A token economy is a system of contingency management based on the systematic reinforcement of target behavior. The reinforcers are symbols or "tokens" that can be exchanged for other reinforcers. A token economy is based on the principles of operant conditioning and behavioral economics and can be situated within applied behavior analysis. In applied settings token economies are used with children and adults; however, they have been successfully modeled with pigeons in lab settings.
Tokens must be used as reinforcers to be effective. A token is an object or symbol that can be exchanged for material reinforcers, services, or privileges (back-up reinforcers). In applied settings, a wide range of tokens have been used: coins, checkmarks, images of small suns, points on a counter. These symbols and objects are comparably worthless outside of the patient-clinician relationship, but their value lies in the fact that they can be exchanged for other things. Technically speaking, tokens are not primary reinforcers, but secondary or learned reinforcers. Much research has been conducted on token reinforcement, including animal studies.
Tokens have no intrinsic value, but can be exchanged for other valued reinforcing events: back-up reinforcers. Most token economies offer a choice of differing back-up reinforcers that can be virtually anything. Some possible reinforcers might be:
- Material reinforcers: sweets, cigarettes, journals, money (although money holds value outside of the patient-clinician relationship)
- Services: breakfast in bed, having your room cleaned, activities
- Privileges and other extras: passes for leaving the building or area, permission to stay in bed, phone calls, having your name or picture on the wall.
Back-up reinforcers are chosen in function of the individual or group for which the token economy is set up, or depending upon the possibilities available to the staff. Prior to starting the staff decides how many tokens have to be paid for each back-up reinforcer. Often, price lists are exposed or given to the clients. Some back-up reinforcers can be bought anytime, for other exchange times are limited (e.g. opening times of a token shop).
Specified target behaviors
There is a broad range of possible target behaviors: self-care, attending activities, academic behavior, disruptive behavior. A token economy is more than just using exchangeable tokens. For a token economy to work, criteria have to be specified and clear. A staff member giving tokens to a client just because he judges he is behaving positively, is not part of a token economy because it is not done in a systematic way. Sometimes client manuals have specifications how many tokens can be earned by each target behavior. For instance, if making the bed is a target behavior, staff and clients have to know how a well-made bed looks like: do the sheets have to be put under the mattress, cushion on top? However, often these specifications are hard to make: behavior such as eating politely and positive cooperation are hard to specify. While planning how many tokens can be earned by each target behavior some factors have to be considered: on the one hand clients should be able to earn a minimal amount of tokens for a minimal effort, and on the other hand clients should not earn too much too soon, making more effort useless.
Sometimes the possibility of punishment by token loss is included, technically called 'response cost': disruptive behavior can be fined with the loss of tokens. This also should be clearly specified before the application starts. Clients can be involved in the specifying of the contingencies.
Other features of a token economy
Token reinforcement is essential, but is always accompanied by social reinforcement. Tokens are intended to make reinforcement explicit and immediate, and to strengthen behavior, but in the end social reinforcement should be sufficient to maintain what’s been learned.
All principles of operant learning are applied within a token economy. Shaping implies clients aren’t expected to do everything perfectly at once; behavior can be acquired in steps. Initially clients can be reinforced for behavior that approaches the target. If the target behavior is keeping attention during a 30 minutes session, clients can initially already get (perhaps smaller) reinforcement for 5 minutes of attention.
Immediacy of reinforcement
Reinforcement will greater influence behavior if given shortly after the response is emitted. The longer people have to wait for a reward, the less effect and the less they will learn. This is the principle of delay discounting. Immediate token reinforcement can bridge later reinforcement.
Learning to plan ahead and save earnings
Sometimes clients can earn larger rewards like the permission to spend a weekend at home, going to a movie, or having a class excursion. When such rewards would be given at once for one instance of a target behavior, the scarce resources would soon be depleted and consequently the incentives would be lost. One advantage of tokens is they can be used to divide larger rewards into parts: clients can save tokens to buy more expensive rewards later. This implies they shouldn’t immediately spend all earned tokens on attractive smaller rewards, and instead learn to plan ahead. This way they can acquire self-control. (See Delayed gratification.)
Individual and group contingencies
Mostly token economies are designed for groups. The system is running for a whole ward or class. Within this group contingency specific individual goals and reinforcers can be added. Though sometimes a token economy is designed for only one specific individual.
The power of a token economy largely depends on the consistency of its application. To achieve this thorough staff training is essential. Some token economies failed exactly on this point. Token economies imply rights and duties for clients as well as for staff. When, according to the system, a client deserves tokens, he should get them, even when a staff member judges he doesn’t deserve them because he has been impolite the day before. Family education and involvement is very important. They can support the system or they can undermine it, for instance by secretly giving undeserved rewards.
Often, token economies are leveled programs. Clients can pass through different levels until they reach the highest level. At that point, behaviors are performed without token reinforcement. Higher levels require more complex behaviors. The incentive to progress from one level to the next is the availability of increasingly desirable reinforcers.
History of the token economy
In the early 19th century, long before there was any knowledge about operant learning, there were some precursors of token economies in schools and prisons. In those systems points could be earned and exchanged for many different items and privileges. Only in the 1960s the first real token economies arose in psychiatric hospitals. Teodoro Ayllon, Nathan Azrin and Leonard Krasner were important pioneers in these early years. The very first token economy bearing that name was founded by Ayllon and Azrin in 1961 at Anna State Hospital in Illinois. In the 1970s the token economies came to a peak and became widespread. In 1977 a major study (a randomized controlled trial), still considered a landmark, was published. This study showed the superiority of a token economy compared to standard treatment and specialized milieu therapy. Despite this success token economies declined from the 1980s on. It became fairly quiet on that front due to a variety of problems and criticism.
Problems and controversy
Especially the application of token economies with adults became a matter of criticism. In addition some impediments and the evolution of mental health care caused troubles.
Changes in patient care
Token economies have proven their effectiveness and utility for chronic psychiatric patients, despite requiring months or even years to achieve optimal results. This causes problems when insurance and government policies increasingly require the shortest possible hospital stays. Because emphasis has shifted to community-based treatment, outpatient and home-based care is often the preferred choice over institutionalization. This decentralization of patient care methods makes it difficult to further study and develop token economies in a scientific, research-oriented method.
Legal and ethical issues
Token economies can present issues with concern to patient rights. The right to have their personal properties, basic comfort and freedom of choice of treatment constrained the possibilities for token economies. In addition, ethical and personal concerns of staff members arose: is it ethically justified to use cigarettes as secondary reinforcers; is it human to rely on ‘reward and punishment’ as means for treatment (behavior modification often is reduced to this by opponents); isn’t sincere human contact much more valuable and effective; and do token economies reduce human interaction to trade, therefore centering patients' attention (and behavior) upon materials?
Response maintenance and generalization
Problems with maintaining what’s been learned and the generalization toward new situations have been signaled. When the token programs stops the acquired behavior might disappear again.
Extrinsic versus intrinsic motivation
Recent applications and findings
In the last 50 years much research has been conducted on token economy. The first 20 years were especially productive. Despite controversy and a lack of implementation token programs are alive and well in several settings.
In adult settings token economies are mostly applied in mental health care. The criticism that clients have no choice and are being forced, is countered by offering them the choice to enter the token program or not to enter, or to leave again once entered. The vast majority of clients in past studies voluntarily chose to stay in the program. Research shows the effects of token economies can more or less be divided into three categories:
- No effect: 5 to 20% of the clients do not (or minimally) respond to the token economy;
- Only effect while the token economy is active and no effect once stopped: in this case the token economy is functioning as a prosthesis (like a wheelchair); it does not permanently help the patient once terminated but is necessary to maintain normal functioning;
- Well established long-term effects (including community reintegration).
The first token economies were designed for chronic, treatment-resistant psychotic inpatients. Even now token economies are applied to clients with schizophrenia, who are often resistant to common behavioral treatment approaches. Sometimes the token economy is used as a lasting prosthesis. Sometimes it’s used to help such clients reach resocialization. A token economy (of course always in combination with other interventions) succeeded in the community reintegration of 78% of the clients within an average period of 110 days, after more than 7 years of uninterrupted hospital stay. Research shows clients experience the token economy with positive reception. Several recent reviews of psychosocial treatment for schizophrenia explicitly mention token economy as an effective, evidence-based treatment.
The application of token economies has been extended from psychiatric rehabilitation services to acute psychiatric units. A token economy was successful in decreasing the aggression on a ward where clients on average stayed for less than three weeks.
As a result of heavy ethical criticism, token economies developed a negative stigma and, as a result, systems were sometimes introduced with aliases. This was especially the case in substance abuse treatment settings (although some systems for smoking cessation continue to use the term token economy). For some time, systems derived from token economies were used under the name contingency management; initially this was more broadly defined and referred to any direct coupling of consequences (reinforcements or punishments) with behavior (for example staying clean ); later it referred specifically to one kind of token economy.
The community reinforcement approach can be combined with contingency management; ‘tokens’ are used, whereas contingency management employs the term "vouchers" (or related terms). Research shows this kind of token economy is easily applied outside of hospitals and is effective, allowing for less hospital-based treatment - although contingency management is used in the treatment of drug abuse in both inpatient and outpatient settings.
Token economy is also being applied in settings for adults with developmental disabilities. Target behaviors can vary in types of social behavior and self care, or the decreasing of inappropriate and/or disruptive behavior.
Children and adolescents
Token economies have been applied to children and adolescents with developmental disabilities as well as in schools.
A token economy has proven effective in increasing attentiveness and motivation in completion of tasks for children with developmental disabilities. Research shows it can help to diminish disruptive behavior and promote social behavior.
Token economies have been applied in schools, particularly special education programs as well as in other programs. Positive results can imply increased attention and decreased disruptive behavior. In educational settings token economy seems to raise the intrinsic motivation to complete assigned tasks. But there’s still need for more research.
- K. Jackson and T.D. Hackenberg : Token reinforcement, choice, and self-control in pigeons. Journal of the Experimental Analysis of Behavior. 1996 July; 66(1): 29–49. doi: 10.1901/jeab.1996.66-29 PMCID: PMC1284552
- A.E. Kazdin : The Token Economy. A review and evaluation. New York: Plenum Press, 1977. ISBN 0-306-30962-9
- T.D. Hackenberg: Token reinforcement: a review and analysis. Journal of the Experimental Analysis of Behavior, 2009, 91, 257-286.
- Ormrod, Jeanne Ellis (2009). Essentials of Educational Psychology (2nd ed.). Upper Saddle River, N.J.: Merrill. p. 338. ISBN 978-0-13-501657-2.
- S.M. Glynn: Token economy approaches for psychiatric patients. Progress and pitfalls over 25 years. Behavior Modification, 1990, 14, 383-407.
- R.P. Liberman: The token economy. American Journal of Psychiatry, 2000, 157, 1398.
- T. Ayllon & N.H. Azrin : The Token Economy: a motivational system for therapy and rehabilitation. New York: Appleton-Century-Crofts, 1968. ISBN 0-390-04310-9
- G.L. Paul & R.J. Lentz : Psychosocial treatment of chronic mental patients: milieu versus social-learning programs. Cambridge, MA: Harvard University Press, 1977. ISBN 0-674-72112-8
- A.E. Kazdin: The token economy: a decade later. Journal of Applied Behavior Analysis, 1982, 15, 431-445.
- C.M. Coelho, e.a.: Rehabilitation programs for elderly women inpatients with schizophrenia. Journal of Woman & Aging, 2008, 3, 283-295.
- S.M. Silverstein, e.a.: Behavioral rehabilitation of the "treatment-refractory" schizophrenia patient: conceptual foundations, interventions, and outcome data. Psychological Services, 2006, 3, 145-169.
- M.F. Lin, e.a.: Significant experiences of token therapy from the perspective of psychotic patients. Journal of Nursing Research, 2006, 14, 315-323.
- G.D. Shean: Evidence-based psychosocial practices and recovery from schizophrenia. Psychiatry, 2009, 72, 307-320.
- L.B. Dixon, e.a.: The 2009 schizophrenia PORT psychosocial treatment recommendations and summary statements. Schizophrenia Bulletin, 2010, 36, 48-70.
- J.P. LePage, e.a.: Reducing assault on an acute psychiatric unit using a token economy: a 2-year follow-up. Behavioral Interventions, 2003, 18, 179-190.
- V. McLeod: How to quit smoking by using a token economy method. EzineArticles, 2011.
- N.M. Higgins & N. Petry: Contingency Management Incentives for sobriety. Alcohol Research & Health, 1999, 23, 122-126.
- M.W. Lewis: Application of contingency management-prize reinforcement to community practice with alcohol and drug problems: a critical examination. Behavior and Social Issues, 2008, 17, 119-138.
- H.G. Roozen, e.a.: A systematic review of the effectiveness of the community reinforcement approach in alcohol, cocaine and opioid addiction. Drug and Alcohol Dependence, 2004, 74, 1–13.
- L.A. LeBlanc, L.P. Hagopian & K.A. Maglieri: Use of a token economy to eliminate excessive inappropriate social behavior in an adult with developmental disabilities. Behavioral Interventions, 2000, 15, 135-143.
- J.L. Matson & J.A. Boisjoli: The token economy for children with intellectual disability and/or autism: a review. Research in Developmental Disabilities, 2009, 30, 240-248.
- K. Zlomke & L. Zlomke: Token economy plus self-monitoring to reduce disruptive classroom behaviors. The Behavior Analyst Today, 2003, 4, 177-182.
- G. LeBlanc: Enhancing intrinsic motivation through the use of a token economy. Essays in Education, 2004, 11.
- H.A. Filcheck & C.B. McNeil: The use of token economies in preschool classrooms: practical and philosophical concerns. Journal of Early and Intensive Behavior Intervention, 2004, 1, 94-104.