Several predominant trends can be observed across the eight annual cohorts:
- Exits from foster care in general, and exits by reunification in particular, occur with higher incidence during the earlier months of spells in foster care. This is most readily observed in Panels E and F, which show that monthly exit (reunification) rates during the first month in care tend to be almost double the rates for months 2 and 3, which in turn are larger than the rates for months 3-6, etc. For example, of the 1990 entrants to care, 9.3 percent were reunified during the first month. Of those that remained in care after one month, an average of 4.8 percent per month were reunified during the next two months. Of those that remained in care at the end of 3 months, an average of 2.8 percent per month were reunified during the next three months. The likelihood of exit of any type, and reunification in particular, decreases as the elapsed time in a foster care spell increases. Another way of saying this is that reunifications are most likely to occur early in spells. The longer a child remains in care, the less likely it is that he/she will be reunified. This fundamental pattern is repeated for each entry cohort.
- Foster care exit rates dropped between 1990 and 1996. With a few small exceptions from the 1995 cohort, every column in Panels C and E decreases as we move downward from the 1990 entrants to the 1996 entrants. Because exits are occurring progressively later in spells for each entry cohort, the average duration of spells in care was increasing across this time period. Panels D and E show that reunification rates also dropped steadily between 1990 and 1996, with a minor exception in the 1993 cohort. Thus, from 1990 to1996, a period in which the number of entrants to care was increasing, the prevailing trend in exit rates and reunification rates was downward.
- On the whole, the decreases in reunification were sharper than the decreases in exits. Panel G presents reunification as a proportion of all exits. Except for a small reversal between the 1992 and 1993 entry cohorts, the share of exits by reunification decreased steadily from 1990 to 1996.
- These decreases in exits and reunifications over time were most pronounced among shorter spells, those less than six months in duration. Looking at Panels E and F, we see that the differences between the rates for the 1990 and 1994 cohorts are much larger for the 0-1, 2-3 and 3-6 month duration intervals than for the later duration intervals.