74
between  four  age  categories  (14-17;  18-20;  21-24;  25-27)  and
find  that  instability  rises  monotonically  throughout  the
observed  age  range.
An  additional  important  short-coming  of  our  work  history
data  is  that  they  pertain  to  a  historical  period  (1968-1980)
of  full  employment  in  Israel.  In  periods  of  full  employment
job  shifts  are  more  likely  to  be  initiated  by  employees
rather  than  by  employers.  Thus  our  findings  are  probably
applicable  primarily  to  voluntary  job  instability  as  it
relates  to  age  in  the  early  life  course  of  Israeli  men.  At
the  present-day  unemployment  rates  in  Israel,  job  shifts  are
probably  more  often  initiated  by  employers.  Age  may  have  a
very  different  relationship  to  involuntary  job  terminations
than  our  data  and  analysis  suggest.