The question
What does the evidence support—and not support—about how scarcity and constraint affect attention, cognition and behaviour?
Scarcity demands attention. When rent is due, food is running low or several urgent costs compete for the same money, those problems cannot simply be placed outside awareness. They require calculation, prioritisation and repeated decisions about what can be delayed, reduced or abandoned.
An influential research account proposes that these demands impose a tax on cognitive bandwidth. The mind becomes occupied with what is missing, leaving fewer cognitive resources available for other tasks. This account offered an important alternative to the claim that decisions made under poverty reflect inferior character or ability. It located at least part of the difficulty in the conditions people face.
But an explanation can be humane and plausible without every part of it being established.
The research does not support one simple conclusion in which scarcity uniformly reduces cognition. Findings differ according to what is scarce, how scarcity is measured, how long it lasts, which outcome is tested and whether the study examines association or causal change.
The central question is therefore:
What does the evidence support—and not support—about how scarcity and constraint affect attention, cognition and behaviour?
Scarcity is not one condition
Scarcity can refer to several different exposures.
Objective scarcity concerns measurable insufficiency: too little income, food, housing, time or access to necessary resources.
Perceived scarcity concerns the experience of having less than one needs relative to current demands. It may depend on income, debt, volatility, responsibilities, expected costs and available support.
The two can overlap without being identical. Income alone is an imperfect measure of financial scarcity. Two households with the same income may face different debts, medical costs, family responsibilities or access to secure housing.
At the same time, describing scarcity as a subjective experience must not reduce material hardship to mindset. A person may feel that resources are insufficient because they are insufficient.
Research also examines different timescales. An experiment might ask someone to imagine an expensive repair or give them too few turns in a game. Chronic material constraint involves repeated consequences that such experiments cannot recreate: unpaid bills, unstable housing, food insecurity, administrative demands, restricted opportunities and cumulative stress.
A hypothetical emergency may reveal how financial concern affects a particular task. It does not reproduce chronic poverty.
Nor should findings move automatically between scarcity domains. Most evidence reviewed here concerns financial scarcity. Time pressure, food insecurity and housing insecurity create different risks, demands and bodily consequences. They belong to the broader family of consequential constraint, but they cannot inherit financial-scarcity findings merely because they share the label "scarcity."
Scarcity is therefore a family of exposures, not one uniform treatment.
What is cognitive bandwidth?
"Cognitive bandwidth" is an umbrella term rather than one directly observed mental resource.
Scarcity studies have measured functions including:
- attention;
- working memory;
- inhibitory control;
- cognitive flexibility;
- reasoning;
- performance on executive tasks.
These functions are related, but they are not identical. A person can perform worse on one task without showing the same change on another.
Some cognitive tasks are designed to index abilities related to fluid reasoning or executive control. Their results are meaningful within the limits of the task. But observed performance remains sensitive to state and context. A lower score can reflect distraction, worry, fatigue, unfamiliarity, time pressure, stress or other features of the testing conditions. It does not by itself establish reduced intelligence or a stable loss of underlying capacity.
Bandwidth can be useful as a metaphor for competing cognitive demands. It becomes misleading when treated as though researchers have identified one mental reservoir that financial scarcity reliably drains.
Strong claims must keep three levels of evidence separate:
- Exposure: What changed or was compared?
- Outcome: What was measured?
- Mechanism: What process was proposed to connect them?
A financial concern can be manipulated and task performance can be measured. That does not mean cognitive load or attentional capture has been demonstrated unless the proposed process was itself tested.
Why the original findings mattered
Two related research programmes shaped the scarcity account, but they did not test identical things.
Anandi Mani and colleagues examined financial concern and cognitive performance. In one set of studies, lower- and higher-income participants considered financial problems of different severity before completing cognitive tasks. Difficult scenarios were associated with worse performance among lower-income participants but not among higher-income participants.
In a field study, Indian sugarcane farmers completed cognitive tasks before and after harvest. The same farmers performed worse before harvest, when money was scarcer, than afterwards.
These findings mattered because they suggested that measured performance could change within the same person as financial conditions changed. The proposed problem was not lower intelligence among people living in poverty. It was temporary interference from managing financial scarcity.
Yet the exposure and outcome did not directly establish the mechanism. Pre- and post-harvest periods may also differ in workload, stress, nutrition, sleep, physical demands, expectations and familiarity with the tests. The financial scenarios showed effects under specific experimental conditions, but imagining a difficult expense is not identical to sustained material insecurity.
A separate programme led by Anuj Shah and colleagues used resource-budget games. Participants with smaller budgets sometimes became more efficient on the immediate task, yet were also more likely to borrow from future rounds in ways that reduced their total payoff.
These experiments helped develop the idea of tunnelling: pressing needs become more prominent while information or consequences outside the immediate focus may receive less consideration.
The early studies established important hypotheses about financial concern, focal performance and neglect. They did not establish one universal mechanism covering every form of scarcity.
Replications did not produce one answer
Later studies have reported positive, null and contradictory results.
A US payday study compared low-income participants before and after receiving income. Participants clearly differed in immediately available financial resources, but researchers found no general evidence that cognitive performance was worse before payday.
Later analyses argued that the precise distance from payday may matter. This keeps timing open as a boundary condition, but it does not turn the initial null finding into broad confirmation.
A Colombian replication of Mani and colleagues' financial-scenario approach did not find that harsh scarcity scenarios impaired cognitive or executive performance among lower-socioeconomic-status participants.
Other studies have reported related seasonal or payment-linked performance differences in some settings and not others. These should be described as direct replications only when their procedures, exposures and outcomes match closely. Many are better understood as conceptual replications testing similar propositions under different conditions.
Cash-transfer and resource-change studies have likewise failed to show one consistent cognitive effect. Where improvements appeared, they were often substantially smaller than the largest early estimates.
A 2025 online study also reported that exposure to scarcity-related words did not differentially alter sustained attention. That finding does not show that consequential scarcity never captures attention. It shows that seeing scarcity-related cues is not equivalent to confronting an unmet demand with real consequences.
The replication record is therefore mixed and context-sensitive. Successful findings support narrower propositions in particular settings. Null results constrain their generality.
Two meta-analyses answer different questions
Recent syntheses make the importance of research design especially clear.
A 2024 meta-analysis combined 256 effect sizes from 29 datasets involving 111,852 respondents. Under the authors' broad model, financial scarcity showed a detrimental relationship with cognitive performance of approximately Hedges' g = −0.43.
The estimate associated with studies accounting for education was approximately g = −0.15—about 60% smaller. Severity, timing and design also contributed to differences among results, while substantial between-study variation remained.
These estimates do not identify education as the sole causal explanation. Education can reflect accumulated opportunity, environmental exposure, institutional access and skill development. It may operate as a confounder, moderator, mediator or part of a reciprocal relationship with income.
The education-adjusted estimate is therefore not the one "true" scarcity effect. It shows that the observed relationship is sensitive to major contextual differences. The synthesis also combines correlational and non-correlational evidence, so its broad estimate should not be interpreted as one causal effect of acute scarcity.
A 2025 systematic review and Bayesian meta-analysis asked a narrower question: when financial resources change, does cognitive performance change?
It examined 26 effect sizes from 14 studies with 30,341 pooled participants. The pooled estimate was small—approximately g = 0.06, with an interval from 0.00 to 0.18—and the Bayesian evidence was inconclusive. The review did not decisively establish either a general improvement or its absence. It also reported moderate support for no effect on memory.
If a broad average causal effect exists under the resource changes studied, its pooled estimate is far smaller than the largest original claims. This does not exclude larger effects in particular populations, settings or time windows.
The review itself had important limits:
- relatively few eligible studies;
- heterogeneous interventions and natural experiments;
- limited measurement of proposed mechanisms;
- uncertain subgroup effects;
- sparse long-term causal evidence;
- concurrent changes surrounding payday and harvest;
- non-preregistered review methods.
The two meta-analytic estimates should not be subtracted from one another or treated as rival estimates of the same parameter.
The 2024 analysis asks about a broad relationship across varied evidence. The 2025 synthesis focuses more narrowly on whether changing financial resources changes measured cognition. A broad association can coexist with a small or uncertain short-term causal effect.
A short-term increase in available cash is also not equivalent to removing chronic scarcity. Payment may leave debt, insecurity, accumulated stress and institutional barriers unchanged. Null findings after payday constrain claims about acute resource fluctuations more directly than claims about the cumulative consequences of long-term hardship.
The safest conclusion is:
Association is clearer than a general short-term causal bandwidth effect.
Focus can change without cognition uniformly declining
Evidence is more consistent with context-specific changes in focus and task performance than with uniform cognitive depletion.
Financial scarcity can make prices, deadlines, debts and immediate trade-offs more salient. In some controlled tasks, participants with fewer resources used those resources more efficiently or remembered scarcity-relevant information more accurately.
This has been described as a focus dividend. It is not a general benefit of scarcity. It appeared in particular tasks under studied conditions.
The same focal orientation may coincide with neglect of information outside the immediate problem. Future costs, peripheral tasks or emerging demands can receive less consideration. But "redistribution of attention" should be reserved for studies that directly measured or validly operationalised attention. Borrowing behaviour or lower task performance alone does not demonstrate how attention was allocated.
A critical review found stronger laboratory support for scarcity-related focus and resource efficiency than for a general bandwidth reduction. Evidence connecting focal attention to real-world neglect and overborrowing remained weaker.
The supported conclusion is narrow: pressing financial demands can alter focus and performance under some conditions. They do not uniformly lower cognition.
Several pathways can produce the same result
Suppose a financially constrained participant performs worse on a working-memory task.
The exposure is financial constraint. The measured outcome is task performance. The mechanism remains open unless it is directly tested.
Possible explanations include:
- attention occupied by unresolved expenses;
- concurrent calculation of trade-offs;
- stress;
- sleep disruption;
- pain or poor health;
- nutritional conditions;
- unfamiliarity with the task;
- educational opportunity;
- environmental instability;
- reduced motivation or increased urgency;
- a combination of these factors.
Many scarcity studies infer cognitive load without measuring it directly.
Stress is especially important. Material constraint can create uncertainty, threat, conflict and repeated administrative demands. Over time, such demands may contribute to cumulative physiological burden. But scarcity and stress are not interchangeable, and worry does not affect every task in the same direction.
One experiment found that exposure to worrisome topics improved incentivised performance under some conditions. Its relevance is not that worry generally improves cognition. It demonstrates that effects can vary with task, motivation and the person's ability to respond.
Debt relief illustrates similar causal complexity. A Singaporean programme was associated with improved psychological functioning and changes in some decision measures. But relief changed more than available money: it could alter creditor demands, administrative burden, uncertainty, perceived control and social pressure. Evidence that cognitive control mediated the decision changes was weak.
These findings support the importance of material and administrative conditions. They do not isolate restored cognitive bandwidth as the mechanism.
Behaviour can change without cognitive impairment
Scarcity can change behaviour by changing prices, risks, penalties and available options even when no cognitive process has been impaired.
A person with stable resources may be able to wait for a larger future reward. Someone who needs food, transport or medicine today faces a different cost of waiting. Borrowing at an unfavourable rate can worsen future finances while protecting an immediate necessity. A short planning horizon may reflect an unstable environment in which distant opportunities are unreliable.
None of this means every decision under financial scarcity is optimal. "Adaptation" means a context-sensitive response, not that the response is harmless, freely chosen or best in the long term.
Scarcity can contribute to decisions with damaging consequences. But those decisions must be evaluated relative to the options, risks and opportunity costs actually present.
Tighter budgets may also increase attention to prices and trade-offs. Research provides some support for this pattern, although many studies rely on hypothetical purchases and imperfect scarcity measures.
A constrained adaptation is not the same as a stable personal deficit. Narrow options can produce costly behaviour without defective reasoning. Institutional rules and penalties can amplify hardship without first reducing cognitive capacity.
What the evidence can safely support
Supported
- Financial scarcity creates real practical and psychological demands.
- Pressing unmet demands can alter focus.
- Some controlled tasks show greater efficiency or memory for focal scarcity-related information.
- Chronic disadvantage is associated with differences in cognitive performance.
- Material, environmental and institutional restrictions help shape behaviour.
- Decisions under constraint must be evaluated relative to available options and risks.
Provisionally supported
- Financial scarcity may produce small, context-dependent changes in cognitive performance.
- Financial concern may interfere with unrelated tasks in some settings.
- Focal concentration may coincide with neglect outside the immediate demand.
- Stress, worry, sleep, health and concurrent problem-solving may contribute to observed differences.
- Resource changes may affect some subgroups more strongly than the population average.
Mixed or contested
- Whether acute financial scarcity reliably impairs cognitive performance.
- Whether increasing financial resources generally improves cognition.
- Whether cognitive load is the primary mechanism.
- Whether experimentally observed tunnelling explains real-world overborrowing.
- Whether different scarcity domains share one psychological process.
- Whether "cognitive bandwidth" identifies one measurable capacity.
Not supported
- Poverty automatically lowers intelligence.
- People under constraint are universally irrational.
- Scarcity drains a fixed quantity of mental capacity.
- A brief manipulation reproduces chronic poverty.
- A change in mindset removes material scarcity.
- One cognitive task measures a person's general ability.
- One theory fully explains behaviour under constraint.
What remains defensible
Financial scarcity and consequential constraint can alter what receives attention, which decisions become urgent and how demanding action becomes. Cognitive-performance changes have been observed, but their magnitude, mechanism and generality remain unsettled.
Evidence is stronger for context-sensitive differences in focus, task performance and prioritisation than for a large, universal reduction in cognitive capacity. Associations between financial scarcity and cognition are clearer than evidence that short-term resource increases reliably restore performance. Proposed pathways—including attention, cognitive load, stress, health and environmental restriction—are often insufficiently separated.
"Cognitive bandwidth" remains useful as an organising metaphor if it directs attention to the demands competing within constrained circumstances. It becomes misleading when transformed into a fixed mental resource that poverty automatically depletes.
The most defensible conclusion is therefore bounded:
Scarcity can change what receives attention, which trade-offs become unavoidable and what actions are realistically available. Its cognitive effects are heterogeneous, its mechanisms remain partly unresolved, and constrained behaviour must not be converted into evidence of inferior intelligence or character.