Psychology by BogdanovaThe Encyclopedia of Psychology for the Artificial Era

RESEARCH DOSSIER

Wealth and Risk Perception

Abstract. Wealth shapes the consequences of adverse events, the resources available for recovery, and the circumstances in which financial threats are assessed. This monograph examines the relationship between wealth and risk perception through a critical narrative synthesis of research on domain-specific risk attitudes, affect, financial threat, liquidity, socioeconomic expectations, relative wealth, portfolio choices and resource shocks. It distinguishes perceived probability, severity, uncertainty, perceived control, emotional response, preference and risk-bearing capacity. Evidence from surveys, experiments, linked portfolio data and lottery-based research demonstrates the need to distinguish resource variation from differences in beliefs and observed behavior. The available evidence does not identify a universal directional effect of wealth on subjective financial risk. The monograph develops an analytical synthesis of measurement and identification problems and specifies implications for research on cross-border wealth. Offshore asset ownership and distributional evidence cannot identify psychological motives without direct measurement. The work is a bounded narrative research synthesis; no original dataset, pooled causal effect, or new theoretical system is claimed.

Keywords: wealth; financial risk perception; perceived financial threat; economic insecurity; liquidity; relative wealth; risk-bearing capacity; perceived control; financial expectations; cross-border wealth

Scientific object and scope

Wealth alters the conditions under which people encounter uncertainty. A household with readily accessible savings, a business owner whose fortune is concentrated in one enterprise, and a beneficiary whose assets are administered by another person may possess comparable net wealth yet confront radically different consequences from an adverse event. Their assessments of danger depend on the resources available to them, the durability of their claims on those resources, the obligations they must satisfy, and their expectations about the future. Wealth is therefore a psychological context of risk appraisal as well as an economic stock.

This dossier studies the relationship between economic resources and subjective assessments of financial threats. It distinguishes the perceived probability of an adverse outcome, its perceived severity, the uncertainty surrounding its probability, the perceived ability to mitigate its consequences, and the emotional response associated with anticipation of loss. It examines absolute and relative resources, resource composition, liquidity, financial commitments, expectations, perceived control, economic insecurity, and investment experience. Risk-taking behavior is relevant because it may express such assessments, although its interpretation requires information about preferences, opportunity sets, and constraints.

The scientific question is whether and through which mechanisms wealth changes the perceived significance of a risky outcome. A subsidiary question concerns the extent to which existing empirical designs can separate perception from other wealth-dependent determinants of choice. The unit of explanation is an individual or household situated within an institutional environment. The dossier does not classify wealthy people as a psychological type, and it does not treat a country's financial arrangements as evidence of the motives of every person using them. Its connection to the Global Offshore System concerns the appraisal of wealth exposed to different institutions and jurisdictions, a topic requiring evidence particular to those settings.

Method and evidence boundaries

The method is a critical narrative synthesis organized by research question and inferential design. Sources include psychometric studies, behavioral surveys, experiments, quasi-experimental wealth variation, institutional portfolio records, and distributional research. The analytical contribution is to identify where these designs measure the same construct, where they support different explanations, and how those differences constrain conclusions about wealth-related risk perception. A source is given weight according to the claim at issue rather than its prestige alone: a well-designed lottery study may establish an effect on asset allocation without identifying subjective probabilities.

Searches targeted work on domain-specific risk attitudes, affect and financial threat, wealth and background risk, liquidity and payday, relative wealth and control, socioeconomic status and expectations, millionaire investment surveys, lottery windfalls, poverty and nonlinear risk-taking, international preferences, and offshore inequality. The evidence base is purposive. It does not support a claim of systematic-review completeness, pooled effect size, publication-bias adjustment, or formal meta-analysis. The date of evidence inspection is November 18, 2026. The source register records the actual access level for each item. Several journal articles were available through author or institutional abstracts; findings from those records are restricted to the reported scope. Theoretical claims and illustrations are identified as such, and the working papers retain their provisional status.

The dossier uses three inferential distinctions consistently. Descriptive evidence establishes an association within observed persons or settings. Experimental or quasi-experimental evidence can identify an effect of a particular assigned treatment under its design assumptions. Interpretation connects such results to psychological processes, but cannot transform a measured choice into a directly observed belief. This distinction is essential for cross-disciplinary evidence: economic data often measure behavior, whereas the psychological explanation may concern appraisals that those data do not contain.

Wealth as a multidimensional exposure

Net wealth equals the value of assets minus liabilities at a specified date. It differs from current income, disposable income, consumption, and access to credit. Gross assets and gross liabilities also matter independently because contractual deadlines, refinancing conditions, and the differing liquidity of assets can create vulnerabilities obscured by a net figure. Wealth has a temporal dimension: a household may expect a secure income stream while possessing little wealth today, or may possess valuable property while anticipating a sharp income decline.

An economically meaningful measurement should distinguish financial assets, real assets, business equity, pension entitlements, liabilities, and liquid resources. The relevant boundary depends on the risky event. A pension account may offer strong protection at retirement but limited protection against an immediate cash obligation. A privately held enterprise may provide substantial expected income and yet be difficult to sell in a downturn. Legal title and effective access can differ where assets are jointly owned or held through fiduciary arrangements. These differences create heterogeneous psychological conditions even at identical net worth.

The implications are methodological as well as economic. A regression using only income categories cannot establish a wealth effect. An aggregate measure of assets can obscure the distinction between accessible funds and locked assets. A question about perceived financial security captures an appraisal that incorporates beliefs, obligations, and comparison standards, rather than a direct measurement of net wealth. Combining objective accounts with subjective measures permits examination of discrepancies instead of defining one measure as the other.

Consider two illustrative households, each with net assets of $500,000. One has a liquid diversified portfolio and modest fixed obligations; the other has equity in a business, secured debt, and high unavoidable operating costs. An unexpected $20,000 payment may threaten the second household's ability to keep its enterprise functioning while requiring only a portfolio withdrawal for the first. The identical nominal event has distinct practical implications. Any difference in its perceived danger can be reasonable given these circumstances. The illustration does not imply that one household necessarily feels greater anxiety; the psychological outcome requires measurement.

Perceived probability, severity, and uncertainty

Perceived probability is a person's judgment of how likely a specified event is within a specified period. Perceived severity concerns the magnitude or significance of its consequences conditional on occurrence. Subjective uncertainty concerns how confident the person is in the probability estimate or in the definition of the underlying event. A person may believe that a particular loss has a small probability but remain uncertain whether that probability can be estimated at all. These appraisals can move independently.

A meaningful measurement fixes the event and time horizon. Asking whether investing is risky leaves open the type of investment, size of exposure, possible outcomes, time frame, liquidity need, and meaning of risk. A better elicitation asks separately about the perceived chance of a specified decline, the perceived impact of such a decline on obligations, and confidence in the estimate. These answers can be compared with externally documented exposures where available. Numerical probabilities require attention to comprehension, rounding, and the interpretation of percentages.

The domain-specific study by Weber et al. (2002), measuring risk perceptions and behaviors, provides a basis for separating appraised danger and benefits from declared willingness to take risk. Its empirical setting is a student sample and its implications for wealthy households require independent validation. Nevertheless, its measurement distinction is general: a person who assigns a relatively low risk to an activity need not possess unusually high tolerance for a given perceived threat.

Financial dangers may also be mutually dependent. The loss of employment, fall in asset prices, need for emergency borrowing, and impairment of collateral can occur together. If an appraisal questionnaire isolates a single event without mentioning background conditions, differences in responses may reflect differing assumptions about those conditions. Wealth may influence those assumptions. A financial-risk survey should therefore specify whether a threat is considered alone or as part of a broader economic shock.

Risk-bearing capacity and the scale of loss

Risk-bearing capacity is the ability to absorb adverse consequences while maintaining essential obligations and valued future options. It is partly observable through accessible reserves, debt service, contractual commitments, income stability, and available support. Its psychological counterpart is a person's estimate of that ability. An increase in liquid resources can improve actual capacity while leaving the probability of an external shock unchanged. It may change the perceived consequences of the shock without changing its perceived probability.

The distinction resolves a frequent ambiguity. A wealthy investor may accept a larger nominal loss than an investor with little savings because the loss would leave more resources available for essential needs. The same behavior might result from a difference in preferences, perceived financial capacity, or subjective belief about the probability of the loss. A choice between investments cannot by itself distinguish these explanations.

Guiso and Paiella (2008), in Risk aversion, wealth, and background risk, report household associations between risk aversion, uncertain income, and indicators of liquidity constraints. Their evidence supports the importance of background conditions when interpreting willingness to bear risk. It does not establish that income uncertainty directly changes perceived event probabilities. The difference matters: reducing exposure to a second correlated risk can alter an optimal position even if beliefs concerning the first risk remain constant.

Loss magnitude may be expressed in dollars, proportion of net wealth, months of essential expenditure, or impairment of a specific goal. The denominator has substantive meaning. A fixed monetary loss compares different relative burdens across persons; a fixed percentage of wealth compares different absolute amounts; a fixed number of months' expenses adjusts for commitments but may conceal unequal lifetime assets. A research design should choose the denominator that matches the psychological question and test whether conclusions survive alternative definitions.

Liquidity, solvency, and anticipation

Liquidity is the ability to mobilize resources promptly at tolerable cost. Solvency is the ability to cover obligations through assets and expected cash flows under the relevant accounting or legal definition. Anticipated illiquidity can make an otherwise manageable expected loss feel threatening because the moment of payment is decisive. The psychological significance of a risk thus involves the sequence of cash needs and the timing of access, not just the expected balance at a distant date.

Carvalho et al. (2016), in their payday study of financial resources, assigned low-income participants to survey timing around payday. Their reported results distinguish changes in monetary intertemporal decisions from a lack of detected change in risk-taking. This is valuable precisely because it limits broad claims: short-run resource fluctuation in that design did not establish a universal change in willingness to accept risk.

The experiment concerns a specific liquidity setting and time horizon. Payday is normally anticipated; some obligations may also fall close to payday. A temporary resource fluctuation differs from a permanent wealth increase, an unanticipated emergency, or a loss of access to previously owned assets. It would be inappropriate to generalize the result into a statement that financial scarcity never affects subjective threat. A study of longer-term insecurity needs a different treatment, different outcomes, and sufficient observation time.

Anticipation creates another empirical difficulty. If a person expects a future payment, the current account balance may poorly represent perceived resources. If a person expects declining income, a generous current balance may coexist with high threat. Repeated measurement before and after known and unexpected events can help establish whether changes in appraisal precede changes in recorded wealth, follow them, or track an external shock affecting both.

Affect and financial threat

Risk appraisal includes cognitive evaluation and affective anticipation. Loewenstein et al. (2001) developed the risk-as-feelings account, explaining why immediate emotional responses can diverge from cognitive evaluations of future events. The account offers a conceptual interpretation of financial decisions in which an investor understands a probability statement yet experiences a strong emotional reaction to a possible loss.

Financial threat is a related but more specific construct. Marjanovic et al. (2013), in their Financial Threat Scale evaluation, report psychometric evidence concerning a short measure of perceived financial threat and its association with psychological health. A threat scale captures experienced apprehension and insecurity; it does not substitute for a calibrated judgment of the probability of unemployment, insolvency, or an asset-price fall.

Wealth can affect threat through material protection, the anticipation of responsibilities, uncertainty about future access, or perceived consequences for a valued way of life. Each is an empirical possibility rather than an automatic effect. An increase in wealth may coincide with acquisition of additional liabilities or responsibilities, including employees, dependents, and long-term commitments. A growing asset stock does not guarantee a reduction in apprehension if the perceived gap between resources and commitments remains large.

The direction of causation also deserves attention. People experiencing persistent worry may adjust their financial behavior, hold different amounts of liquid savings, seek advice, or avoid uncertain investments. Their resulting balance sheets can then reflect prior psychological conditions. A correlation between financial threat and wealth thus admits both resource-to-appraisal and appraisal-to-resource pathways. Longitudinal measurements can assess temporal ordering, while causal identification requires a design that addresses common causes and selection.

Perceived control and effective agency

Perceived control concerns the extent to which a person believes that their actions can influence a situation or its consequences. In financial settings it includes the ability to meet obligations, diversify exposure, obtain information, access funds, negotiate contractual terms, and recover from loss. Actual control depends on enforceable rights, available alternatives, skills, and institutional conditions. A person's self-assessment may be accurate, pessimistic, or overconfident.

An increase in wealth may expand the attainable set of actions. Whether the person recognizes those opportunities depends on information, expertise, legal access, or the delegation of management. A beneficiary who lacks decision rights may possess a high beneficial interest while experiencing limited control over the timing of distributions. Conversely, a household with modest wealth but dependable contractual rights and clear access to funds may perceive substantial control over a defined emergency.

Fehr and Reichlin (2025), in Status, Control, and Risk, report experimental evidence concerning relative wealth information, perceived control, and incentivized risk-taking. Their working paper addresses risk-taking under a manipulated rank-related information condition, not the causal effect of changing actual wealth. It suggests that the perception of position can matter even when the immediate material resources involved in a task do not change.

The distinction between control over occurrence and control over consequences is important. An investor usually cannot prevent a market-wide decline but may alter exposure, choose a liquidity reserve, or revise spending. Confusing these forms of control can yield inappropriate measurement. Researchers should ask whether the participant expects to influence the event itself, the level of exposure, or their ability to recover. Each is a different psychological assessment.

Relative resources and social comparison

Economic position has both absolute and comparative dimensions. A person can compare their resources with neighbors, coworkers, family members, professional peers, a nationally defined percentile, or their own past circumstances. These comparison groups need not coincide. Population rank measured by a statistical agency is a different variable from perceived rank within a person's salient social environment.

Relative wealth can influence the meaning of a potential loss by altering perceived standing, future opportunities, or expectations regarding ordinary commitments. A researcher examining this possibility should separate observed position, perceived position, perceived mobility, and the personally relevant comparison target. A measure asking whether someone feels poorer than others cannot determine whether they possess fewer assets in an objective distribution.

The experimental work of Fehr and Reichlin (2025) on relative wealth perceptions offers a relevant identification strategy: manipulating information concerning rank can isolate a psychological component of perceived position within that experiment. Its inference should remain attached to the treatment and outcome. Rankings may change beliefs about social status without changing material capacity; greater willingness to take risk in that task is not direct evidence that the participant estimates a lower probability of loss.

A comparison-based appraisal may also be asymmetric. Losing a rank position and failing to attain an expected rank need not be experienced in the same way. A longitudinal design could ask participants to record their perceived economic position and expectations before an external change, then measure event probabilities, severity, control, and behavior separately afterward. This would provide a more informative test than deriving social comparison from consumption or wealth records alone.

Expectations and the interpretation of information

Perceived financial risk is prospective. Judgments about future employment, inflation, economic expansion, returns, and household income influence whether a potential loss appears tolerable and whether protection appears necessary. Expectations can differ with economic experiences, information access, education, and the institutions in which people participate. The study of wealth-related risk appraisal therefore overlaps with the study of economic beliefs.

Das et al. (2020), in Socioeconomic status and macroeconomic expectations, report systematic differences in macroeconomic optimism associated with income and education and variation in these differences across economic conditions. These are measures of socioeconomic status, not direct measurements of net wealth. The result is evidence that economic expectations vary with social and economic position; its application to the wealth distribution requires independently measured assets and liabilities.

Kuhnen and Miu (2017), investigating learning from financial information, report socioeconomic differences in financial beliefs and investment decisions across their studied settings. Learning is particularly relevant to perceived risk because two individuals observing the same sequence of returns may interpret its significance differently. Evidence concerning socioeconomic gradients does not alone establish that increasing a person's wealth would change the way they process information. Education, prior experience, opportunities to participate in markets, and background conditions are plausible related variables.

An important measurement consequence is that researchers should elicit expectations before presenting common information, then record updated expectations and confidence. A comparison of only final responses combines prior beliefs with updating. A comparison of initial and revised estimates under a shared information treatment can separate them more credibly, provided participants understand the event and the probability scale. Researchers should also distinguish expected return, probability of a loss, expected loss conditional on its occurrence, and perceived ambiguity.

Risk preferences and domain specificity

Risk preference expresses willingness to accept uncertain alternatives given beliefs and consequences. It cannot be read directly from an person's wealth position or inferred simply from how much money they hold in risky assets. Models of choice formalize preferences over probability distributions, but estimates depend on assumptions concerning beliefs, consequences, background resources, and the decision environment.

Dohmen et al. (2011), in Individual risk attitudes, examined a general survey measure of willingness to take risk together with behavioral validation and domain-specific variation. Such a measure can be valuable in wealth research because it provides a distinct outcome or covariate. Its validity as a measure of willingness does not make it a direct measure of perceived probability or financial threat.

Weber et al. (2002) show in their domain-specific risk-attitude scale why a person's behavior may vary across domains with differing perceptions of danger and benefit. Financial, health, recreational, and social decisions pose different consequences. A general risk-willingness measure can therefore be combined with domain-specific assessments rather than used to explain all risky behavior through one invariant trait.

A conceptual example clarifies identification. Two people choose between a certain payment and an objectively specified lottery. Their choices may differ because they evaluate the consequences differently, distrust the stated probabilities, use different reference amounts, or possess unequal capacity to bear a loss. Even objectively assigned probabilities need not be accepted as credible by respondents. An experiment seeking preference parameters should check probability comprehension and credibility; an experiment seeking perceptions must measure them directly.

Background risks and correlated exposures

The portfolio of risks confronting a household includes income, health, housing, family obligations, business liabilities, and institutional commitments. Risk exposure is rarely a set of independent events. A person who depends on a highly cyclical employer may simultaneously hold employer-related stock, possess real estate in the same region, and depend on local credit availability. A downturn can damage all of these resources.

Wealth may increase diversification opportunities, but composition and concentration remain decisive. A successful business owner can have a high net worth with substantial exposure to a single business. A salaried household with modest assets may face comparatively diversified income and savings. The economic relationship between net wealth and total background risk is therefore contingent on composition. Guiso and Paiella's (2008) analysis of background risk provides empirical motivation for incorporating other uncertainties in the interpretation of risk attitudes.

For psychology, perceived correlations are themselves an object of study. A participant may underestimate the possibility of joint adverse events, or may treat several reasonably independent dangers as a single looming crisis. Measuring isolated probabilities and the anticipated simultaneous occurrence of shocks can reveal whether perceived vulnerability originates partly in assessments of dependence. The resulting measure should be interpreted alongside actual constraints and exposures; the absence of precise external probabilities does not justify assigning an arbitrary benchmark.

A practical experimental vignette can hold wealth fixed while varying whether an additional loss is independent of or correlated with an income shock. A complementary design can hold correlation fixed while varying accessible liquidity. Separate probability, severity, and control questions allow the researcher to identify which aspect of appraisal responds to each manipulation. These are research proposals, not reported experimental results.

Accumulated wealth and investment behavior

Observed investment portfolios represent decisions made under contractual, tax, institutional, informational, and personal constraints. A portfolio is a realized allocation, potentially influenced by defaults, inertia, advisory arrangements, and prior market returns. Treating the proportion of risky assets as a direct scale of perceived financial danger discards this complexity.

Giglio et al. (2021), in Five facts about beliefs and portfolios, link investor surveys with administrative portfolio information. Their reported evidence includes substantial variation in beliefs and relatively modest average sensitivity of portfolio allocations to elicited beliefs. This is especially relevant because beliefs and behavior are observed separately. Their relation is empirically informative, but its magnitude is not a universal coefficient across all investor populations.

The study by Bender et al. (2022), Millionaires speak, examines 2,484 Americans with at least one million dollars in investable assets. Its survey findings illuminate stated investment beliefs, constraints, and reasons for investment decisions in an affluent population. The sample definition is investable assets, not the full population of people whose net worth exceeds a million dollars. People whose wealth is primarily tied up in housing or unlisted businesses may face materially different decision settings.

These studies help establish an important boundary for the interpretation of wealthy investors' behavior. A person may report substantial concern about a market decline yet retain equity exposure because of long-term goals, tax consequences, or advice. Another may report confidence in returns yet hold substantial cash because of an imminent commitment. Measuring both the perceived danger and the reason for the allocation is more informative than placing investors on a single risk-tolerance continuum.

Wealth shocks and causal inference

A causal effect of wealth requires variation that separates wealth change from prior preferences, economic opportunities, and personal circumstances. Ordinary cross-sectional comparisons rarely achieve this. Individuals who accumulated wealth through risky entrepreneurship are selected partly through past decisions and outcomes. Comparing them with those who did not undertake such activities may capture differences that existed before wealth accumulation.

Lottery prizes offer a particularly useful source of externally induced variation. Briggs et al. (2023), in Financial Windfalls, Portfolio Allocations, and Risk Preferences, use Swedish lottery data and report a decrease in the risky share of portfolios after financial windfalls. Their working paper discusses interpretations under alternative assumptions about risk preferences. The result addresses the composition of investments after a particular change in resources. It does not directly establish that recipients became more fearful of adverse investment outcomes.

The denominator of a risky asset share can also change immediately following a windfall if the resources initially arrive as cash. Subsequent purchases, debt repayment, transfers, consumption, and financial planning then change the observed portfolio. Researchers need an observation horizon appropriate to the proposed mechanism. If the psychological hypothesis concerns lasting changes in perceived security, short-lived portfolio adjustments provide incomplete evidence.

There is a general lesson for designs using wealth shocks. The exclusion restriction for an instrumental variable may fail for a psychological outcome if the event affects that outcome through channels other than resources. An inheritance commonly involves bereavement; a promotion changes work responsibilities and perceived status; a windfall changes expectations as well as the balance sheet. A lottery design can improve causal leverage on the effect of a prize, but a mediation claim about perceived security still requires measurements and assumptions beyond assignment of prizes.

Economic insecurity and nonlinear responses

A monotonic relationship between wealth and risk-taking is not implied by the distinctions developed above. Greater resources can reduce the severity of a loss while reducing the attractiveness of an additional gain that would make a critical difference to a low-resource household. Different choices can result from changing consequences even if preferences and objective probabilities remain fixed.

De Courson et al. (2025) report poverty-related risk avoidance and risk-taking in an observational preregistered survey of 472 adults in the United Kingdom and France. The work examines hypothetical gambles and provides model-dependent evidence concerning nonlinear relationships with subjective resources. The design does not establish that material deprivation caused risk-taking, and its nonrepresentative sample and the distinction between subjective and objective resource measures constrain generalization. The later theoretical treatment by de Courson et al. (2026), the desperation threshold model, develops an explanation for apparently divergent responses to resource insufficiency.

The proposed threshold logic concerns the consequences of outcomes relative to a resource requirement. An individual close to meeting a critical obligation has a reason to avoid a loss that would place them below it. Someone currently unable to meet the obligation may find a risky prospect attractive if a favorable outcome can restore sufficiency. This reasoning concerns the structure of outcomes. It does not require either person to misperceive the probability, and it does not prove that subjective threat increases or decreases in a particular form.

A psychological test must therefore measure the subjective assessment of the relevant requirement and the feasibility of alternatives alongside choice. The threshold itself can vary with commitments; it can be temporary, domain-specific, or socially defined. Assigning everyone the same income cutoff could conceal the very mechanism the explanation addresses. Claims of a universal threshold or a universal U-shaped wealth effect exceed the evidence examined in this dossier.

Time horizon, loss, and recovery

Risk appraisal depends on the duration over which an event threatens valued outcomes. A temporary decline in the market value of assets is different from a permanent impairment of future consumption, the forced sale of an asset, or an inability to meet an immediate payment. An investor whose expected recovery horizon is long may regard a short-term decline as tolerable. A person approaching a binding deadline may perceive the same event as threatening.

Wealth can increase the range of available recovery options, including the capacity to wait, obtain advice, or renegotiate a commitment. Such options may remain highly contingent on liquidity and legal rights. A person may possess assets whose long-term value is substantial but cannot readily realize that value without discount or delay. Evaluating a threat requires both estimated adverse outcomes and estimated recovery possibilities.

A study of wealth and risk perception should explicitly distinguish current appraisal from expectations of adaptation. The question “How damaging would this loss be immediately?” differs from “How damaging would it be after one year?” and from “How likely are you to regain your prior position?” Each may have independent predictive relevance for financial behavior.

Temporal framing also creates potential reporting inconsistencies. Specifying a percentage decline without a starting value can conceal what amount respondents imagine; asking about future recovery without specifying whether new earnings can contribute leaves different participants with different assumptions. Carefully designed vignettes should define the event, the relevant resource baseline, and the recovery window before soliciting judgment.

Wealth, expertise, and delegation

Wealth creates opportunities to purchase specialized information and management services. Expertise may improve appraisal by enabling a clearer understanding of exposures, probability statements, asset composition, and contractual terms. Delegation can also produce an epistemic division of labor: an individual may rely on an adviser, trustee, or manager to monitor threats and make decisions. This arrangement complicates measurement of whose perceptions determine a given allocation.

The affluent-investor survey by Bender et al. (2022), personal investment decisions of millionaires, addresses stated reasons for investment decisions and the relevance of advisory relationships. It supports treating advice as a substantive component of the decision environment. It does not show that delegation always improves calibrated risk perception or that advisers uniformly cause a reduction in perceived threat.

There are several plausible psychological responses to delegation. Confidence in a trusted institution may reduce apprehension; incomplete understanding of the institution's decisions may increase uncertainty; perceived responsibility may shift between the manager and asset owner. These are hypotheses requiring evidence. A researcher should distinguish trust in the adviser's competence, trust in the adviser's alignment with the client's interests, and the perceived ability to verify decisions.

The choice recorded in a portfolio database may be the recommendation of a professional implemented with the client's consent. If the client and adviser have different beliefs, a survey of only the client cannot explain the complete decision process. Where feasible, dyadic study designs can elicit both sets of assessments and distinguish recommended, accepted, and executed decisions.

Wealth preservation and the meaning of adverse outcomes

Financial threats are evaluated relative to goals. Wealth used to support near-term consumption, preserve a business, finance retirement, care for relatives, or transfer resources across generations carries different consequences if impaired. The same monetary reduction may be experienced as temporary inconvenience, loss of autonomy, disruption of commitments, or damage to a valued long-term project.

As wealth grows, goals and obligations can change. This means an observed reduction in perceived severity per dollar need not translate into a proportionate reduction in overall perceived financial threat. A larger asset base may be accompanied by more extensive commitments and more complex management. Alternatively, a person can retain modest commitments despite increasing wealth and consequently experience greater security. The empirical object is the relation between resources and goals, not wealth alone.

Risk perception research needs to avoid treating preservation motives as equivalent to risk aversion. An investor seeking to preserve the real purchasing power of assets over decades may accept substantial nominal fluctuations. A person who prioritizes an imminent payment may prefer a stable nominal value despite other risks. Under varying time horizons these objectives can imply different strategies without requiring different perceived probabilities.

Research should record both the stated goal and its objective financial relevance. Questions about intended use, expected horizon, acceptable impairment, and available substitutes make the meaning of loss more identifiable. They also facilitate comparisons across populations without imposing a single social or moral interpretation on wealth accumulation.

Institutional context and cross-border wealth

Risk appraisal of cross-border assets may involve currency fluctuations, political and regulatory changes, market liquidity, contractual enforcement, financial-intermediary reliability, communication, and access to information. The significance of each depends on the institutions involved, the owner's legal position, the asset's composition, and the nature of the obligations. These domains deserve separate treatment in research concerning the Global Offshore System.

Offshore ownership is an institutional and legal arrangement, not a directly observed psychological variable. Individuals with similar arrangements may have very different expectations and purposes; conversely, people facing similar perceived threats may use different arrangements. An investigation of institutional-risk perception should distinguish perceived probability of an adverse institutional event from perceived ability to respond. It should control, where feasible, for actual contractual and jurisdictional differences.

Alstadsæter et al. (2019), in their study Tax evasion and inequality, link offshore-related records with Scandinavian wealth and tax information to estimate the distribution of tax evasion. The work makes the distributional importance of otherwise difficult-to-observe offshore arrangements empirically relevant. Its subject is tax evasion in particular settings and its data do not identify subjective financial risk perceptions or general motives among offshore asset owners. Its findings cannot be applied to offshore wealth as a whole or to any individual solely because they own cross-border assets.

An offshore-specific psychological study would require lawful recruitment, informed consent, credible confidentiality protections, and measurements that distinguish objective institutional exposure from reported fear or confidence. A comparison of domestic and cross-border asset holders would also need to address selection into ownership, differences in wealth composition, and access to specialized advice. The present literature supports this research agenda but does not establish a unique offshore psychological profile.

Measurement across countries and populations

Cross-national comparisons are complicated by differences in economic security, financial institutions, language, social expectations, and access to services. Equal nominal wealth across countries need not represent equal purchasing capacity or equal ability to meet foreseeable obligations. Relative rank and subjective comparisons may also vary independently of adjusted absolute wealth.

Falk et al. (2018), in Global evidence on economic preferences, use standardized measures in approximately 80,000 respondents across 76 countries and report substantial heterogeneity in economic preferences. Their work demonstrates the feasibility of large-scale comparable elicitation while providing a reminder that preference measures, even when internationally standardized, are not direct measures of risk appraisal.

Cross-cultural studies of wealth-related perception should establish measurement equivalence. A translated question concerning financial security or control may evoke different institutional or personal assumptions. Multi-group analysis can help identify whether items display comparable relationships with the construct and comparable thresholds or intercepts; meaningful comparisons of means require an appropriate measurement model. Numerical probabilities also require comparable comprehension.

The wealth distribution sampled matters. Representative surveys frequently contain relatively few respondents from the extreme upper tail; specialized high-net-worth samples may have nonresponse and recruitment limitations and underrepresent people outside particular intermediaries. Oversampling high-wealth groups, carefully documenting recruitment, and independently evaluating sample coverage can improve estimates, but weighting cannot recover a group absent from the sampling frame.

Research designs for separating appraisal and choice

A credible wealth-perception study should distinguish at least three types of outcome: direct judgments about risk, emotional or threat responses, and behavior under uncertainty. Direct judgments can use defined events and numerical or ordinal scales. Emotional measures can capture financial apprehension under a specified scenario. Behavioral tasks can examine choices whose available options and consequences are documented. Linking these measures makes it possible to test rather than assume their relationship.

One design combines a financial questionnaire with randomized hypothetical scenarios varying liquid reserves, debt obligations, and the probability of loss. The scenarios can isolate perceived severity from willingness to accept a prospect, while keeping the actual resources of participants unchanged. The resulting causal estimate concerns the scenario manipulation. A separate natural experiment or longitudinal design is needed to establish a change arising from actual wealth.

Another design observes participants before and after an externally induced resource change, with repeated assessments of probability, severity, perceived control, and affect. A comparison group and prespecified outcomes are needed to distinguish the event's effect from simultaneous market changes or secular trends. If the resource change is assigned, noncompliance and missing follow-up still require attention; if it is observational, selection and confounding require stronger assumptions.

A third design links elicited beliefs to subsequent verified decisions, in keeping with the logic of Giglio et al. (2021), matched beliefs and portfolio evidence. Recording decision constraints, delegated management, and current obligations helps explain why stated perception may only weakly predict behavior. The temporal alignment between elicitation and recorded choice is crucial because both beliefs and available opportunities may change.

Identification, measurement error, and analytical checks

The empirical association between wealth and perceived risk may be shaped by age, education, prior financial experiences, family commitments, employment, business ownership, health, macroeconomic conditions, and institutions. The correct adjustment set depends on the estimand. Some variables can be common causes of wealth and appraisal, whereas others can lie on the causal pathway from wealth to appraisal. Conditioning automatically on all observed variables can remove part of a total effect or introduce bias.

A simplified causal question is the effect of accessible resources on perceived severity of a fixed loss. If liquidity is increased by an assigned transfer, the comparison can estimate the effect of the transfer within the experiment. If liquidity is merely observed, prior worry may influence how much cash participants save, and outstanding obligations may determine both reserves and severity. Temporal data, appropriate design and a clearly stated causal diagram or equivalent assumptions are required before a causal interpretation is defensible.

Wealth measurements introduce error through omitted assets, uncertain valuations, reluctance to disclose, unrecorded liabilities, and differences in access. Psychological measurements introduce response noise, scale-use differences, demand effects, and ambiguity of question interpretation. Researchers should collect multiple indicators where possible, report missingness and uncertainty, and examine how alternative wealth definitions affect the conclusions. A single self-reported wealth category should not support fine-grained estimates at the extreme upper tail.

Nonlinear analysis requires particular care. Researchers should specify plausible functional forms before observing outcomes or clearly mark exploratory analysis. Flexible models can reveal departures from monotonicity, but a fitted turning point in a small sample can reflect sparse observations, scale choices, or model specification. The model-dependent findings of de Courson et al. (2025) in the UK and France survey illustrate why objective and subjective measures, alternative specifications, and out-of-sample validation matter.

The analysis should also distinguish failure to reject a difference from evidence of practical equivalence. Carvalho et al.'s (2016) payday evidence informs the question of short-run resource changes in that setting; the absence of a detected risk-taking difference should not be described as a mathematically zero effect in every population. Precision, outcome sensitivity, and the exact contrast studied determine what conclusions follow.

Ethical and interpretive boundaries

Research on wealth can involve sensitive personal and financial information. Collecting detailed asset data, account locations, ownership structures, and perceived vulnerabilities creates privacy and security responsibilities. Investigators should minimize identifying information, use appropriate data protection and informed consent, and avoid publishing granular combinations of characteristics that facilitate reidentification. Data linkage requires a lawful basis and safeguards suited to the jurisdiction and purpose.

Interpretation carries a separate ethical burden. High wealth does not establish recklessness, dishonesty, immunity to distress, or superior judgment; low wealth does not establish irrationality or defective self-control. A person's risk decisions may reflect the real cost of a loss under their circumstances. Psychological explanation should be grounded in measured processes, particularly when the objects of study include unsampled or inaccessible populations.

The same applies to policy uses of risk measures. A reported willingness to accept uncertainty is only one component of a person's financial circumstances. A threat scale developed for research should not be substituted for a diagnosis or a complete individualized assessment. The purpose of this dossier is explanatory and methodological, and its findings do not prescribe a particular investment, insurance, tax, or jurisdictional strategy.

Synthesis and research implications

The literature supports a differentiated account of the link between resources and risk-related outcomes. First, wealth changes the feasible consequences of adverse events through liquidity, commitments, diversification possibilities, and recovery options. Second, socioeconomic position is associated with financial and macroeconomic expectations, and relative-position information can influence risk-taking in particular experimental settings. Third, subjective appraisal and actual behavior are empirically distinct, and studies that measure both provide stronger evidence for connecting them.

The evidence is strongest for the necessity of these distinctions and for context-dependent associations. It is less complete for the causal effect of durable changes in wealth on separately measured probabilities, severity, perceived control, uncertainty, and affect. Experiments around payday and lottery-based variation improve causal identification of resource-related effects on selected behaviors, but the outcomes differ from direct perceptions. Wealth survey data and socioeconomic gradients establish additional associations rather than closing that gap.

Several apparently contradictory findings become intelligible when their measured outcomes and treatments are compared. Reduced relative risk-taking after a windfall, lower financial optimism among some lower-status groups, and nonlinear risk-taking associated with subjective resource insufficiency are not estimates of a single common quantity. They concern distinct populations, levels of resources, horizons, constraints, and psychological or behavioral variables. Their coexistence calls for careful specification of the dependent variable rather than a universal wealth coefficient.

An informative next research program would prioritize directly measured judgments of defined financial events, alongside validated threat and willingness measures; collect asset composition and access rather than only income; document subjective and objective rank; prespecify the causal contrast; and examine heterogeneity across actual exposure and institutional environment. For cross-border wealth, such research must additionally identify the specific assets, rights and risks being appraised and the role of delegated decisions. These proposals are methodological implications of the synthesis, not claims of a new named theory or an original empirical discovery.

Conclusion

Wealth is a resource condition that changes the potential impact of loss, the set of feasible responses to uncertainty, and the economic contexts in which financial judgments are formed. Its psychological consequences may depend on the meaning and accessibility of resources, obligations, expectations, perceived control, economic position, and the nature of the threatened goal. The same nominal wealth can therefore coexist with markedly different risk appraisals.

A scientifically defensible account separates objective exposure, perceived probability, perceived severity, uncertainty, affect, willingness to take risk, risk-bearing capacity, and observed action. The distinctions prevent a portfolio choice or a wealth category from being treated as a direct measurement of the person's perceived danger. They also clarify why findings on risk-taking under temporary scarcity, financial windfalls, and relative economic position need not establish a common directional relationship between wealth and risk perception.

For the Global Offshore System, the principal implication concerns evidence. The ownership or location of an asset can establish facts about institutional exposure; explanations of the owner's perception and motives require appropriate psychological measurements and causal designs. Existing distributional and financial research motivates those investigations. It does not replace them.

The most productive continuing question is which particular aspect of perceived financial risk changes when a specific resource condition changes, for whom, over what period, and under what actual constraints. Answering that question requires combining economic measurement and psychologically precise elicitation. This is the basis for cumulative research on the psychology of wealth and risk.

References

  • Alstadsæter, A., Johannesen, N., & Zucman, G. (2019). Tax evasion and inequality. American Economic Review, 109(6), 2073–2103. DOI: 10.1257/aer.20172043. Research record
  • Bender, S., Choi, J. J., Dyson, D., & Robertson, A. Z. (2022). Millionaires speak: What drives their personal investment decisions? Journal of Financial Economics, 146(1), 305–330. DOI: 10.1016/j.jfineco.2021.09.016. Research record
  • Briggs, J., Cesarini, D., Lee, S. C., Lindqvist, E., & Östling, R. (2023). Financial Windfalls, Portfolio Allocations, and Risk Preferences. SOFI Working Paper 15/2023. Research record
  • Carvalho, L. S., Meier, S., & Wang, S. W. (2016). Poverty and economic decision-making: Evidence from changes in financial resources at payday. American Economic Review, 106(2), 260–284. DOI: 10.1257/aer.20140481. Research record
  • Das, S., Kuhnen, C. M., & Nagel, S. (2020). Socioeconomic status and macroeconomic expectations. Review of Financial Studies, 33(1), 395–432. DOI: 10.1093/rfs/hhz041. Research record
  • de Courson, B., Frankenhuis, W. E., & Nettle, D. (2025). Poverty is associated with both risk avoidance and risk taking: Empirical evidence for the desperation threshold model from the UK and France. Proceedings of the Royal Society B: Biological Sciences, 292(2040), 20242071. DOI: 10.1098/rspb.2024.2071. Research record
  • de Courson, B., Frankenhuis, W. E., & Nettle, D. (2026). Explaining the paradoxical effects of poverty on risk taking: The desperation threshold model. Behavioral and Brain Sciences, advance online publication. DOI: 10.1017/S0140525X26104567. Research record
  • Dohmen, T., Falk, A., Huffman, D., Sunde, U., Schupp, J., & Wagner, G. G. (2011). Individual risk attitudes: Measurement, determinants, and behavioral consequences. Journal of the European Economic Association, 9(3), 522–550. DOI: 10.1111/j.1542-4774.2011.01015.x. Research record
  • Falk, A., Becker, A., Dohmen, T., Enke, B., Huffman, D., & Sunde, U. (2018). Global evidence on economic preferences. Quarterly Journal of Economics, 133(4), 1645–1692. DOI: 10.1093/qje/qjy013. Research record
  • Fehr, D., & Reichlin, Y. (2025). Status, Control, and Risk: How Relative Wealth Perceptions Shape Risk-Taking. CESifo Working Paper No. 9253, 2025 revision. Research record
  • Giglio, S., Maggiori, M., Stroebel, J., & Utkus, S. (2021). Five facts about beliefs and portfolios. American Economic Review, 111(5), 1481–1522. DOI: 10.1257/aer.20200243. Research record
  • Guiso, L., & Paiella, M. (2008). Risk aversion, wealth, and background risk. Journal of the European Economic Association, 6(6), 1109–1150. DOI: 10.1162/JEEA.2008.6.6.1109. Research record
  • Kuhnen, C. M., & Miu, A. C. (2017). Socioeconomic status and learning from financial information. Journal of Financial Economics, 124(2), 349–372. DOI: 10.1016/j.jfineco.2017.03.002. Research record
  • Loewenstein, G. F., Weber, E. U., Hsee, C. K., & Welch, N. (2001). Risk as feelings. Psychological Bulletin, 127(2), 267–286. DOI: 10.1037/0033-2909.127.2.267. Research record
  • Marjanovic, Z., Greenglass, E. R., Fiksenbaum, L., & Bell, C. M. (2013). Psychometric evaluation of the Financial Threat Scale (FTS) in the context of the great recession. Journal of Economic Psychology, 36, 1–10. DOI: 10.1016/j.joep.2013.02.005. Research record
  • Weber, E. U., Blais, A.-R., & Betz, N. E. (2002). A domain-specific risk-attitude scale: Measuring risk perceptions and risk behaviors. Journal of Behavioral Decision Making, 15(4), 263–290. DOI: 10.1002/bdm.414. Research record