Equality of Opportunity

Equality of Opportunity in Health and Health Care

Equality of opportunity is a principle of equity that sorts the causes of unequal outcomes into two groups, circumstances for which people are not held responsible and choices for which they are. Only the first group makes an inequality unfair, so a gap in life expectancy between children of rich and poor parents calls for action, while a gap that follows from freely chosen behaviour may not. For health economists the principle matters because it decides which health inequalities an evaluation should try to reduce. In England the National Health Service Act 2006 asks NHS England to have regard to reducing inequalities in access and outcomes, and the NICE manual for technology appraisal (PMG36) lets committees weigh health inequality impacts alongside total health.

Circumstances, effort and the line between fair and unfair inequality

Gideon Elford, in the Stanford Encyclopedia of Philosophy, finds the common theme of every version in a justified hierarchy that classes some factors as inappropriate determinants of a person's success. Versions differ in how many factors they rule out. Formal versions ask that positions and goods be open to all, without legal barriers or direct discrimination, and substantive versions also look at the background conditions that decide who can use a formal opening.

Rawls's fair equality of opportunity is a prominent substantive version. It requires that people with the same talent and ability, and the same willingness to use them, have the same prospects of success whatever their social class of origin. Luck egalitarianism goes further and holds that nothing other than a person's own responsible choices should create differences in prospects. It rests on Ronald Dworkin's distinction between option luck, the outcome of deliberate gambles a person could have declined, and brute luck, the outcome of risks that were not chosen. For luck egalitarians, inequalities in brute luck are unfair and inequalities in option luck are fair.

John Roemer turned these ideas into a tool for policy. He separates circumstances, the parts of a person's environment beyond their control, from effort, and sorts the population into types that share the same circumstances. Equality of opportunity then requires equal outcomes across types for people who exert the same effort, while differences within a type that reflect effort may remain.

Formal, fair and luck-egalitarian versions in health care

Each version advises a health system differently. The Equality Act 2010 uses the phrase in its public sector equality duty, section 149, which requires a public authority to have due regard to the need to advance equality of opportunity between people who share a protected characteristic and people who do not. The section adds that compliance may involve treating some people more favourably than others.

Norman Daniels extended Rawls's principle to health. He argues that health, understood as normal functioning, protects the range of opportunities people can reasonably exercise, so obligations to protect opportunity imply obligations to promote and protect normal functioning for all. On this view access to public health and medical services should be universal and based on health need, under reasonable resource limits, with special weight on spreading the risks of disease equitably. Daniels notes that a luck-egalitarian version would also support universal health coverage but leaves more room to exclude conditions for which a person is substantively responsible.

Elford gives a UK health example, the postcode lottery, in which a drug is far less available in one county than in another. Framed as a failure of equality of opportunity, the complaint is that care should follow need and not the arbitrary fact of where a person lives, which links the principle to access to care and horizontal equity.

Why the principle breaks with welfarism

Welfarist welfare economics judges a policy by the utility levels it produces. Roemer and Alain Trannoy write that equal-opportunity theory sharply challenged that welfarist assumption, because a social judgement also needs to know the extent to which individuals are responsible for their outcomes. A social welfare function defined over outcomes alone cannot tell a gap caused by poverty in childhood from one caused by later choices. Brouwer, Culyer, van Exel and Rutten (2008) label any normative framework that does not restrict social welfare to individual utilities as extra-welfarist, so on their definition an equal-opportunity criterion falls outside welfarism.

Roemer and Trannoy show the difference with a health ministry that spends a fixed budget treating cancer and tuberculosis in a population of rich and poor types, where effort is lifestyle quality and the outcome is life expectancy. With 25% of people poor and a budget of USD 3,000 per head, the policy that equalises opportunity spends USD 686 per case of cancer and USD 13,027 per case of tuberculosis. The utilitarian policy, which maximises average life expectancy, spends USD 1,915 and USD 10,571, so it spends almost three times as much on cancer. At the equal-opportunity solution the rich still live longer, by less than a year. Both policies are Pareto efficient, and the authors call it a confusion to label the utilitarian answer the efficient one. Spending depends only on the disease, so the policy is horizontally equitable and no patient need ever be asked about type or lifestyle.

Measuring inequality of opportunity in health

Empirical work asks how much inequality is due to circumstances. Under Roemer's identification axiom, two people at the same quantile of their own type's outcome distribution count as having exerted the same effort. Roemer and Trannoy describe two measures that Fleurbaey and Schokkaert named. Direct unfairness is the inequality left once the effect of effort is removed, such as the inequality of mean outcomes across types. The fairness gap is total inequality minus the inequality left once each person is given the mean outcome of everyone with the same effort.

$$DU = \mathrm{Var}(\bar{y}_{t(i)}) \qquad FG = \mathrm{Var}(y_i) - \mathrm{Var}(\bar{y}_{e(i)})$$

where $DU$ is direct unfairness, $FG$ is the fairness gap, $y_i$ is the health outcome of person $i$, $\bar{y}_{t(i)}$ is the mean outcome of the type to which person $i$ belongs, $\bar{y}_{e(i)}$ is the mean outcome of everyone with the same effort as person $i$, and $\mathrm{Var}$ is the population variance. Roemer and Trannoy suggest that variance may be the better index for health, because it is an absolute measure and decomposes.

Studies reviewed by Roemer and Trannoy find inequality of opportunity in self-reported health between people of different parental background in the UK and in France, with types defined by the father's occupation. Using life histories of Europeans aged 50 and over, Bricard and colleagues found that inequality of opportunity accounted on average for half of the health inequality due to circumstances and lifestyles together. Health gaps of this kind are the raw material for distributional cost-effectiveness analysis.

Worked example with two backgrounds and one lifestyle choice

All numbers in this example are illustrative. A population has four groups of equal size. Type A grew up in advantaged households and type D in disadvantaged ones, and half of each type smokes. Because smoking is equally common in both types, absolute effort and Roemer's within-type rank coincide. The outcome is healthy life expectancy at birth.

1. Total inequality. The mean is (70 + 62 + 64 + 56) / 4 = 63. The squared deviations are 49, 1, 1 and 49, so the variance is (49 + 1 + 1 + 49) / 4 = 25.

2. Direct unfairness. Giving everyone the mean of their type leaves values of 66 and 60, each 3 years from 63, so $DU = 3^2 = 9$.

3. Fairness gap. Giving everyone the mean of their effort group leaves 67 and 59, each 4 years from 63, with variance 16. The fairness gap is $FG = 25 - 16 = 9$. The two measures agree here, and circumstances account for 9 / 25 = 0.36 of total inequality.

4. Two programmes of equal cost. Programme X, a cessation service, adds 3 years for every smoker in both types, a mean gain of 1.5 years per person. Programme Y, an early years and housing package, adds 2.4 years for everyone in type D, a mean gain of 1.2 years.

5. Effect on unfair inequality. Under X the type means become 67.5 and 61.5, still 6 years apart, so direct unfairness stays at 9 while total variance falls to 15.25. Under Y the type means are 66 and 62.4, so direct unfairness falls to $DU = 1.8^2 = 3.24$ while total variance is 19.24.

6. Interpretation. A decision maker who maximises total health prefers X, which gives 0.3 more years per person. Roemer's criterion, the average across effort levels of the worse-off type's outcome, gives 61.5 under X and 62.4 under Y, so it prefers Y. X narrows total inequality more because it narrows the smoking gap, which the principle treats as fair, while Y cuts the unfair part by 64%. The choice is a value judgement of the kind priced by the equity-efficiency trade-off.

How NICE and the NHS treat responsibility and circumstance

NICE's principles address personal responsibility directly. They say that NICE does not alter its normal approach because a condition may have been caused by a person's behaviour, though behaviour can be taken into account if it is likely to continue and to make a treatment less effective or cost effective. The same principles aim to reduce health inequalities, considering the protected characteristics in the Equality Act 2010 and inequalities that arise from socioeconomic factors. Under section 13G of the National Health Service Act 2006, NHS England must have regard to the need to reduce inequalities in people's ability to access health services and in the outcomes achieved for patients.

Under sections 6.2.35 to 6.2.38 of the NICE manual for technology appraisal (PMG36), committees can weigh substantial health inequality impacts, but they will not make optimised recommendations for subgroups based solely on social characteristics.

Boundaries with neighbouring equity concepts

Equality of opportunity is one principle within equity. Health equity states the goal that everyone should have a fair chance to reach full health, and health inequity labels the differences judged unfair. Equality of opportunity supplies one test for that judgement, the line between circumstance and choice. Horizontal equity, equal treatment for equal need, is a rule about delivering care that Roemer's health ministry keeps while pursuing equal opportunity for life expectancy. Vertical equity concerns how much more care greater need should receive.

Common misreadings and limits

Most errors make the principle look easier to apply than it is. Each misreading below has a counterpoint in the sources.

1. Lifestyle counts as effort. Elford notes that dispositions to make an effort can depend on brute luck, such as social circumstances, which is why Roemer measures effort within a type.
2. Measured inequality of opportunity is complete. Roemer and Trannoy note that circumstances left out of the definition of types are counted as effort, so a measured figure is a lower bound on the true inequality of opportunity.
3. Equalising opportunity is inefficient. Both policies in Roemer's health example are Pareto efficient, and they differ in the social objective.
4. The principle settles rationing. Daniels describes fair equality of opportunity as too general and indeterminate for many rationing problems, which is why he pairs it with a fair process, accountability for reasonableness.

Sources

• Brouwer WBF, Culyer AJ, van Exel NJA, Rutten FFH. Welfarism vs. extra-welfarism. Journal of Health Economics. 2008;27(2):325-338. https://doi.org/10.1016/j.jhealeco.2007.07.003
• Daniels N. Justice and access to health care. In: Stanford Encyclopedia of Philosophy. First published 29 September 2008, substantive revision 20 October 2017. Sections 3 and 4. https://plato.stanford.edu/entries/justice-healthcareaccess/
• Elford G. Equality of opportunity. In: Stanford Encyclopedia of Philosophy. First published 3 August 2023. Sections 1 to 3, 5, 6, 8 and 10. https://plato.stanford.edu/entries/equal-opportunity/
• Equality Act 2010, section 149 (public sector equality duty). https://www.legislation.gov.uk/ukpga/2010/15/section/149
• National Health Service Act 2006, section 13G (duty as to reducing inequalities), as inserted by the Health and Social Care Act 2012. Revised text up to date to 30 October 2026. https://www.legislation.gov.uk/ukpga/2006/41/section/13G
• National Institute for Health and Care Excellence. NICE technology appraisal and highly specialised technologies guidance: the manual (PMG36). Published 31 January 2022, last updated 13 October 2026. Sections 6.2.35 to 6.2.38. https://www.nice.org.uk/process/pmg36
• National Institute for Health and Care Excellence. Our principles. Last modified 31 March 2026. Principle 9. https://www.nice.org.uk/about-us/our-principles
• Roemer JE, Trannoy A. Equality of opportunity. JEL version, 18 November 2013. https://milescorak.com/wp-content/uploads/2014/06/roemer-trannoy-equality-of-opportunity-journal-of-economic-literature.pdf. Published in revised form as Equality of opportunity: theory and measurement. Journal of Economic Literature. 2016;54(4):1288-1332. https://doi.org/10.1257/jel.20151206