Motivation and emotion/Book/2022/Hidden costs of reward

Overview


Offering reward is a well-known and traditional approach for inducing desirable behaviour or compliance. This extrinsic motivation works well because it is able to offer to a person immediate and obvious results for changing their behaviour. There follows two important questions: how effective is reward in instances where results are not obvious or immediate over the long-term, and how can we encourage someone to develop an intrinsic desire to behave in the preferred manner?

While extrinsic motivators have particular efficacy in certain common behaviour management strategies, reward does also result in significant hidden costs. These unintended consequences may undermine a person's medium and long-term motivation and ultimately harm the effort to encourage particular behaviours. This chapter explores how reward is best used, how improper use of reward-based extrinsic motivators can undermine the internalisation of behaviour, and what strategies can be used to aid reinforcement and development of intrinsic motivation.



Imagine for a moment you are a new primary school teacher in a class full of 9 year old students. You have inherited a classroom with a sticker-chart based behaviour reward system. While many of the children appear to remain on task regardless of if you remember to reward them with stickers or not, some will only return to on-task behaviour when you mention the chart and not before. Some even verbalise they want a sticker before they will return to their task.

Consider:
 * With limited time in each lesson, how can you be sure that this reward system is the best result for effort?
 * Could this dependence on reward be causing more or new problems?
 * Does this approach effectively teach self-regulation?


 * What is the function of reward as an extrinsic motivator?
 * What is the theoretical underpinning of intrinsic motivation?
 * What are the hidden costs and risks of using reward?
 * How do we effectively encourage intrinsic motivation?

Operant conditioning


Operant conditioning is a theory of human learning first developed in the late 19th century by Edward Thorndike and was extensively studied throughout the early to mid 20th century by B. F. Skinner, who also coined the term (Morris et al., 2022). It is the basis of applied behavioural psychology. Operant conditioning sees behaviour as an associative process where consequences affect the likelihood of a behaviour to be elicited (Stone et al., 2008). These consequences (also called feedback) are categorised as either reinforcement or punishment (see figure 2).

Reward can be any number of positive feedback options, not limited only to tangible items like money, prizes, or opportunities, but also intangible items and social rewards such as praise or affection. Reward may also be in the form of a negative reinforcer such as escape from performing a boring activity or the removal of an unpleasant situation. Positive (or additive) reinforcement is considered the most effective form of consequence within operant conditioning (Skinner, 1950). Reward is particularly effective because it is the most directly informative of the behaviour event consequences. Most simply, reward demonstrates that the action most recently taken is an appropriate behaviour and repeating that action in similar circumstances in future will also likely elicit further reward. This situation mimics the neurological circumstances from which classical conditioning was derived (Skinner, 1950).

While reward can be used in numerous behaviour management strategies, such as intermittent reinforcement or nudge theory, operant conditioning makes the purpose of reward straight-forward and uncomplicated: if you want someone to behave a certain way, give them a reward when they do so (Hall et al., 1972).

Self-determination theory
Self-determination theory is a theory of human motivation developed in the late 20th century by Edward Deci and Richard Ryan focusing on self-motivation (Deci et al., 1994). The theory was predicated on Ryan and Deci's research regarding intrinsic motivation, or the initiation of an activity for the purpose of satisfying an internal goal or interest. This is opposed to the classical use of extrinsic motivation, or the fashion by which external events and stimuli encourage the initiation of behaviour (Deci et al., 1999). This research was used to develop a model of motivation which explains intrinsic motivation as a function of the internal psychological satisfaction a task may produce (Ryan & Deci, 2000). This psychological satisfaction is divided into three constituent perceptual ingredients; individual autonomy, a sense of competence, and relatedness to the task (as shown in figure 3).

Subsequent research has expanded self-determination theory to include a theoretical taxonomy demonstrating the extent to which motivation exists within three major categories; low motivation, externally derived motivation, and internally derived motivation (as shown in figure 4). While self-determination theory had been instrumental in categorising motivation types and describing their use, it was not able to specify the conditions upon which motivation may change between these categories. Ryan and Deci developed the cognitive evaluation sub-theory to describe this process (Ryan et al., 1983).

Cognitive evaluation theory
Cognitive evaluation theory is a sub-theory of self-determination theory first suggested by Deci and Ryan in 1980 that investigates how external influences affects intrinsic motivation (Ryan & Deci, 2000). The theory refers to an individual’s social environment and in particular events such as feedback or rewards. It asserts that external stimuli that promote self-confidence may enhance intrinsic motivation, while events that harm self-confidence may negatively impact intrinsic motivation (Ryan et al., 1983).

Ryan (1982) explains that cognitive evaluation theory is predicated upon a set of three propositions. First, that if an event exists to control behaviour then autonomy and intrinsic motivation will be harmed and vice versa. Second, it proposes that events which are informative and positive regarding the competence of the individual will increase perceived competence and intrinsic motivation while negative informational events will harm perceived competence and intrinsic motivation. Third and lastly, the theory proposes that the extent to which an event embodies the first two proposals will determine the overall effect upon motivation (as shown in table 1).

The usefulness of reward to developing motivation and the movement of that motivation up or down the self-determination model is contingent on the pressure that reward gives to the psychological satisfaction of an individual (Ryan et al., 1983). Therefore, when considering what efficacy contrived consequences like rewards have upon motivation, it is best framed within a delicate balance of enhancing autonomy and competency or enticing behaviour at the expense of that person's autonomy. It is through this lens we may specify how reward comes with hidden costs below.

 Take a moment to return to the previous application example. It may now be evident that the efficacy of our sticker-chart is now contingent on the specific behaviour we are seeking to alter. Stickers may likely perform an effective role in aiding motivation to tasks which already contain low levels of individual autonomy for our students such as tasks regarding cleaning the room on a time limit or encouraging an unmotivated student to participate in a physical education activity. However, serious questions now arise regarding the usefulness of the reward system when it comes to encouraging students to work through a worksheet assessment or to stay silent during a reading task. Consider:
 * Is a reward-based system going to undermine students already exhibiting an intrinsic desire to perform well?
 * How will reward affect motivation on an assessment task?
 * Will reward undermine pro-social motivations?

 {In operant conditioning, reward is what kind of feedback? - Positive reinforcement. + Reinforcement (positive and negative). - Punishment. - Absence of punishment.
 * type=""}
 * Reward is often a positive reinforcement, and that is its most common understanding, but it is also more than that.
 * That's correct. Reward can be both a positive or negative reinforcement.
 * Reward is not a form of punishment.
 * Reward as a concept is inclusive of negative reinforcers such as removal of negative stimuli, but it is also more than that.

{Which of the following is not an ingredient to psychological satisfaction as shown by self-determination theory? - Competence. - Autonomy. + Motivation. - Relatedness.
 * type=""}
 * Incorrect. A sense of competence is important for psychological satisfaction in SDT.
 * Incorrect. Autonomy is essential for motivational satisfaction under SDT.
 * Correct. SDT seeks to discover how motivation develops. Motivation is not in and of itself a feature of psychological satisfaction according to the theory.
 * Incorrect. Feeling related to a task or event is an essential part for psychological satisfaction.

{According to cognitive evaluation theory, if a reward is used to control your behaviour, is this likely to increase or decrease reliance on extrinsic motivation? + Increase. - Decrease. - Neither.
 * type=""}
 * Correct. People are more likely to demand more rewards to maintain a demanded behaviour.
 * Incorrect. Reward is more likely to increase a need for extrinsic motivation for behaviour-controlling events.
 * Incorrect. Reward is more likely to increase a need for extrinsic motivation for behaviour-controlling events.

The hidden costs of reward
Unintended consequences of the misuse of reward for behaviour modification and instruction can be grouped into three broad categories: reward which undermines intrinsic motivation; reward which interferes with the process of learning; and reward which compromises personal autonomy. It should also be noted that while the vast majority of research below has been focused on the cognitive and psychological aspect of self-determination theory, medical research by Murayama et al. (2012) using fMRI imagining appears to indicate reward may cause a decrease in activation of the midbrain relating to cognitive engagement thereby demonstrating at least some biological support of the theory.

Undermining intrinsic motivation
One aspect by which reward appears to undermine motivation is the overjustification effect, so called because of the way intrinsic justification for a behaviour is replaced by a justification for the reward. The overjustification effect not only appears to be partially responsible for how extrinsic motivation can reduce intrinsic drive, but also appears to show that even in instances where intrinsic desire still exists, rewards for behaviour will decrease a person's desire to test their competence resulting in an extrinsic justification to participate in simpler tasks (Pittman et al., 1982; Ryan et al., 1983).

Issues related the overjustification effect have also been applied to macroeconomic theory. Crowding theory was identified by Richard Titmuss in the 1970s and formalised by Burno Frey in 1997 (Frey & Jegen, 2002). Frey and Jegen (2002) notes that many economic incentives rely on the intrinsic motivations of individuals to effect meaningful productivity like in markets where access to financial capital or incentives do not meet the level necessary to ensure proper management such as the environment or organ donation. The extrinsic financial incentive has been found to 'crowd-out' the internalised incentive, ultimately resulting in less market activity for these areas. Research by Frey and Goette (1999) indicated that when volunteers were paid, pay needed to approach US$50 an hour before productivity and work hours approached that of the non-remunerated volunteers. Though it should be noted that a study by Niza et al. (2013) did not find evidence of crowding-out when analysing quantity of blood donations against payments. Further discussion on the topic seems to indicate that it is most prevalent in working environments than in discrete donations, but more research is needed to understand the connection between crowding theory, cognitive evaluation theory, and macro economics.

Finally it should be noted that reward only appears to interfere with intrinsic motivation when the person is already aware the task will result in a real and tangible reward such as payment or permission. Unexpected rewards do not appear to interfere with intrinsic motivation and neither do praise, verbal feedback, or other non-tangible rewards (Ryan & Deci, 2000).

Interference with the learning process
The second area where reward may interfere with motivation is within the learning process itself. Ryan and Deci (2000) discuss how the seeking of reward can undermine internal awareness of how well a person is performing a task. This makes it difficult to focus on building competence on a task as the finer details of skill acquisition is lost in simply completing the task to achieve the reward. From a motivational standpoint, this makes reward uninformative. Research by Koestner et al. (1984) confirms that while reward can be informative from an operant conditioning framework (this is a behaviour the environment wants performed), reward does not necessarily provide any direct information as to if you are improving or stagnating at learning a new skill.

This process is further complicated when we consider activities of which someone is already highly motivated. When tangible reward is utilised after a subject already exhibits intrinsic or autonomous motivations, not only is there a risk of the overjustification effect, but also research by Daniel and Esser (1980) found that overall interest in the task was undermined by the reward. This was particularly true for complex or difficult tasks which the participants where already well skilled. Daniel and Esser continue to note that intrinsic motivation is essential to autonomous learning and that reward-based motivators impede the investment of free-time to the learning process.

Compromising personal autonomy
The final area where reward may interfere with motivation is the sense of personal autonomy and self control. Reward is, as a contrived extrinsic motivator often offered by a supervisor, teacher, parent, or other authority, generally used as a form of incentive against the status quo or as a means of behaviour control. Koestner et al. (1984) discusses not only how reward may interfere with the learning process, but also demonstrates that reward appears to lessen the frequency of spontaneous autonomous behaviours - that tasks previously taken up by motivated individuals of their own accord are put off until the opportunity for reward presents itself and therefore limiting the activity not to the desires of the person, but to the needs of the supervisor or environment. This confirms research by Calder and Staw (1975) which also found evidence of reward undermining the sense of self and replacing their desire to perform a task for self-satisfaction with an internalised perception that they only needed or wanted to complete the task to receive the reward.

Personal autonomy counts for more than task-oriented behaviour. It also is a feature of social behaviour and motivation (Pavey et al., 2012). The introduction of reward into this process interferes with the informative social functions of empathy and replaces that social behaviour with contrived actions which seek to earn tangible extrinsic reward instead of supporting social relationships (Pavey et al., 2012). Importantly, social considerations may limit the costs of reward to intrinsic motivation if reward is designed such that it minimises the aspect of control it often embodies. Jang (2008) found that if rewards are offered using language which confirms the autonomy of the individual and refrains from overt manipulation as well as specifying the relation of the task to the individual and its meaningful importance, then the negative effects may be significantly mitigated. Notably, Jang shows that if reward is used in such a way that it fosters ownership of the task (often, but not always, with intangible rewards), it may then be effective in not only simple or low-motivation tasks, but also developing motivation in moderately complex or medium-motivation tasks. She posits this may be because the motivator of the reward lessens the cognitive requirements of the moderate task allowing more energy to be expended on the skill or behaviour itself. However, even this socially-conscious attempt to use reward still harms the highest levels of intrinsic motivation as above.

 Considering the above, we can now imagine some clear limitations to our sticker-chart reward system. As an exercise, can you identify: Write these down on a piece of paper or leave your ideas in the Talk page.
 * Two occasions where this system may undermine intrinsic motivation?
 * Two occasions where it may interfere with the learning process?
 * Two occasions where it may compromise student autonomy?
 * Also, can you imagine any ways our reward system can be used as intended?

 {True or false: There is no evidence self-determination theory applies beyond the individual. - True. + False.
 * type=""}
 * Incorrect. Disruptions to intrinsic motivation has been researched by economists in parallel to and sometimes independent of psychologists.
 * Correct. Disruptions to intrinsic motivation has been researched by economists in parallel to and sometimes independent of psychologists.

{Which of the following is not a category upon which reward may harm motivation? - Interference with the learning process. - Undermining intrinsic motivation. - Compromise of personal autonomy. + Devalues efficacy of punishment.
 * type=""}
 * Incorrect. Reward may interfere with the learning process by disrupting understanding of end-goals.
 * Incorrect. Reward may undermine intrinsic motivation by crowding out or the overjustification effect.
 * incorrect. Reward is often used as a form of behaviour management and often interferes with a sense of autonomy.
 * Correct. Reward does not interfere with punishment as a feedback mechanism.

Conclusion
While reward is the motivator of choice for applied behavioural psychology and especially operant conditioning and has been shown to be particularly effective in developing the foundations of behavioural motivation, an understanding of the theory of self-determination demonstrates that a person's cognitive and psychological needs play directly into the development of autonomous self-driven intrinsic motivation. This internalisation of motivation is the highest level of motivation and is developed through a sense of competence, autonomy, and relatedness.

Reward comes with significant costs to the sense of self-determination by undermining intrinsic motivation, by interfering with the learning process of complex tasks, and by compromising the sense of personal autonomy and control. While reward may be a good starting motivator for simple or new tasks, it is a distraction from the processes which develop motivation in the medium and long-term. Remember, if a reward doesn't build upon the sense of self-satisfaction, then reward will likely interfere with it.