Equity, Fairness, and Distributive Justice

Why this matters in public value

Every allocation decision a public or social-sector body makes is also, without anyone necessarily intending it, an answer to a question of distributive justice: who gets how much of a good that cannot be given to everyone in full. A hospital that ranks patients on a waiting list, a housing authority that scores applicants against a points system, a tax office that sets thresholds and reliefs, and a grant-making foundation that chooses which applications to fund are all doing the same underlying thing — deciding, case by case, what fairness requires — even when none of them uses the word 'equity' in the policy document that authorizes the rule.

The stakes are not abstract. A benefit that is universal reaches everyone, including people who do not need it, and costs more per person helped; a benefit that is targeted reaches fewer non-recipients but routinely misses some of the very people it exists for, because eligibility rules, stigma, and administrative friction all suppress take-up. Neither design is simply 'the fair one' — each embodies a different, defensible theory of what equity means, and a leader who cannot articulate which theory a programme rests on cannot defend it when it is challenged, nor redesign it competently when circumstances change.

This topic gives public managers and social-sector leaders two rigorous philosophical traditions for reasoning about fairness — the work of the American philosopher John Rawls and the capability approach associated with the economist Amartya Sen and the philosopher Martha Nussbaum — plus the practical vocabulary of horizontal and vertical equity that turns those traditions into allocation rules a real organization can operate. Equity is one of the values that Topic 1.4 — Public Goods, Merit Goods, and Value Pluralism identifies as incommensurable with efficiency and liberty (see Topic 1.4 — Public Goods, Merit Goods, and Value Pluralism): it cannot always be converted onto the same scale as cost per outcome, and pretending otherwise is itself a choice, not a neutral default.

Core concepts

Distributive justice is the branch of political philosophy concerned with how a society ought to allocate benefits and burdens — income, services, opportunities, risks — among its members. Public value theory inherits this question directly: every budget, formula, and eligibility rule distributes something, and distributive justice supplies the standards against which that distribution can be judged good or bad, rather than merely legal or popular.

Rawls's 1971 book A Theory of Justice is the most influential modern attempt to derive fair principles for a society's basic structure from first principles rather than from custom or utility. Rawls asks readers to imagine designing the rules of a society from behind a veil of ignorance: a hypothetical state in which the people choosing the rules do not know what position — rich or poor, healthy or disabled, majority or minority — they will occupy once the veil lifts and the rules apply to them. Reasoning under that uncertainty, Rawls argues, rational people would not gamble on rules that could leave them destitute; they would instead choose rules that protect the worst-off position, because any one of them might turn out to hold it. From this thought experiment Rawls derives two principles: equal claim to the most extensive basic liberties compatible with the same liberties for everyone else, and permission for social and economic inequalities only where they attach to positions open to all under fair equality of opportunity and work to the greatest benefit of the least advantaged. This second half of the second principle is the difference principle: inequality is not inherently unjust, but it is only justified when it improves the position of those who have least, not merely the average.

Applied to service allocation, the difference principle gives a concrete test beyond 'be equal' or 'be efficient': does this formula, this eligibility rule, make the worst-off group better off than the next-best alternative would? A formula weighting deprived areas more heavily than an equal-shares formula is a difference-principle argument, whether or not the document ever mentions Rawls. The veil of ignorance is a useful design discipline in its own right: a team that genuinely tries to design a rule without knowing which side of it they would occupy tends to produce more defensible rules than one that starts from an existing distribution of advantage.

The capability approach, developed by Sen from the 1980s onward and extended by Nussbaum into a more fully specified list of entitlements, asks a different question: not how resources or utility should be distributed, but whether people have the real freedom to do and be the things they have reason to value — Sen's distinction between a functioning (a state a person actually achieves, such as being adequately nourished) and a capability (the genuine opportunity to achieve it, whether or not they choose to). Two people given an identical resource — a wheelchair ramp, a cash transfer, a training place — can end up with very different capabilities, because conversion factors, personal, social, and environmental, determine how effectively a resource turns into a real opportunity: a ramp at a building with no accessible transport route delivers the resource without delivering the capability. Nussbaum went further, proposing a list of 10 central human capabilities — among them life, bodily health, bodily integrity, and control over one's environment — that she argues any decent society is obliged to secure as a threshold. For service design, the capability approach corrects a resourcist error equity policy makes constantly: counting a service as delivered because an input was provided, without asking whether the recipient can convert that input into the outcome intended.

Sitting underneath both traditions is a distinction public finance uses every day: horizontal and vertical equity. Horizontal equity means treating people who are alike in the relevant respect alike: two patients with identical clinical need should receive identical treatment regardless of where they live; two taxpayers with identical income should pay identical tax. Vertical equity means treating people who differ in the relevant respect differently, in proportion to that difference: a progressive tax that takes a higher share of income from a higher earner is a vertical-equity instrument, as is a service allocating more support to greater assessed need. The two principles are not in tension in the abstract — they apply to different situations — but a single allocation rule still has to decide, for any given pair of people, whether they are 'alike' or 'different', and that classification is where most real equity disputes live.

That classification question is also where equity meets efficiency and universalism. A universal service — free at the point of use to everyone — maximizes horizontal equity by construction: no one is excluded or means-tested, and no one bears the stigma a targeted scheme imposes. Universal basic services is the contemporary name for extending that logic further across the welfare state. A targeted, means-tested service concentrates the same budget on those with the greatest assessed need, which looks more vertically equitable per unit of spending, but a means test is never free: it costs money to administer, it creates a threshold some genuinely eligible people fail to cross through complexity or stigma (the take-up gap), and it produces a cliff edge where two people in near-identical circumstances receive very different treatment — a fresh horizontal-equity problem generated by an attempt to fix a vertical-equity one. Topic 2.2 — Social Cost-Benefit and Cost-Effectiveness Analysis shows the efficiency side of this same trade-off through distributional weighting, which adjusts a cost-benefit appraisal to reflect that a unit of resource is worth more, in welfare terms, to a poorer recipient (see Topic 2.2 — Social Cost-Benefit and Cost-Effectiveness Analysis); this topic supplies the philosophical grounding for why that adjustment is defensible, not merely a technical convenience.

Best practices

1. Name the equity principle before you design the formula. Decide explicitly whether a given allocation is meant to satisfy horizontal equity (same need, same service) or vertical equity (different need, proportionately different service) before drafting the rule, because the two principles imply different formulas and cannot both be optimized by the same one.
2. Run separate audits for horizontal and vertical equity. A horizontal audit asks whether people with matching assessed need receive matching service across every delivery unit and location; a vertical audit asks whether the gap in service between differing needs tracks the gap in need itself. Neither audit substitutes for the other, and a programme can pass one while quietly failing the other.
3. Use the veil of ignorance as a live design test, not a classroom exercise. Before a policy panel signs off an eligibility rule, ask each member to imagine not knowing which side of the threshold they would fall on; rules that survive that test tend to need fewer emergency amendments once real cases start arriving.
4. Check conversion factors, not just resource delivery. Before declaring a service equitable because an input has been distributed, verify that the people receiving it can actually convert it into the outcome intended — a training place is not equal opportunity if childcare, transport, or language support is missing.
5. Pair every targeted programme with an explicit take-up strategy. A means-tested benefit's vertical-equity case collapses if a large share of eligible people never claim it; budget for outreach, simplified applications, and automatic enrolment wherever the data allows it, and report take-up alongside spend.
6. Model the equity and cost consequences of universal versus targeted design side by side, before choosing. Present decision-makers with both the horizontal-equity case for universalism and the vertical-equity, cost-per-outcome case for targeting in the same document, rather than letting one framing quietly dominate the choice.
7. Smooth cliff edges wherever a threshold creates a large step change. Where feasible, taper support near an eligibility boundary instead of withdrawing it all at once, because a hard cliff edge routinely produces worse horizontal equity than the untargeted alternative it replaced.
8. Disaggregate outcome data by the groups most likely to reveal a vertical-equity gap. An aggregate service-satisfaction score can conceal a large gap between, for instance, rural and urban users, or between language groups; report the disaggregated figures routinely, not only after a complaint.
9. Involve the people affected in defining what counts as equivalent need. Professionals and affected communities frequently disagree about which differences are relevant to an allocation decision; treat that disagreement as information to resolve through engagement, not as noise to be designed around.
10. Review allocation formulas against outcomes on a fixed cycle, not only after a complaint or a scandal. A formula that was defensible when it was written can drift out of alignment with need as demographics, costs, or service models change; schedule the review rather than waiting for a crisis to force it.
11. Document the trade-off explicitly whenever equity and efficiency conflict. When a more equitable option costs more per unit of outcome than a less equitable one, say so in the business case in plain terms, so the decision-maker who accepts the extra cost is accountable for having chosen equity deliberately (see Topic 2.2 — Social Cost-Benefit and Cost-Effectiveness Analysis).

Questions to discuss with your team

Which idea of fairness actually governs our largest allocation decision, and have we ever said so out loud? Most organizations can point to a formula, a waiting list, or an eligibility rule without being able to state which equity principle it implements. Walk through your single largest allocation decision — a budget formula, a service-eligibility rule, a grant round — and ask whether it is designed to deliver horizontal equity, vertical equity, or some blend, and whether that blend was chosen deliberately or inherited from a previous administration. Ask, too, who in the room is actually accountable for that choice, since a principle nobody owns is a principle nobody can revise. If nobody can answer confidently, that is itself a finding: an unnamed principle cannot be defended when challenged, and cannot be redesigned competently when circumstances change.

Where do we run targeted and universal versions of a similar service, and would either group be better served by swapping the model? Many organizations operate a mix of targeted and universal provision across different services without ever comparing them directly. Pick one targeted and one universal programme addressing a related need, and discuss what the targeted programme's take-up rate actually is, and what the universal programme's cost per beneficiary who did not strictly need it looks like. Ask what would happen to each measure if the two designs swapped places, and whether anyone has actually modelled that swap rather than assumed the answer. This question often surfaces uncomfortable evidence that a targeted scheme is quietly failing on vertical equity because of low take-up, or that a universal scheme has drifted into subsidizing people it was never meant to reach.

What conversion factors are we assuming away when we count a service as delivered? Teams routinely measure success by outputs distributed — places allocated, ramps built, vouchers issued — rather than by whether the recipient could use them. Discuss your organization's flagship service and list the personal, social, and environmental conditions a recipient needs in order to convert that service into the outcome it is meant to produce. Ask whether frontline staff have ever been asked systematically which of those conditions are missing most often, since they usually know before any dashboard shows it. Where those conditions are absent for a predictable group of recipients, decide whether that gap belongs on the programme's own risk register rather than being treated as somebody else's problem.

Where is our sharpest eligibility cliff edge, and who falls just on the wrong side of it? Every threshold-based rule creates a boundary, and boundaries generate the starkest horizontal-equity complaints an organization receives, because two nearly identical cases receive very different treatment. Identify the threshold in your organization that produces the most appeals or complaints, and discuss whether a tapered alternative was ever costed, and if not, why not. Consider, too, how long it has been since the threshold itself was reviewed, since a figure fixed in cash terms quietly tightens every year that prices rise around it. This question is worth revisiting whenever a threshold has not moved in real terms for several years, since inflation alone can quietly widen the group caught on the wrong side.

If we designed this service from behind a veil of ignorance, would we design it the way it exists today? Rawls's thought experiment is a discipline as much as a theory: imagine the design team does not know whether they will end up as the highest-need or lowest-need user of the service under discussion. Walk through your service's rules as if for the first time under that constraint, and discuss which existing features would survive and which look like they were shaped by whoever had the most influence over the original design. Notice, in particular, which features benefit whoever is already best placed to navigate a complex system, since those are usually the ones a veil-of-ignorance test removes first. Where the answers diverge sharply from current practice, treat that gap as a prompt for redesign, not merely an interesting philosophical observation.

How do we know whether our vertical-equity commitments are being met, in outcomes, rather than merely declared in policy? A formula that allocates more resource to greater need is only a vertical-equity success if the resulting outcomes actually narrow, rather than merely redistributing inputs. Discuss what outcome measure would tell you, within a year, whether your highest-need group is closing the gap with your lowest-need group, and whether that measure is currently tracked, by whom, and reported to whom. Ask what the organization would actually do differently if that measure showed the gap widening rather than narrowing. If the honest answer is that nobody currently owns that comparison, decide in the room who will from today.

In practice: a public value example

A regional social-care authority in a fictional mid-sized country runs a home-support programme for older residents who need help with daily living. For a decade, the programme allocated a flat weekly entitlement of six hours of home-care support to anyone over 75 who registered — a universal design, chosen because it was simple to administer and avoided the stigma of a means test. As the region's older population grew faster than its budget, the flat entitlement began to feel wrong: a person managing early-stage memory loss alone received the same six hours as an independent resident who had simply turned 75, while a small number of residents with severe, complex needs were clearly under-served.

The authority's leadership team, redesigning the programme, deliberately worked through both traditions in this topic rather than defaulting to whichever consultant's template arrived first. Applying a Rawlsian test, they asked which redesign would most improve the position of the worst-off residents, even at the cost of a less generous entitlement for residents with light needs; this argued for moving from a flat allocation toward a needs-weighted one, the difference principle applied directly to a care-hours formula. Applying the capability approach, an occupational therapist on the redesign panel pointed out that hours of care are not the same as the capability to live independently: two residents with identical assessed need could have very different real independence depending on family nearby, a step-free home, or transport to appointments — conversion factors the flat-hours model had never measured.

The authority adopted a tiered model: a smaller universal baseline of two hours for every registered resident over 75, preserving horizontal equity and avoiding a gatekeeping step for light needs, combined with an assessed, needs-weighted allocation above that baseline informed by conversion-factor questions about housing, transport, and isolation, not clinical need alone. They built a deliberate taper rather than a hard threshold between tiers, having reviewed complaint data from a neighbouring authority whose earlier, more abrupt redesign had generated a wave of appeals from residents just below its new cut-off. The redesigned programme cost slightly more to administer, a trade-off the authority's board accepted and minuted explicitly, on the basis that the extra cost bought a defensible answer to both the horizontal- and vertical-equity questions the flat model had left open. This scenario is illustrative; it does not describe any real region, authority, or care programme.

Four sector lenses

Local government

Local authorities usually apply equity principles that national government has partly pre-decided — a national funding formula, a national minimum entitlement — but they retain real discretion over local eligibility criteria, waiting-list prioritization, and how far to taper thresholds within their own budget. Because local boundaries are visible and comparable to residents in a way national averages are not, horizontal-equity complaints concentrate here: two households either side of a district boundary can receive markedly different services for identical need, a pattern residents notice and organizations are frequently unable to justify beyond administrative history.

National government

National government sets the vertical-equity architecture that most public services operate within — the progressive tax system, the national minimum entitlements, the formulae that redistribute funding toward higher-need regions. Because national programmes operate at a scale where the take-up gap in targeted schemes is measurable in the hundreds of thousands or millions of people, the universal-versus-targeted design choice carries larger fiscal and political consequences here than in any other sector, and is correspondingly harder to reverse once entrenched.

Social sector and nonprofit

Nonprofits and non-governmental organizations (NGOs) frequently exist specifically to serve a narrowly defined, high-need population, which makes vertical equity close to their founding mission rather than a policy choice layered on top of it; the harder question for this sector is usually horizontal equity between similar organizations serving adjacent but not identical populations, where funders' targeting criteria can arbitrarily include one group and exclude a near-identical one. Because most social-sector organizations depend on restricted funding tied to a specific need category, they often have less freedom than government to redesign an allocation rule that internal evidence shows is producing a poor result.

Multilateral and international

Multilateral bodies apply equity reasoning at two levels simultaneously: within a member state's population, through the conditions attached to development finance and technical assistance, and between member states, through the formulae that decide how aid, loan terms, or vote weight are distributed across very different national circumstances. The capability approach has had particular influence here, most visibly through its role in shaping the United Nations Development Programme's Human Development Index, which explicitly rejects ranking countries on national income alone in favour of measures closer to real capability, such as health and education outcomes.

Common failure modes

Equity theatre. An organization publishes an equity strategy, commissions an equity audit, or appoints an equity lead, while the underlying allocation formula that actually determines who gets what remains unchanged. The cost is not merely wasted effort: it teaches affected communities that engagement produces documents rather than different outcomes, which erodes trust faster than having made no equity claim at all.

The postcode lottery. Identical need produces different service purely because of which local delivery unit, health area, or jurisdiction a person happens to live in, with no defensible vertical-equity justification for the difference — only administrative history. This is a horizontal-equity failure that is politically corrosive precisely because it is easy for affected residents to see and compare, unlike more diffuse national statistics.

The silent take-up gap. A means-tested or targeted programme is designed with a sound vertical-equity rationale, but a large share of the people it is meant to reach never successfully claim it, because of complexity, stigma, or simple unawareness, so the programme's real-world equity performance falls far short of its formula's intent. Organizations that measure only spend per claimant, rather than the ratio of claimants to the estimated eligible population, routinely miss this failure entirely.

The eligibility cliff edge. A hard threshold creates two nearly identical cases with very different outcomes, converting an attempt to fix a vertical-equity problem into a fresh horizontal-equity one at the boundary. This failure is especially persistent because the boundary itself is usually the cheapest part of the system to design and the most expensive part to redesign once entitlements have become embedded around it.

Counting the input, not the capability. A programme reports success by resources distributed — training places filled, devices issued, appointments offered — without checking whether recipients could convert those resources into the outcome intended, the resourcist error the capability approach was built to correct. This failure is dangerous precisely because the output metrics look healthy on a dashboard while the underlying inequity persists untouched.

Maturity model

Checklist

• State, in writing, which equity principle — horizontal, vertical, or a named blend — governs each major allocation rule.
• Run a horizontal-equity audit comparing outcomes for matched-need cases across every delivery unit or area.
• Run a vertical-equity audit comparing the size of the service gap against the size of the assessed-need gap.
• Estimate the eligible population for every targeted programme and report take-up as a ratio, not only spend.
• Identify every hard eligibility threshold and cost a tapered alternative before ruling one out.
• Test at least one major allocation rule against a veil-of-ignorance design exercise with the actual decision panel.
• List the conversion factors a typical recipient needs to turn the service into the outcome intended.
• Disaggregate outcome data by the groups most likely to reveal a hidden vertical-equity gap.
• Document the cost of any equity-motivated design choice that is more expensive than the efficient alternative.
• Schedule a fixed-cycle review of every allocation formula against current, not historical, need.
• Involve affected communities directly in defining what counts as equivalent need for at least one major rule.

Key sources

The philosophical grounding for this topic rests on John Rawls's A Theory of Justice and its account of the original position, the veil of ignorance, and the difference principle, together with the Stanford Encyclopedia of Philosophy's entry on Rawls for a rigorous secondary treatment of how these ideas connect. The capability approach draws on Amartya Sen's Development as Freedom and Martha Nussbaum's Creating Capabilities: The Human Development Approach, alongside the Stanford Encyclopedia of Philosophy's dedicated entry on the capability approach. The applied, public-finance treatment of horizontal and vertical equity, and the trade-offs between universal and targeted service design, draws on the World Bank's World Development Report 2006: Equity and Development and the OECD's applied work on equity in service systems, exemplified by its report on equity in education.

References

1. John Rawls — Wikipedia — https://en.wikipedia.org/wiki/John_Rawls
2. A Theory of Justice — Wikipedia — https://en.wikipedia.org/wiki/A_Theory_of_Justice
3. Veil of ignorance — Wikipedia — https://en.wikipedia.org/wiki/Veil_of_ignorance
4. Difference principle — Wikipedia — https://en.wikipedia.org/wiki/Difference_principle
5. Amartya Sen — Wikipedia — https://en.wikipedia.org/wiki/Amartya_Sen
6. Capability approach — Wikipedia — https://en.wikipedia.org/wiki/Capability_approach
7. Martha Nussbaum — Wikipedia — https://en.wikipedia.org/wiki/Martha_Nussbaum
8. Distributive justice — Wikipedia — https://en.wikipedia.org/wiki/Distributive_justice
9. Equity (economics) — Wikipedia — https://en.wikipedia.org/wiki/Equity_(economics)
10. Progressive tax — Wikipedia — https://en.wikipedia.org/wiki/Progressive_tax
11. Means test — Wikipedia — https://en.wikipedia.org/wiki/Means_test
12. Universal basic services — Wikipedia — https://en.wikipedia.org/wiki/Universal_basic_services
13. Justice — John Rawls's theory of justice — Stanford Encyclopedia of Philosophy — https://plato.stanford.edu/entries/rawls/
14. Capability Approach — Stanford Encyclopedia of Philosophy — https://plato.stanford.edu/entries/capability-approach/
15. World Development Report 2006: Equity and Development — World Bank — https://documents.worldbank.org/en/publication/documents-reports/documentdetail/435331468127174418/world-development-report-2006-equity-and-development
16. Development as Freedom — Amartya Sen (Open Library catalogue record) — https://openlibrary.org/works/OL116534W
17. Creating Capabilities: The Human Development Approach — Martha Nussbaum (Open Library catalogue record) — https://openlibrary.org/works/OL19364306W
18. Equity in Education: Breaking Down Barriers to Social Mobility — OECD — https://doi.org/10.1787/9789264073234-en