A market alternative to child adoption and foster care.

Blackstone, Erwin A. ; Hakim, Simon

Every year almost one million children in the United States are victims of neglect or abuse. At a given time, about 581,000 children are in foster care and more than 125,000 are available for adoption. About 17 percent remain at least five years in foster care, a status that is intended to be temporary (U.S. Department of Health and Human Services 2001).

The average age of children in foster care is 10 years. Many older children cannot even be placed in foster care; 8 percent are in group homes and 10 percent are in institutions. Unfortunately, a significant number of children who spend large periods of time in foster care, which usually involves many different foster homes, tend to become criminals. A study in Rochester, New York, found that 90 percent of youths who experienced at least five family transitions became delinquent (U.S. Department of Justice 1999). Minorities seem to be over-represented in the foster care population. Black children, for example, comprise 39 percent of the foster care population and 42 percent of those awaiting adoption, while they represent only 12.3 percent of the population.

It is more difficult to find adoptive homes for older or minority children. The waiting time nationally to adopt a healthy infant ranges from one to seven years. It takes four to eighteen months to adopt a typical child from foster care where the average child is 10.3 years old. Adopting a foreign child takes from 6 to 18 months. A study of Michigan adoptions found that black children were adopted at one-third the rate of Caucasians (Barth 1997). Further, even once adopted, older children tend to return at a higher rate to state custody. Five percent of children between three and five years old are returned compared with 17.1 percent for those between 12 and 14 (Evan B. Donaldson 2000).

Some states have tried to improve the process of foster care and adoption by shifting more of the activities to private providers. The intent is to utilize the power of competition and incentives. In this article we review and evaluate the pioneering privatization efforts of Kansas and Michigan. We will then suggest possible market-based improvements to achieve appropriate and expedited placements. The objective is to maximize child welfare while recognizing budgetary limits.

Privatization in Kansas

The American Civil Liberties Union (ACLU) brought a class action suit against Kansas, claiming that its child welfare system had excessively large caseloads and inadequate monitoring of children. As part of the 1993 settlement, Kansas moved toward privatization of its system.

For foster care a bidding process was employed to select a private contractor to serve in each of five regions. In February 1997 three nonprofit contractors from Kansas won the right to be the monopoly provider in each region. Responsibilities included providing homes, medical, and all other required services. The state was broken up into regions in order to maintain the foster care child in close proximity to his or her family because the goal of foster care is family reunification if possible. If reunification is undesirable, adoption is then normally tried. A fixed payment was set for the entire duration of the child in foster care. The 1999 amount was $15,511, and the payment was adjusted annually. This amount had to be estimated because under the state system costs were not maintained by activities. The contractor was required to accept every child but could receive special allotments for difficult cases. At the inception of privatization, the contracts allowed the providers to keep profits not exceeding 10 percent per child. Symetrically, Kansas would reimburse any excess costs above 10 percent. One contractor estimated that if a child remains in foster care longer than six months it loses money.

Under the original contract, costs were on the average 65 percent above the set price. Accordingly, in July 2001, four years after the initiation of privatization, contracts were changed to provide a fixed price per month of $1,958 to $2,200, depending on the region (James Bell Associates 2001).

As a result of the ACLU suit, adoption services were also privatized in October 1996. Lutheran Social Services (LSS) of Kansas and Oklahoma was the only bidder for the statewide contract. LSS and its 12 subcontractors provided adoption services for 1,400 to 1,600 children. Unlike foster care, proximity of the child to his or her original family is not important, and therefore a statewide company was contracted to handle all cases. The nonprofit contractor received $13,556 to provide all adoption services including maintenance of the child. The child was referred to the adoption provider upon the termination of one parent's rights. This early referral contributed to significant losses for the adoption provider because termination of both parents' rights is required for a child to be adopted. While the legal process, which is outside the control of the contractor, worked to free the child for adoption, the provider had to cover all foster care expenses. At the end of two years, LSS had lost $5.5 million and its subcontractors had also incurred substantial losses (Mainstream 2001).

In July 2000, a new statewide contractor was selected, and a year later the fixed price contract was replaced with a per month payment of $2,101. The contractor was still responsible for all costs for children returned to state custody within 18 months of the adoption placement. The change in payment was a result in part of the cost overruns, which were beyond the control of the contractor.

Some positive outcomes are clearly evident. Adoptions increased from 352 in the year before privatization to 546 in the year after, or by 55 percent. Further, in the year before privatization, 43 percent of the children were placed within one year, and the comparable post-privatization figures ranged between 51 and 68 percent. During the four years of privatization only 2 percent of all children were returned to state custody compared with 12 percent for the nation as a whole. Administrative costs of foster care declined from 18 percent of the budget to 8 percent after privatization. Average cost per day of foster care generally declined after privatization (Legislative Division of Post Audit, State of Kansas 2001).

Some shortcomings should be noted. As in many other privatization efforts, it is not clear whether the state saved resources. A legislative study stated that it would be an extremely difficult task to determine the cost of foster care and adoption when it was state run (Legislative Division of Post Audit, State of Kansas 2001).

Implementation of privatization requires the shift from agency-wide budgeting to Activity Based Costing (ABC) as pioneered by former Mayor Steven Goldsmith of Indianapolis. Under ABC direct or avoidable cost is calculated for a governmental activity and should be compared with the prices offered by the contractors. In the absence of such accounting the government agency cannot determine whether any savings are realized when the service is contracted out. Clearly, cost per adoption by the public and private entities can be compared only for similar performance measures. In the case of Kansas we only know that the actual costs were at least $125 million above the projections. There is no evidence as to the cost associated with the improved service and whether it is cost effective.

Kansas did not even have explicit outcome performance measures before privatization to benchmark the private contractors. The state should first assess its performance by shifting to the ABC accounting system and then measuring the level and quality of its service. Based upon the cost and quality information, the state is in a position to decide whether the proposals by the private providers are attractive. Under the pressure of the court suit, Kansas privatized the service without being able to determine whether the private providers' offers were more efficient or effective. In any event, privatization means that Kansas now has better knowledge of the costs of providing adoption and foster care services.

Kansas could have encouraged more competition than it did, leading to lower cost and better performance. Only one company bid for the statewide adoption contact and three companies bid for the right to provide foster care in each of five regions. Clearly, the level of competition was limited, and in both adoption and foster care a private monopoly replaced the public monopoly.

The advantage of the private versus government monopoly is the reduction in the organization's bureaucracy attributed to greater cost pressure. However, the lack of competition prohibits efficient provision and leads to the other familiar ills of monopoly. One way to facilitate competition is to allow the current employees to submit a bid and to compete with private providers. This concept is termed "managed competition" and has been successful in other areas (Blackstone and Hakim 1997).

Actually, it seems that Kansas could enjoy more competition and a private monopoly could be avoided. LSS had 12 subcontractors and the new provider, Kansas Children's Service League, had 6. These subcontractors could probably compete statewide. Hence, there seems to be a sufficient number of firms and children available for adoption to achieve effective competition without selecting a monopoly provider.

In general, incentive contracts are intended to promote efficiency. However, when significant costs are outside the control of the provider, fixed-price contracts are unlikely to be successful. The courts determine the length of time a child remains in foster care or awaits adoption and then largely determine the cost for foster care or adoption. The provider obtained a fixed amount no matter how long the child was in foster care or waited for parental rights to be terminated to become available for adoption. In order to correct for the existence of unknowable costs, the contract was changed to provide a fixed amount per month. This change introduces incentives for the provider to maintain the child in foster care for an extended time. On the other hand, the earlier fixed-price contract for foster care encouraged rapid return of the child to the original home with the hope that she is not returned within 12 months when her care is still the responsibility of the foster care provider. Any contract may cause unintended changes in the behavior of the provider.

Privatization works best when a single easily measurable output exists. For example, in the case of trash collection the output is clearly stated and measured. As such there are many similar transactions in the marketplace. Government can easily contract out trash collection or even allow consumers to select government and nongovernment providers.

Privatization becomes less successful as the number of outcomes increases and their quantification is on different measurement scales. Quality of adoption consists of the length of the process, the return rate, and the attributes of the adoptive family, its home and environment. Clearly, these quality indicators vary in importance, are nonmeasurable on the same scale, and are nonadditive. It is therefore difficult to compare public and the private provision and monitor the private providers.

An important indicator of whether privatization is an option is termed the "yellow pages test." The existence of alternative private adoption services in the region suggests that the private option is viable. The key for success is increased competition and not necessarily shifting services to a private monopoly. The fact that only one company bid for the statewide contract to provide adoption services suggests that the terms of the contract were perceived as unprofitable. Not surprisingly, LSS lost money on its contract.

The government should rebid after clearly specifying expected quality of outcomes with an initial expected price obtained from an ABC-driven cost per adopted or foster care child. A Dutch auction mechanism can be employed where the bid starts with the government cost for a given quality and declines from there. The bidding is terminated when a predecided number of bidders remain. In the long run, creating a market with substantial competitors will improve the quality and price of service. Government workers should be allowed to compete with firms that can perform the service at the government's cost. A merit award system for government social workers may also prompt them to compete with private providers.

Privatization in Michigan

The Michigan legislature in 1989 began the process of encouraging private companies to compete with government in adoption services. The objective was to expedite the return of the child to a permanent family structure by adoption if reunification was not possible. The state established the Michigan Adoption Resource Exchange (MARE) that uses public channels to disseminate information about children available for adoption. The information includes a picture, age, and other demographic characteristics, interests, and the number of siblings also available for adoption (Craig et al. 1998). The state licenses about 80 private adoption agencies to place children.

Prior to 1992 adoption agencies were paid either their actual cost or an average price for placing a child. The larger agencies that were able to accurately track their cost for each child were typically paid $15,000 to $18,000. The smaller agencies that could not track their actual cost were paid only $3,900, a price that would probably make it impossible for them to handle many adoptions. Such a cost-plus system for the larger agencies meant that children were mired in foster care. Therefore, different fixed payments were established depending upon the difficulty and speed of placing the child. For example, the highest payment of $10,000 is given for placing a child directly from residential care like group homes or institutions for delinquent children. The lowest payment of $1,300 goes to a private foster care agency for a child placed by another adoption agency. Placing a child within five months of its availability for adoption yields a payment of $8,600; after seven months the payment declines to $3,535 (Snell 2000).

Under the post 1992 system, the 80 licensed adoption agencies that have contracts with Michigan's welfare department--Family Independence Agency (FIA)--can compete to place any child available for adoption. Every child is under the care of a particular foster care provider. If two or more adoption agencies find families wishing to adopt a particular child, the child's foster care provider determines the best home for the child.

The 1992 incentive program appears to have at best mixed results. Comparing 1999 with 1991 shows that the number of adopted children overall, black children, and disabled children have increased greatly. Adoptions overall increased by 83 percent. However, the number of children available for adoption rose by 116 percent. The adoption of black and disabled children, who are often difficult to place, increased by 82 and 52 percent, respectively, suggesting no obvious improvement. On the positive side Michigan had only 3.5 percent of its adoptions disrupted compared with 12 percent for the United States as a whole (Snell 2000).

The advantages of Michigan's program are the ubiquitous dispersion of information about the children available for adoption and the large number of companies able to compete in placing children. Another advantage is that the FIA can also place children, a movement toward the desired model of managed competition. Also, payments for foster care are not the responsibility of the adoption agencies for the period while the agencies are awaiting adoption approval by the court to place a child.

Economists strongly believe that market-determined prices lead to efficiency. Michigan has started in the right direction in using prices as incentives. However, prices established by Michigan for adoption providers are arbitrary; prices neither reflect the opportunity cost for government nor are they the result of market forces. Prices reflect the length of time for adoption and some proxies for characteristics of the child. Since the state or the court needs to approve the adopting family, it would appear that the length of time is somewhat out of the control of the private provider. Moreover, the payment system is so complicated and difficult to comprehend that it weakens the incentives for providers. Further, the a priori uncertain time of the adoption process is a problem for providers in choosing what cases to take and the effort to devote to them.

The FIA still places children for adoption like the private agencies do. In practice, it appears that the FIA places the more difficult cases. Employing the full concept of managed competition, where workers are rewarded for excellence, can further enhance competition and improve performance.

Michigan still could further improve its process. It is unclear whether the payments for private agencies reflect savings compared with government operation. ABC accounting would enable comparison of public costs and private pricing. We suggest that the ABC accounting system be established to set the initial prices for private agencies. The cost and adoption time by government for the major characteristics of adoptable children will set the initial prices and standards for at least one year. Government employees will be allowed to compete. Clearly, government employees will act like a private provider and set their own internal incentives.

At the end of the designated period, FIA will adjust the prices for the various categories of children to reflect the accumulated experience. For example, suppose that the rate of adoption of disabled children is low, then the price paid by FIA to the adoption agencies needs to be raised for the next time period. At the same time if at given prices, demand for white infants is greater than the number of available children, indicated by a rapid adoption rate, then the price set by FIA should be lowered. The savings generated from white infants will provide needed resources for the hard-to-place groups. Obviously, this pricing mechanism will help ensure that only efficient providers, whether private or public, will survive. The range of prices offered by the FIA is, of course, constrained by its budget.

Conclusion

This article reviewed and evaluated pioneering privatization efforts of foster care and adoption by the states of Kansas and Michigan. Evaluation of these two experiences is aimed at guiding other states that intend to improve their child welfare programs. Kansas contracted out the service to one provider for adoption and one for each of five regions for foster care, establishing undesired and unnecessary private monopolies. Michigan allowed for more competition in adoption, by advertising information and enabling 80 companies to compete. Kansas experimented with pricing policies and Michigan set differentiated prices by ease and time to adoption.

Economic theory suggests that social welfare rises as the level of competition increases. Michigan's model that incorporated ubiquitous information and large number of providers is more promising in achieving an efficient solution. However, Michigan's arbitrary non-market pricing does not ensure timely and best placement practice by providers. Privatization led to some distinct improvements; both states know better the cost of providing foster care and adoption services by private providers, and the number of adopted children increased. These are clearly significant improvements.

However, there is a more direct way to shorten foster care time and improve adoption placement. Privatization works in Kansas and Michigan on the intermediaries to facilitate the transactions. We can and probably should increase competition as we have already suggested, and in that vein managed competition should not be overlooked. Can we do more to improve the situation? As Dave Thomas, who was adopted at six weeks and founded Wendy's hamburger chain, said: "I know firsthand how important it is for every child to have a home and loving family. Without a family I would not be where I am today" (Philadelphia Inquirer 2002).

Our review of public and private adoption practices and economic theory suggests a more direct way to shorten foster care time and improve adoption placement. We suggest a market approach.

The adoption market is in fact operating inefficiently. For white infants there is excess demand. Interested adoptive families must wait a long time for a child or attempt to adopt a foreign child. Occasionally those wanting a child resort to black markets, in the process violating U.S. or foreign laws. At the same time, there are more older, minority, and disabled children awaiting adoption than families desiring them, resulting in children living in foster care or institutions for a long time. The reason for this failing process is simply that prices are not allowed to fluctuate to eliminate shortages or surpluses. Instead the emphasis has been on improving the work of the intermediaries, the adoption agencies, rather than improving the transaction itself.

Economic theory again comes to the aid. As in the case of Michigan, let the state widely advertise the attributes of and children available for adoption. Interested families may contact the state child welfare agency or any licensed private adoption agency. In the application form, families already approved for adoption will indicate the child of their choice and the price they are ready to pay or how much they have to be given for adopting the particular child. Markets for children in demand like infants will clearly generate substantial revenues while markets for less demanded older and disabled children may require payments.

The state should establish a fund that includes the revenues generated from the adoption of the desired children, the existing budget, and any revenues for adoption from the federal government. These funds will enable payments for the less desired children. Since this process will shorten the period children spend in foster care, some additional revenues will become available.

As with any market, children will end up in the homes of those who can most afford and desire them and are willing to pay the most. As a result, the adopting parents are likely to better care for them. This is also likely to be a more suitable home than in the existing system where adopters who happen to be next on the waiting list get an unknown child and effectively pay just a nominal price. In the case of the less desired children, the family that requests the lowest amount will get the child. Again, this family wants a particular child and is ready to make the greatest financial sacrifice. It is most likely that this family will provide a better home for the child. The process will shorten the painful waiting time for adoption for both the children and interested families.

Creating such markets will eliminate black markets existing today and will reduce the influence and financial rewards of intermediaries like lawyers and private adoption agencies, hence making more resources available for the hard-to-place children. Social worker supervision and monitoring of the adoption process will of course be maintained by public and private agencies.

Economists claim that markets become more efficient as more suppliers and consumers participate. In the case of adoption, we suggest that information about available children be disseminated nationwide. Some barriers exist for out-of-state adoption. Some states may have a surplus of children while other states have excess demand. Also, the "quality" of interested parents may vary among the states. There is no need for the child to remain within the state. Converting the adoption process to a nationwide market will improve the "quality" and fit of the adopting parents, shorten the time for adoption, and reduce costs for the states.

Markets are working to allocate resources and achieve efficiency for almost all goods and services. New laws may be required but states and possibly the federal government should experiment with the power of a competitive market to improve child welfare services. Rhetoric about selling children aside, there is no reason to prevent markets from benefiting children.

References

Barth, R. P. (1997) "Effects of Age and Race on the Odds of Adoption versus Remaining in Long-Term Out-of-Home Care." Child Welfare 76 (2) (March/April): 285-308.

Blackstone, E. A., and Hakim, S. (1997) "Private Ayes: A Tale of Four Cities." American City & County 112 (2) (February): PS4-PS12.

Craig, C.; Kulik, T.; James, T.; and Nielsen, S. (1998) "Blueprint for the Privatization of Child Welfare." Reason Public Policy Institute, Policy Study No. 248 (December).

Evan B. Donaldson Adoption Institute (2000) "Adoption in the U.S." (www. adoptioninstitute.org).

James Bell Associates (2001) "External Evaluation of the Kansas Child Welfare System: July 2000-March 2001." 3rd Quarterly Report. Arlington, Va. (9 August).

Legislative Division of Post Audit, State of Kansas (2001) "The State's Adoption and Foster Care Contracts: Reviewing Selected Financial and Service Issues." Topeka, Kansas (January).

Mainstream Inc. (2001) "Privatization of Adoption in Kansas: A Report to Dave Thomas Foundation for Adoption." Topeka, Kansas (May).

Philadelphia Inquirer (2002) Obituary of Dave Thomas, 9 January: B7.

Snell L. (2000) "Child-Welfare Reform and the Role of Privatization." Reason Public Police Policy Institute, Policy Study No. 271 (October).

U.S. Department of Health and Human Services, Administration for Children and Families (2001) "The AFCARS Report" (June).

U.S. Department of Justice, office of Juvenile Justice and Delinquency Prevention (1999) "Family Disruption and Delinquency." Juvenile Justice Bulletin (September).

Cato Journal, Vol. 22, No. 3 (Winter 2003). Copyright [c] Cato Institute. All rights reserved.

Erwin A. Blackstone and Simon Hakim are Professors of Economics at Temple University and are associated with the Center for Competitive Government at Temple.