The Alameda project: A two year report & one year follow up

The Alameda project: A two year report & one year follow up

Child Abuse and Neglect, Vol.3, pp. 521 ,-528. @ Pergamon PressLtd., 1979. Printed inGreatBritain. Theodore J. Stein School of Social Work Calif. Sta...

730KB Sizes 0 Downloads 33 Views

Child Abuse and Neglect, Vol.3, pp. 521 ,-528. @ Pergamon PressLtd., 1979. Printed inGreatBritain.

Theodore J. Stein School of Social Work Calif. State Univ.-Sac.CA.

01.45.2134/79/0601-0521 $02.00/O.

Eileen D. (iambrill School of Social Work Univ. of Calif. Berk.-Perk.CA.

Data wggeats that more than .?Q% of the estimated420,ooO children in foster care in the United States will spend their lives in out of home placement. ('I)Rather than being the outcome of psqx~+efuldecisionmaking, the situationin which these children live is said to result from an absence of case planning. (2) These childran are describedin the literatureas "drifting" in the limbo of long term foster care, (3) Concern for the plight of such yaungstersled to the developmentand implementationof a two year experimentalproject which ia describedin this paper. Providing permanenthomes for chil&en in care was the primary objective.Intensive serviceswere offered to natural parents ta facilitatetheir participationin decisionmaking such that permanentplans could be made. One of the followingfour outcomes could result from such planning: (a) the child could be re&ored to natural parents; (b) the child could be placed for adoption followingparental relinquishmentor terminationof parsntaf right; fc) an adult could be designatedby the court as the child's legal guardian,or fdf the child could remain in long term placement, Contparing the effectivenessof behavioralprocedureswith those typicallyused by child welfare workers to resolve identifiedproblems w&s a second objective,Testing the feasibility of having two workers &are the managementof each case was a third area for investigation. One of the workers bore responsibilityfor providing services to natural parents; the other focused attention on the child and foster parents.This article describesthe experimentalphase of the project and presents data concerningthe whereaboutsof the children one year after the project terminated.(4) METHOD Cases had to meet the following criteriato be included: ta) at least one biologicalparent had to be present in the county; (b) the child must be under 16 years of age; fc) the child muat be in foster home placement. Children in treatment institutionswere excluded since other resourceswere primary treatmentagents and because geographicdistance made frequentcontact between parents and children difficult,and; cd) a decisionas to the child*s future mu& not have been made. Such decision makkng was an objective of the project. The initial caseload was volunteeredfrom the active files of workers in two out of six foster care placementunits selected for inclusion in the project. One unit waa designatedaa experimental,the other as CC!lt~Cl. New intake was randomlyassigned to all units. Accepting new cases was contingent on openings in the caseloadsof workers. Caseload limits were reached,inthe experimentalunit, by the end of the first month at which point a second experimentalunit was added in order to continueaccepting new cases. There were two experimentalunits and one control for the remainder of the project. Three project workers were hired especiallyfor the study. A total of 428 childrenparticipatedin the project over two year&, 227 in experimentalunits and 201 in control. Volunteeredcaees accountedfor 54% in the experimentalunits and 66% in the control,whereas new intake comprised46% of the cases in the experimentalunit and 34% in the control.Maximum caseload size in the experimentaland control units was 35 childrenin the former and 49 in the latter. In viewing caseload size it is importantto remember that all experimentalcaseB were in various stages of decision making and thus were active, Once a case was referred for adoption or guarctkanehip or designatedlong term care, the case was closed to

522

T.J. Stein and E.D. Gambrill

the project to make room for new intake. In control, all long term cases as well as guardianship on which the court dependencywas not dismissed,remained part of the worker's caseload. It is reasonable to conclude that such cases were less active than those in various stages of decision making. Sample Description Ninety-ninepercent of all childrenwere court dependents,voluntary placementsrepresenting only 1 percent in each unit. Eighty-sixpercent of the children in the experimentalunit and 92 percent of the children in the control unit entered care under neglect petitions. Thirteen percent in the former and six percent in the latter entered care under either abuse or combination abuse and neglect petitions. The majority of children in the experimentalgroup were six years old or less (58%) compared to 31% for children in the control units, whereas control cases had a slightly higher percentageof children 7 to 9 years of age (31% compared to 22%).* There were equal percentagesof 10 to 12 year olds in both units. More control than experimental childrenwere over 12 years of age (22% compared to 6%). Males and females were almost equally representedacross units, as were Caucasian (34% experimental;j'&control),MexicanAmerican (9% and 8%) and Bi-Racial children (9% and 10%). Black children representeda greater percentageof experimentalthan control cases (48% compared to 3%), while 8% of the control and none of the experimentalchildrenwere of American-Indiandescent. Females headed the majority of families (71% experimental;74% control). Two-parentfamiliesrepresented23% of the cases in each unit. Male headed familiesaccounted for only 6% and 4% of the experimenal and control cases. The only significantdifferencebetween units was in the child's age (*= 42.95;5 d of f;p.&CCl). Reason for placement. For 54 of the 227 experimentalunit children (24%) and 47 of the 201 children in the control group (23%) the general definitionof the petition under which they entered care (e.g. neglected or abused) was the only informationavailableregarding reason for placement. Additionalproblem informationwas obtained from case records for the remaining 173 experimental(76%) and 145 control yo gsters (7%). Significantdifferencesbetween units existed in 3 out of 10 problem areas (x30.6 (9 d of f) p.-< .OOl):parentaldrug abuse, cited as the reason for placing 53 experimentalcases (21%) compared to 20 in control (100~); alleged physical abuse of the child which was reported for 28 children in the experimentalunit (11%) and 9 in the control (5%), and being left unattendedwith a neighbor or relative reported for 56 children in control (29%) compared to 36 in the experimentalunits (14%). Other problem areas mentioned included parental use of alcohol, parent in prison, in a psychiatricinstitution or said to be mentally ill but not hospitalized,children "out of parental control, and the condition of the parent's home.n There was a miscellaneouscategory that included small numbers of youngsterswhose parents had been hospitalizedfor physical illness, childrenwhose medical needs were unmet, and childrenwhose parents had marital difficulties. Framework for DecisionMeking Ninety-fivepercent of natural parents in the experimentalunits stated that they wished to have their children restored to their care. There was no attempt at this point to judge the sincerityof their verbal statement. Their motivationand capacity to parent was assessed through the efforts they extended in working toward that goal. Counselor-clientcontracts were drawn up in all cases and the parents encouragedto sign these. Initial contract6"setthe-stage"for gathering needed assessmentinformation,during prearrangedparent-childvisits, and they specified outcomes in other areas that could be determinedat en early point. If additional problems were identifiedas ca6es moved toward restoration,new objectiveswere added. Minimum standards of parentingwere our concern; hence, contingenciesfor restorationfocused only on the relationshipbetween parental behavior and a child's well being. Detailed plans for achieving each objectivewere describedin writing and attached to the contract. An important part of case planning entails decidingwhen to "alter coursen and pursue new options. Contracts facilitatedthis process, since they included statementsdescribingalternatives for each case, such as referringa case for adoption, that a worker could pursue if clients did not participatein plans, as well as the time limits within which this would occur. Because of the specificitywith which they describedobjective6and documentedprogress toward

*Further descriptiveinf,ormation is available from the author6 upon request.

The Alameda Project

523

their attainment,they could facilitatethe transfer of a case to a new worker. (5) RESULTS When the project ended in April of 196, decisionshad been made for 145 of the 227 children in the experimentalunits (64%) and for 148 of the 201 children in the control unit (74%). (see Table 1). Premature closings precludedmaking decieione,or, if they were made, precluded followingthem through to completionfor the remaining82 experimentaland 53 control unit children. Premature closings included case6 where the court jurisdictionchanged when a parent moved out of the county, caseB where children"ran away*Ifrom foster homes and whose status under the law changed from that of a neglectedchild to a child "in danger of becoming a delinquent." Cases were closed followingrestorationto natural parent6 and dismissalof the court dependency or after adoption or guardianshipactions were completedfor 41% of the experimentalunit children compared to 25% of those in the control unit (see Table I). Cases were to be closed in the near future for 37% of the children in the former units compared to 1% in the latter. These would be closed once dependencieswere dismissedfor childrenwho were restored, or when cases were completed for children to be restored,adopted or who were to have court appointed guardians. Overall, 114 experimentalunit children (79%) compared to 59 in control (4C%) were either out or headed out of c&a-e when the project ended. Only 31 children from the former unit (21%) compared to 89 from the latter (60%) were designatedlong term placement. The differencein outcome6 between unite was significant(p.=~.CCl). gutcome and PredictorVariables The question addressedhere is what variablesmight be useful to predict differentialoutcomes for children in care? Correlationsfor demographicvariables such as family composition,child's age, ethnicity,6ex, type of placement,reason for placementor petition under which a child warnplaced, were low, ranging from zero (age) to .38 (ethnicity) .* Since age was the only demographic variable for which there wae a significantdifferencebetween groups thie was further analyzed controllingfor length in care. If age had predictiveutility, one would expect older children,in care for long periods of time, to comprise the majority of long term cases. The redte of this analyeis yielded a purely random distributionby age and outcome, controlling for time in care. In general, age had no predictiveutility. Children in the experimentalunits had a significantlygreater chance of leaving foster placement than thoee in the control unit regardlessof the length of time they had been in care (lees then 3 years; p.=s.OOl; three yeare or more; p .=s.Ol).There was a pronouncedchange,however at the three year point in time. For experimentalcases, the percentagerestored or to be restored decreased from 8% to 55% and for control from 44% to 26 percent.There were concommitant increases in the percentageof children designatedlong term. For experimentalthe incream wa8 from 18 to 45%; for control from 56 to 74 percent.? Written contractewere employed by experimentalunit workera only. Seventy percent of the childrenwhose parents signed a written agreementwere reetored. Eighty-fivepercent of those whom parent6 would not sign a contractwere either referred for adoption, court appointment of a guardian or were designatedaa long term care. The predictabilityof restorationcompared to alternativeoutcomes given informationas to whether a contractwae signed wae 57 percent. ONEYEARFOI&OWUP As noted in the introduction,one objectiveof the project was to test the feasibilityof joint case managementbetween a project and county worker. Project workers had been hired expressly for the experimentand their employmentterminatedwhen the project ended. These workers received training in behavioralmethods of interventionand bore the reeponsibilityfor providing intensive cervices to natural parents. A county worker, who was to provide services to the child in the foster home, did not receive special training. In practice, the divisionof reeponaibilitywae not maintained. As cases moved closer to restoration,project staff worked

+ Cuttmanlslambda was the measure of associationused throughoutunless otherwisenoted. data tables are available from the authors on request.

t Complete

524

T.J. Stein anciE.D, Gambrill

with both the natural parent and child. In all likelihood,collaboratingon caees resulted in some informal learning for county personnel, However, several of these workers left the department or were transferredout of foster care units during the follow up year. The result of this change in staffing is that many of the cases in the follow up period were not carried by county workers who had any gart in initial decision meRing. Due to these overall changes in staff composition,follow up data is not presented seperatelyfor experimentaland control units. Reporting by units would suggest that we were comparingcaseworkmethods used. Such a view would be appropriateonly if staff compositionin the units was the same as in prior years. Since this wae no longer the case, it would be misleadingto report data in such a manner. Reference is made to the Unit8 children had been in only as appropriate. Three question6 are addressed below: (I) Did the childrenwho left care during the project period remain out of placement? (2) Were plan6 that were in process when the project ended completed? (3) What was the status of children classifiedae long term placement at the end of the project? The Sample At the end of the project, decisionshad been made for 145 experimentaland 148 control children. Seventy experimental(48%) and 44 control unit children (3C%) had moved out of foster A total of 179 children remained in placement. (see Table 1). Of these, 44 in the experCal-B. imental (3C%) and 15 in control (10%) were headed out of placement. Thirty-one in the experimental unit (21%) and 89 in the control (6~%) were designatedlong term care. Cases were closed during the follow up year for 7* of these 179 children for miscellaneousreasons, such as change in court jurisdiction,leaving a total of 172 youngster6 for whom follow up data is reported. RESULTS Two children who had been restored to their natural parents, one each from experimentaland control units, reentered care during the follow up year. New decisionshad not been reached for these youngstersat the time of follow up. The remaining 172 children fall into two groups based on decisions made during the project. Fifty-nine children (3&f were headed out of placement end 113 were designatedlong term care (66%). Case movement during the follow up period for the 59 children is &own in Table 2. Decisionswere completedin 64% of the cases, changed in 29% and remained unchanged in 10 percent. Cases that had been referred for adoption accounted for the highest percentage of completed decisions (n=lq:7%). Thirteen of the 22 youngsters headed for restoration (5%) were restored and guardianshipactions were completed for 6 of thirteen children (46%). Initial.decisionswere changed for 15 children.Five who were headed for adoption were referred for guardianship,end one child who was a guardianshipreferral was in the adoption unit. The 6 remaining youngstersfor whom guardianshipwas being pursued, plus two who were to have been restored,were classifieda8 long term placements. The one additional child who was headed for restorationwas reported a8 und tided. The differencebetween completed, unchangedand changed decisionswas not significant. (2=4.43 (2 d of f). t case outcomes for the 713 children classifiedas long term placementsare shown in Table 3. For 54% of the children, there was no change in decision atatua. However, changes did occur for 46%. Sixteen percent of the children from this latter group were restored to their natural parents and 28% are now headed out of placement. No new decisionswere reached for two youngsters. DISCUSSION Ae noted in the introduction,more than 50% of the children in foster home care in the United States will grow to maturity in out-of-homeplacement. The results of the Alameda Project suggest that this figure can be substantiallyreduced. Although the design of the project precludes partiallingout the effects of the various case managementprocedures,the method6 employed were effective for attaining project goals. As such, they provide direction for case

*Six of these children had been in experimental,one in control. All had been long term care. t Rtantg-fiveof the 113 ohikban in the long term categesywere from experimentalunits, the rem&.ning 88 mere in.omtrof.

The Alameda Project

525

planning, decision making and problem solving that have heretoforabeen unavailableto child velfare workers. Only one child from the experimentaland one from the control unit reentered care in the follow up gear. This suggests that decisionsto restore children,whether made by experimentalor control unit workers were appropriate. However, as the data in the first part of this report shows, deaisionsthat result in providingchildrenwith continuity-incare were made for a significantlygreater number of childrenin the former units than in the latter. One factor which is probably highly related to this result is the differentuse of time by project and county workers. The majority of experimentalcase contactswere with natural parents while the majority of control case contactswere with foster parents. (6) Based on this data we suggested that caseload siee, often cited as a barrier to service delivery,may not be an obstacle to providing services. Workers do provide services, exercisingjudgementwith regard to the client group who will receive the ulionBs ahs.re.l' It seems reasonablethat only if time is devoted to determiningwhether natural parents can attain the minimal parenting skills required to regain custody of their children,can decisionsbe made in a timely manner for children in care. Why foster parents rather than natural parents? First, it has been suggestedthat workers avoid biologicalparents because of negative attitudes toward this client group resulting from the stigma associatedwith the labels assigned to clients, such as neglectful or abusive. (7) The possibilitythat workers lack problem solving skills has also been suggested (g), and is supportedby our data which shoved that a significantlygreater number of natural parent difficultieswere resolved by experimentalcompared to control workers. (9) Fifty-fourpercent of the youngstersclassifiedas long term care at the end of the project remained classifiedin this manner one year later. If long term care is to provide continuity this decision should remain stable over time. Qiven that decisions changed for almost !5@6of these children, it is reasonableto conclude that this alternativeis least likely to result in stability for children. The lack of stabilityin this outcome categoryrelates to several factors. First, unlike adoption or guardianship,there are no legal safeguardsfor the foster parent-childrelationship. Such cases are subject to yearly court review at which time a child may be returned to natural parents. With few exceptions,when foster parents move out of the county of jurisdiction,the child is replaced in a new home. The absence of a shared definition regarding a minimum period ot time a child is expected to remain in care when designatedlong term is an additionaldifficulty. The need for guidelinesfor use of this category is recognised in the Foster Care and AdoptionsReform Act of 1977 (IO) which describescriteria for its use. Because of this difficulty,it has been recormnended that long term care be dichotomieed to reflect short term placementsas well as those expected to Last an indefiniteperiod of time. (11) Long term decisionsmay be made for the purpose of %anking**cases. (12) This occurs when workers designate childrenas long term shortly after they enter care without exploring alternativeoutcomes. Our data shoved that this approaoh to case managementwas used by control staff. (13) Other factors which may account for such categorisationof cases include lack of a frameworkfor decisionmaking and deficits in problem solving skills. Once classificationoccurs, expectationsmay be establishedwhich limit considerationof alternatives. The worker responsiblefor the initial decisionmay fully intend to reconsiderother options, but the chances of this happening are reduced by the high rate of worker turnover. (14) In view of the relationshipbetween length of time a child is in care and case outcome, decision making of this type is quite problematic. To date, there have been few controlledinvestigationsdirectedat identifyingmethods and procedure6to correct the problems of children drifting in foster care. The mall number of studies employing experimentaldesigns report either minimal or no success. (15) In view of the magnitude of the problem this study addressed,it is imperativethat this work be replicated and the componentsmost relevant to outcome be identified.

The results of a two year experimentalproject plus a one year follov up have been reported. Eata at the end of the study shoved that a significantlygreater number of experimentalthan oontrol unit childrenwere moved out of foster placementwhen systematiccase managementprocedures were used. Only tvo children,one each from experimentaland control units reentered care in the follov up year. This suggests that decisionsto move youngstersout of placement, whether made by experimentalor control staff, were appropriate. Bovever, such decisionsvere

T.J.

526

Stein and E.D. Gambrill

mad8 for only a small percentageof control cases relative to those in experimentalsections. Referencewee mad8 to data previouslyreported by these authors showing that control workers visited with foster parents with approximatelythe same frequencyas experimentalstaff saw biologicalparents. In consideringwhy this occur6, we suggested that perjorativeattitudes caseworkershave toward natural parents plus deficits in problem solving skills may function as barriers to service delivery and case planning. In the follow up year almost 50% of the long term care decisionsmade during the project period were changed. We hypothesizedthat the in&ability of this outcome category wa6 due to the absence of legal safeguardsprotecting the foster parent-childrelationshipas well a8 the absence of uniform criteria for designating children in this way. In addition, attentionwas directed to the uee of long term care of the project, it was recommended for case managementpurpot3es.Because of the overall i8ucces6 that this work be replicatedand that those aspects of the case managementproces.8having the greatest bearing on outcome be identified.

1.

Emlen, A.C. (1976) Barriers to Planning for Children in Foster Care:A Summary Regional Research Institute,Portland;P.6. Gruber, A.R. (1978) Children in Fostei Care: Destitute,Naglectad,.....B8trayed, Human Services Press, New York; P. 17b. Maas,H. & Enz1er.R. (1959) Children in Need of Parents, Columbia Univ. Press, New York; P. 356. Jeta;, H;R. (1963) 2lhildren, Problema and Services in Child Welfare Programs,Children's Bureau, Washington,D.C.; P.E7.

2. Emlen, ibid. Gruber, ibid. Maas and Engler, ibid. Jeter, ibid. 3.

D. Fanshel, The Emit of Children from Foster Care:An Interim Research Report, Child Welfare L, 65-81 (191). Maas and Engler, ibid.

4.

data relative to the second and third objective6can be found in; Stein, T.J., Gambrill, E.D. and Wiltse, K.T. (in-press)Children in Foster Xomea:AchievingContinuityin Care, Praeger Publ., New York.

5.

additionalinformationdescribingthe u68 of contracts in the Alameda Project can be found in; Stein, Gembrill and Wiltse, ibid.

6.

T.J. Stein and E.D. Gambrill,Early Interventionin Foeter Care, Public Welfare 34, 38-44 (1976). M.K. Rosenheim, Notes on Helping Juvenile Nuisances,in Rosenheim,M.K. (ed.) (1976) PurUniv. of Chicago Press, Illinois;43-66. Jenkins, S.and cement:MothersView Foster Care, Columbia Univ. Press, New York; P.140.

7.

8.

E.D. Gambrill and K.T. Wilts%, Foster Care:Plansend Actualitiee,Public Welfare 34, I& 21 (1974).R.H. Mnookin, Child Custody Adjudication:Judicial Function in the Face of Indeterminancy,Law and ContemporaryProblems 39, 226-293 (1975).

9.

Stein, Gambrill and Wiltse, op cit.

10. U.S. Congress,House of Representatives. Foster Care Adoption and Reform Act of 1977, 95th Congress, 1st session (1977). Il. Pascoe, D.J, (1974) Review, Synthesis and Recommendationsof Seven Foster Care Studies in California , The Children*8Research Institute of California,Sacramento;P.6. 12. Pers, J.S. (1976) Governmentas Parent:Adminiatering Foster Care in California,Institute of GovernmentalStudies, Berkeley:P. 117. 13. Stein, Gambrill and Wiltse,

op cit.

14. Pisani, J.R. (1976) The Children of the State:Barriereto the Freeing of Children for Adoption, New York State Dept. of Social Services,Albany; P.130. Shapiro, D. (1976) Agencies and Foster Children,Columbia Univ. Press, New York; p-18. 15. Sherman, E.A., Neuman,R. and Shyne, A.W. (1974) Children Adrift in Foster Care:A Study of AlternativeApproachss,Child Welfare League of America, New York. Jones, M.A., Neuman,R. and Shyne, A.W. (1976) A Second Chance for Families:Evaluationof a Program to Reduce Foster Care, Child Welfare League of America, New York.

527

The Alameda Prcjecr

TABLE ? NUMBER AND PERCENTAGEOF CLOSED CASES AND CBILDREN BEADED OUT OFFOSTER CARE COMPAREDTOTBOSE DESIGNATEDLONGTERM PLACEgENT Experimental 6x=145) Closed2 Dependenciesdiemiesed: childrenrestored to natural parents: Completed: Adoptions Guardianship

ZlE-

-cont.

45

31

12 3

4 2

?G

7

Control {n=148)

&$41X)

37(2Y.Q

54(3m

22w%)

Children headed Out of placement: RestoredrDependency not yet dismissed To

be restored

Adoption to be completed Guardianshipto be completed Sub-Total:Children out or headed out of care: Long-termplacement Total

14

7

'17 12

7 1

,,4(79%) 31C2'%) 145

59Plo%) 89(6p' 148'

"x2=58.73 (9 d of f)p.=kOOl. Given the small numbers in Borneof the cells it waa necessary to combine the followingcategoriesto do a test of significance:allchildren restored,whether or not dependencieswere dismissed,completedadoption and guardianshipcases with those to be completed.

528

T. J.

Stein

and

E. D.

Gambrill

TABLB 2 DECISION STATUS OF CHILDREN HEADED OUT OF PLACEMENT ONE YEAR LATER Decision Status at End of Pro.iect

No. of Children

Status One Year Later Completed Unchanged

To be restored

22

,3(5W

Incompleteadoptions

24

,#7?%)

Incompleteguardianship

13

G&6%)

($2?%1+

Changed 304%)

$21%) 7(54%)

38W+%)

*Percentagesare to horizontal totals kategories of unchanged and changed were combined for chi-squaretest because of the empty cells in the unchanged category.

TABLE 3

DECISION STATUS OF CHILDREN CATEGORIZEDAS LON~TERMPLACE?ENTONEYEARLATEFf Long-Term (n=llJ)

Status Unchanged

61 (54%)

Decisions Changed: were restored

,8(16%)*

to be restored

,2(11%)

referred for adoption referred for guardianship undecided

6 6%) ,4m%) 2 tab) Total

52 (46%)

L

l

Percentagesare to the total number of children