Self-Control Concept

Self-Control Concept

The self-control concept concerns how people manage a conflict between an immediate temptation or effort and an outcome they value over a longer horizon. In health economics, this can illuminate some gaps between intended and actual behaviour, such as postponing a preventive appointment, but it is not a diagnosis or a complete explanation of any person's choices. This page distinguishes preferences from constraints, illustrates one formal account of present bias, and examines supportive interventions and their welfare implications.

When a choice feels different at the moment of action

A person may sincerely plan to take an action next week and then defer it when its immediate cost arrives. The near-term effort or discomfort can feel more salient than a delayed health benefit, even when both were anticipated. This kind of time inconsistency is one possible self-control problem; ordinary uncertainty, changed circumstances, informed preferences or inability to obtain care can produce the same observed delay.

Observation	Possible explanation	Evidence needed
A person delays an appointment.	Present bias, fear, transport difficulty or an unavailable slot.	Preferences, access, timing and stated reasons.
A person misses a prescribed dose.	Forgetting, side effects, cost, treatment change or conflicting instructions.	Clinical context and opportunity to take it.
A person seeks an advance commitment.	Anticipated temptation or a desire for reminders and structure.	Voluntary demand and consequences after adoption.

Do not label someone as lacking self-control solely because their behaviour differs from a clinician's recommendation. A health choice may reflect genuine trade-offs, limited money, disability, competing responsibilities or mistrust. These conditions should be examined before proposing a behavioural remedy.

A simple model of present bias

One behavioural model applies ordinary discounting between future periods while giving extra weight to the immediate period. A stylised quasi-hyperbolic expression is $U_t=u_t+\beta\sum_{k=1}^{T-t}\delta^k u_{t+k}$, where $u_t$ is utility now, $0<\delta\leq1$ is the usual per-period discount factor and $0<\beta\leq1$ weights all later periods relative to now. When $\beta<1$, a cost becoming immediate can change the ranking of an action and delay; this is a model assumption, not a measured property of everyone.

Consider an invented action that costs 75 utility units on the day it is done and yields 100 units one period later. With $\delta=0.9$ and $\beta=0.8$, a person evaluating tomorrow's action today compares both its cost and benefit as future events: the benefit relative to its future cost is $0.9\times100=90>75$, so the plan looks worthwhile. Tomorrow, the 75-unit cost is immediate but the benefit is weighted $0.8\times0.9\times100=72<75$, so the same stylised decision maker would defer.

Evaluation point	Illustrative comparison	Model prediction
One period before action	0.9*100 - 75 = 15 in units relative to the action date.	Plan to act when both consequences are future.
At the action date	0.8*0.9*100 - 75 = -3.	Prefer not to act with immediate effort.
If beta equals one at action date	1*0.9*100 - 75 = 15.	This particular reversal disappears.

At the earlier date, multiplying both future consequences by the common positive factor $\beta\delta$ leaves their comparison unchanged; the table reports values relative to the action date, not total current utility. The example does not estimate a person's beta or delta from one missed action, and utility units are not money, QALYs or observed clinical benefit. Other models of temptation and self-control use different assumptions.

Supports that can help without blaming the patient

A reminder can address limited attention; a scheduled appointment can reduce planning effort; a stable cue can support repetition; and a voluntary commitment can alter future options or incentives. Each targets a different mechanism. Offer support only after checking that the person can access and wants the underlying care.

Hard commitments may impose penalties or restrict flexibility, while softer commitments create plans or social accountability with less loss from unforeseen circumstances. An intervention can improve the targeted behaviour and still make some participants worse off if they face penalties, prefer to change course or lack resources to comply. Measure consent, uptake, failure, costs and wellbeing, not just adherence.

Evaluate the intervention and its economic consequences

An evaluation should compare the offered support with a real alternative and define outcomes over time. Measure whether people begin and continue appropriate care, subsequent health, adverse consequences, staff and patient time, and who bears any payment or penalty. Heterogeneity matters: the same device may help someone facing a predictable temptation and harm someone whose work schedule changes unexpectedly.

Suppose a fictional service offers voluntary planning support to 100 people at an additional programme cost of £2,000, and 20 more people complete a chosen appointment than under usual scheduling. The incremental programme cost per additional completion is $£2{,}000/20=£100$. That intermediate ratio does not show that self-control was the causal mechanism or that the programme is cost effective; appointment value, clinical outcomes and any burdens remain to be established.

Spreadsheet item	Illustrative formula	Interpretation
Additional programme cost	=2000	£2,000 under fictional assumptions.
Additional completed appointments	=20	Difference against the specified comparator.
Incremental cost per completion	=2000/20	£100, excluding downstream outcomes and costs.

An economic model should not translate every additional appointment into a fixed health gain. The benefit depends on what the appointment changes, for whom, and over what horizon. Report the effect of access barriers and any penalty payments separately from real resources used.

Common misunderstandings and ethical limits

Present bias is a useful theoretical lens, but it cannot be read directly from a single action. Stated long-run preferences may be uncertain or may change as a person learns more about risks and side effects. A model should inform respectful design, not override patient autonomy.

• Distinguish mechanisms: Forgetting, limited access, uncertainty and present bias require different responses.
• Avoid moral labels: A missed action does not demonstrate a flawed character or lack of motivation.
• Preserve flexibility: Commitment devices can have costs when health, finances or circumstances change.
• Measure welfare: Higher uptake is not automatically better if the intervention imposes penalties or unwanted care.
• Check equity: A penalty or digital-only tool may disproportionately burden people with fewer resources.
• Test persistence: Short-term response to a prompt does not prove a durable change in behaviour or health.

Sources and further reading

The NBER study on self-control and preventive health examines commitment devices and cautions that they can reduce welfare in some settings. An original study of appointments and commitment devices for health behaviour compares mechanisms for increasing HIV-test uptake, while the economic review of commitment devices distinguishes different commitment designs. The beta–delta and appointment calculations above are original teaching illustrations, not estimated preferences or evidence for a particular programme.