Health Equity

How health equity connects fairness to health

Health equity is a normative principle: people should have a fair and just opportunity to attain their full health potential. Applying that principle requires more than identifying differences in health. This page explains how values, social conditions, distributional evidence, and decision processes combine to determine whether a health difference is an equity concern and what action may be justified.

Health equity is not the same as equality

Equality usually means giving people the same resources, services, or treatment. Equity asks whether arrangements are fair in light of different needs, barriers, starting positions, and opportunities. Equal provision can therefore be inequitable when people need different forms or levels of support to achieve a fair opportunity for health.

Principle	Central question	Health-system example
Equality	Are people receiving the same input or treatment?	Every clinic offers the same appointment hours.
Equity	Are resources and arrangements fair given differences in need and barriers?	Clinics serving populations with greater need receive additional capacity and offer accessible hours.
Health equality	Are health levels or service-use rates the same across groups?	Two groups have the same vaccination uptake.
Health equity	Are health opportunities and outcomes distributed fairly and justly?	Preventable access barriers are removed so that each group has a fair opportunity to benefit.

Equity does not require every person to have identical health outcomes. Some variation may reflect age, informed preferences, biological factors, chance, or other circumstances that a society does not judge remediable or unjust. The ethical and empirical task is to identify differences that are systematic, avoidable or remediable, and unfair.

Inequality and inequity describe different claims

A health inequality is an observed difference in health, access, quality, exposure, or another measure between people or groups. A health inequity is a difference judged to be unfair or unjust under an explicit ethical standard. Data can establish inequality, but the move from inequality to inequity requires normative reasoning and contextual knowledge.

• Descriptive analysis asks who experiences different health outcomes, risks, services, costs, or opportunities.
• Causal analysis asks which mechanisms and conditions produce those differences.
• Normative analysis asks whether the differences and their causes are unfair, avoidable, remediable, or socially unacceptable.
• Policy analysis asks which actions could reduce the unfair difference without creating greater harms elsewhere.

Fair opportunity is the core idea

The fair-opportunity view focuses on whether people can realistically attain health rather than whether a service nominally exists. Opportunity is shaped by material resources, power, discrimination, geography, information, physical accessibility, administrative rules, and the ability to bear financial and non-financial costs. A fair system must therefore consider both formal entitlement and practical capability to use and benefit from care.

Fair opportunity can be constrained at several stages:

• Exposure to unhealthy living and working conditions can create unequal need before contact with healthcare.
• Eligibility and entitlement rules can exclude people from services or financial protection.
• Availability, distance, waiting time, language, trust, and discrimination can limit access.
• Differences in diagnosis, treatment, continuity, and quality can limit effective coverage.
• Financial costs, time costs, caregiving duties, and lost income can make nominal access unusable.
• Unequal power and representation can prevent affected communities from shaping decisions.

Horizontal and vertical equity

Horizontal equity means treating people with equal relevant need equally. Vertical equity means appropriately different treatment for people with different levels or types of relevant need. These principles help translate the broad idea of fairness into questions about financing, access, use, quality, and outcomes.

• Horizontal equity in service use asks whether people with equal healthcare need receive comparable care.
• Vertical equity in service use asks whether people with greater need receive appropriately greater or different care.
• Horizontal equity in financing asks whether people with similar ability to pay contribute similarly.
• Vertical equity in financing asks whether contributions vary fairly with ability to pay, often through progressivity.

Neither principle determines the correct policy by itself. Analysts must specify which differences count as relevant need, which circumstances should not affect treatment, and what constitutes an appropriate response.

Distributional and procedural equity

Distributional equity concerns who receives health benefits, bears costs, experiences risks, or gains access to resources. Procedural equity concerns how decisions are made, including whose evidence counts, who participates, whether reasons are transparent, and whether decisions can be challenged. A policy with a favourable average distribution can still be inequitable if affected people were excluded or treated without respect.

An equity review should therefore examine both outcomes and process:

• Distributional questions identify the allocation of health, healthcare, financial burden, time, risk, and opportunity.
• Procedural questions examine representation, accessibility, transparency, accountability, appeal, and respectful treatment.
• Recognition questions examine whether institutions understand and value the identities, experiences, and knowledge of affected communities.

Structural and social causes of health inequity

Health inequities are produced through social, economic, commercial, environmental, and political systems rather than through healthcare alone. The distribution of income, education, housing, employment, transport, environmental hazards, political influence, and discrimination shapes both exposure to illness and the resources available to respond. Health policy can reduce, reproduce, or deepen these patterns.

Structural mechanisms include:

• Laws, institutions, and markets that distribute power and resources unevenly.
• Racism, sexism, ableism, class disadvantage, colonial legacies, and other forms of discrimination.
• Residential segregation, environmental exposure, insecure housing, and unequal infrastructure.
• Employment conditions, income insecurity, education, and access to social protection.
• Health-system design, financing, coverage rules, workforce distribution, and administrative burden.

Intersectionality changes what an average group comparison reveals

People occupy several social positions at once, and their combined effects may not equal the sum of separate categories. For example, the barriers experienced by a low-income disabled woman may differ qualitatively from the average patterns reported for income, disability, or sex alone. Intersectional analysis examines how systems of power interact to shape exposure, access, treatment, and outcomes.

Analysts should avoid treating social groups as biologically fixed or internally uniform. Categories should be selected because they illuminate a plausible mechanism or accountability question, not simply because they are available in a dataset.

Choosing dimensions for equity analysis

Equity-relevant dimensions should reflect the decision context and the pathways through which unfair disadvantage may arise. A commonly used prompt is to consider place of residence, race or ethnicity, occupation, gender or sex, religion, education, socioeconomic status, and social capital, while adding age, disability, migration status, language, sexual orientation, and other locally relevant dimensions. The list is a starting point rather than a universal checklist.

For each dimension, an analysis should state:

• Why the characteristic is relevant to the health question or mechanism.
• How the groups are defined and whether people can self-identify.
• Whether sample size and data quality support valid estimates.
• Whether aggregation hides important differences within a category.
• How privacy, stigma, and potential misuse of the data will be managed.

Measuring absolute and relative inequality

Absolute and relative measures answer different questions and can move in opposite directions. An absolute difference describes the size of the gap in the original units, while a relative measure describes proportional disparity. Good reporting normally presents both and includes group-specific levels so that the underlying changes remain visible.

If (y_D) is the outcome for a disadvantaged group and (y_A) is the outcome for an advantaged group, an absolute gap for an adverse outcome can be written as:

$$ Absolute\ gap = y_D - y_A $$

The corresponding relative ratio is:

$$ Relative\ ratio = \frac{y_D}{y_A} $$

Suppose avoidable hospitalisation falls from 30 to 20 per 1,000 in a disadvantaged group and from 10 to 5 per 1,000 in an advantaged group. The absolute gap falls from 20 to 15 per 1,000, but the relative ratio rises from 3.0 to 4.0. Reporting only one measure would give an incomplete account of equity progress.

Measuring inequality across an ordered social gradient

Pairwise comparisons can miss the pattern across the full socioeconomic distribution. The slope index of inequality estimates the absolute difference across the social hierarchy, while the relative index of inequality expresses the gradient relative to an average level. Concentration curves and concentration indices describe how a health variable is distributed across an ordered socioeconomic ranking.

A common concentration-index form is:

$$ C = \frac{2}{\mu}\operatorname{Cov}(y_i,r_i) $$

where (y_i) is the health variable for person (i), (r_i) is that person's fractional socioeconomic rank, and (\mu) is the mean of (y). The sign and interpretation depend on whether the variable represents health, ill health, service use, or another outcome; bounded variables may also require correction. Analysts should state the ranking variable, weighting, standardisation, and interpretation rather than presenting the index alone.

Need adjustment can change the equity conclusion

Differences in healthcare use are not automatically inequitable because groups may have different levels of healthcare need. Horizontal-equity analysis often compares actual use with expected use after accounting for legitimate need variables. The choice of what counts as need and what counts as an avoidable non-need influence is itself normative and should be transparent.

Variables such as illness severity may represent need, while income, insurance status, ethnicity, or geography may reveal barriers that should not determine use after need is considered. Over-adjustment can conceal inequity when a variable lies on the causal pathway from structural disadvantage to poor health or access.

Data quality shapes whose inequity can be seen

Equity analysis depends on data that accurately include and identify relevant populations. Small samples, missing demographic fields, unstable denominators, inconsistent categories, linkage error, and exclusion from surveys or administrative systems can make disadvantaged groups statistically invisible. Suppressing small cells may protect privacy but can also remove precisely the evidence needed for accountability.

An equity data audit should examine:

• Population coverage and groups omitted from the source.
• The validity, consistency, and acceptability of classification fields.
• Missingness and whether it differs systematically across groups.
• Sample size, uncertainty intervals, and instability in small areas or intersections.
• Changes in definitions that affect trends.
• Privacy, governance, community consent, and the risk of harmful inference.

Health equity in priority setting

Priority setting determines who receives health gains, who waits, and which needs remain unmet. Conventional economic evaluation often focuses on maximising total health from limited resources, while equity-focused approaches make the distribution of health and opportunity explicit. Decision makers must state the ethical basis for any equity weight or trade-off rather than treating it as a purely technical parameter.

Relevant questions include:

• Should priority reflect severity, lifetime health, socioeconomic disadvantage, rarity, caregiving responsibility, or another ethically relevant factor?
• How much total health, if any, is society willing to forgo to achieve a fairer distribution?
• Are benefits and costs distributed across the same people and time periods?
• Does the intervention reduce an unfair gap, or does it mainly improve outcomes for people already advantaged?
• Are opportunity costs likely to fall on groups not represented in the evaluation?

Distributional cost-effectiveness analysis

Distributional cost-effectiveness analysis extends economic evaluation by estimating how health benefits and opportunity costs are distributed across equity-relevant groups. It can compare total health with changes in health inequality under alternative policies. Results depend on baseline health distributions, subgroup-specific effects, service uptake, costs, displaced health, and the selected social welfare function.

A simplified equally distributed equivalent health measure can be represented using an inequality-aversion parameter (\varepsilon):

$$ H_{EDE} = \left(\sum_{g=1}^{G} p_g H_g^{1-\varepsilon}\right)^{\frac{1}{1-\varepsilon}}, \quad \varepsilon \ne 1 $$

where (H_g) is health for group (g), (p_g) is the group's population share, and larger (\varepsilon) places greater weight on reducing inequality. The parameter expresses an ethical preference and should be explored through transparent sensitivity analysis rather than presented as an objective constant.

Extended cost-effectiveness analysis

Extended cost-effectiveness analysis can estimate health gains, financial-risk protection, and distributional consequences across population groups. It is especially useful when public financing changes both health outcomes and exposure to out-of-pocket costs. The approach complements rather than replaces analysis of fairness in process, discrimination, rights, and non-financial barriers.

Outputs may include deaths averted, cases prevented, household expenditure averted, catastrophic spending prevented, and benefits by income group. Results should show the incidence of both programme benefits and financing burdens.

Proportionate universalism

Proportionate universalism means providing action across the whole population while increasing scale or intensity according to disadvantage or need. It aims to avoid the stigma and exclusion of narrowly targeted programmes while recognising that equal universal provision may widen gaps. The appropriate gradient of support must be based on evidence and revisited as needs and barriers change.

A universal vaccination programme, for example, may be paired with additional outreach, transport, trusted community delivery, or flexible hours in areas with lower access. The universal entitlement remains, but implementation effort is proportionate to barriers.

Assessing whether a policy advances equity

An equity assessment should begin before implementation so that objectives, affected groups, mechanisms, and baseline differences are explicit. It should then monitor both intended and unintended distributional effects. Participation by affected communities improves the relevance of assumptions and can reveal burdens that routine datasets miss.

1. Define the fairness objective. State which opportunity, outcome, burden, or process should become more equitable and why.
2. Identify affected groups and intersections. Include people who may benefit, bear costs, lose access, or remain unseen.
3. Map the causal pathway. Explain how the policy could change exposures, access, quality, uptake, financial burden, and outcomes.
4. Select distributional measures. Report group levels, absolute and relative gaps, gradients, and uncertainty as appropriate.
5. Assess implementation barriers. Examine administrative burden, trust, accessibility, workforce, affordability, and discrimination.
6. Monitor consequences and revise. Track whether the policy narrows unfair gaps without worsening other important outcomes.

Equity impact is not guaranteed by an equity intention

A policy labelled equitable can still widen disparities if advantaged groups are better able to learn about, access, or benefit from it. Digital-first services, complex applications, co-payments, travel requirements, and inflexible appointment systems can generate unequal uptake. Implementation evidence must therefore test reach, quality, adherence, outcomes, and burden across groups.

Average improvement with widening gaps is a common warning sign. Decision makers should examine whether the policy needs redesign, additional support, or a different delivery route rather than assuming that benefits will eventually diffuse.

Common misunderstandings

Health equity is often reduced either to equal outcomes or to a descriptive table of subgroup differences. Both interpretations are incomplete because equity connects empirical evidence to an explicit judgment about fairness. The following distinctions protect against overclaiming.

• Health equity does not mean that every health outcome must be identical.
• A health inequality is not automatically a health inequity.
• Statistical adjustment does not remove the need for ethical reasoning.
• Equal service provision can preserve inequity when needs and barriers differ.
• Targeting a disadvantaged group is not automatically equitable if it excludes others in comparable need or creates stigma.
• A national average can improve while inequity worsens.
• Healthcare policy alone cannot eliminate inequities produced by wider social and structural conditions.
• Equity does not replace efficiency; policy must make any trade-off between distribution and total health explicit.

Reporting an equity analysis transparently

Transparent reporting separates observed differences, causal claims, and normative judgments. It explains why groups and measures were selected and acknowledges where data or community knowledge are missing. It also shows whether conclusions change across reasonable ethical and statistical choices.

• Report the fairness principle and equity objective before presenting results.
• Report group-specific levels as well as absolute, relative, and gradient measures.
• Explain need adjustment, standardisation, weighting, missing-data handling, and uncertainty.
• Describe intersectional analyses and any limitations caused by small samples or classification.
• Distinguish association from causal evidence about the source of inequity.
• Report participation, governance, privacy protections, and community interpretation where relevant.
• Present sensitivity analyses for equity weights, inequality aversion, and distribution of opportunity costs.

The decision standard

Health equity asks whether people have a fair and just opportunity for health and whether institutions are changing conditions that unfairly constrain that opportunity. A credible conclusion combines distributional data, causal understanding, explicit ethical reasoning, and fair decision processes. The objective is not simply to make a metric more equal, but to remove unjust barriers and create arrangements in which everyone can realistically attain their health potential.