment ends when
> the worker finds another job or gives up looking for work.
> These numbers are based on published BLS data on
> unemployment duration, but are modified to reflect: 1) the
> too narrow definition of "unemployment" used by the BLS,
> and 2) the fact that most workers experience multiple spells of
> "unemployment" (=actively looking for work). This is why
> these numbers differ from those I posted before.
>
> If anything, I think these numbers might be underestimates.
>
> (Note: published BLS numbers for "unemployment duration"
> are not what they seem -- they measure of the average length
> of time people currently unemployed have been unemployed
> (that is, they have not yet completed their spell of unemployment).)
>
> The generation of these numbers are based on many assumptions
> but I think are likely at least as good as other estimates. Of
> course, these numbers will likely undergo some further
> modification before they are finalized.
>
> Eric Nilsson
> Department of Economics
> California State University
> San Bernardino, CA 92407
> enilsson@xxxxxxxxxxxxxxx
>