Human development and the capabilities approach

Development Economics, lecture 3

Prof. Dr. Claudius Gräbner-Radkowitsch

Department of Pluralist Economics, Europa-Universität Flensburg

2 October 2026

Where we are

Last week: the theory-ladenness of observation

Measurement involves choices — concept, conception, indicator, aggregation
The HDI as a prominent illustration
Today: whose conception is that — and how does the theory relate to the indicator?

The capability approach is the theory behind the HDI — and it illustrates its achievements and drawbacks as an operationalization.

Income is a means. Today is about the end

Why income is the wrong yardstick — but far from being unimportant
Functionings, capabilities, agency — three concepts you will reuse a lot
Whether a universal list of capabilities is possible — or dangerous
What the HDI makes of all this, and what it drops on the way

Halfway through you write your own list of capabilities — and compare against Nussbaum’s.

From income to capabilities

Same income, different lives

Person A

30,000 EUR a year

Healthy, lives in a city with good public transport

Person B

30,000 EUR a year

Mobility impairment, lives in a rural area with no accessible transport

Same income, different situation → it matters not what you have but what you can do.

The approach was built against the income view

Portrait of Amartya Sen

Amartya Sen (*1933)

The central concepts: freedom, functionings, plurality, democracy. Nobel Prize 1998.

Portrait of Martha Nussbaum

Martha Nussbaum (*1947)

A fuller account of justice based on dignity, thresholds, and central capabilities.

Both writing against the post-war consensus: development = growth of income.

Sen (1990); Nussbaum (2013). Photos: Elke Wetzig (Elya) and Robin Holland, CC BY-SA 3.0, via Wikimedia Commons; cropped and converted to greyscale.

Resources are means, never ends

Resources: income and wealth · food, goods, land · access to public services and networks
They are an input to a life, never the point of one
What a resource yields depends on the person and the place

Two people with the same basket can end up with entirely different lives.

Conversion factors decide what a resource is worth to you

Personal

Metabolism, disability, literacy, age

Social

Public services, discrimination, social norms

Environmental

Climate, geography, infrastructure

A bicycle is mobility — unless you cannot ride, may not go out alone, or there is no road.

Functionings are what you actually are and do

Functionings: the beings and doings a person achieves
Being well-nourished · being educated · being healthy · taking part in community life · being mobile
They are achievements — ends in themselves, not instruments

The list is open. Which functionings count is already a value judgement.

Capabilities are the functionings you could have chosen

Capability: the real freedom to achieve a functioning
Capability set: every combination of functionings open to you

Person A fasts

Chooses not to eat

Person B starves

Cannot eat

Identical functioning, distinct capability sets — freedom to choose is valuable in itself.

The same distinction separates frugality by choice from frugality as a social condition — Muradian (2019)

Agency: development cannot be done to people

Done to people

Patients receive medicine
Students are taught
The poor receive charity

Done by people

Agency: people choose which functionings to pursue
Participants, not recipients
Change lasts when people want it

The outcomes can be identical. The process is not — and the process is part of the outcome.

The whole framework in one picture

Income enters on the far left. Everything that makes it matter happens to its right.

Your turn

Advise the UN: which capabilities are essential?

Alone, 5 min — list 5–7 capabilities every person should have
Mark each one: a universal right, or culturally specific?
Groups of 3–4, 5 min — where do you agree? Where not?
Plenary, 5 min — one capability, one disagreement, your position

Worksheet: course website → session 3. Write your own list before you read any further — that is the whole point of doing it now.

The Sen–Nussbaum debate

Nussbaum: without a list, dignity has no teeth

Portrait of Martha Nussbaum

Vagueness protects the powerful — no list, no accountability
Human dignity is universal, so the capabilities it requires must be too
Without thresholds, deprivation goes unnoticed and unmeasured

A list is what turns a philosophical commitment into a constitutional claim.

Nussbaum (2013). Photo: Robin Holland, CC BY-SA 3.0; cropped, greyscale.

Nussbaum’s ten central capabilities

Life — a normal lifespan
Bodily health — nourishment, shelter
Bodily integrity — movement, security, reproductive choice
Senses, imagination, thought — education, expression
Emotions — to love, to grieve

Practical reason — to form a conception of the good life
Affiliation — to live with others, with social respect
Other species — concern for nature
Play — recreation, enjoyment
Control over one’s environment — political and material

Each is a threshold, not a quantity to be maximised — and none may be traded against another.

Sen: a fixed list is exactly the danger

Portrait of Amartya Sen

Societies face different problems and hold different values
A list written elsewhere is paternalism — historically, cultural imperialism
Reasonable people disagree; the selection belongs to public reasoning

Sen offers a method where Nussbaum offers a list — which is also the objection to it.

Photo: Elke Wetzig (Elya), CC BY-SA 3.0; cropped, greyscale.

Sen still names five instrumental freedoms

Not a definition of the good life — the freedoms that make the others attainable.

Freedom	Content
Political freedoms	Civil rights, free speech, elections
Economic facilities	Market access, employment, credit
Social opportunities	Education, healthcare, social services
Transparency guarantees	Information, accountability
Protective security	Safety nets, disaster relief

They are mutually reinforcing: each one makes the others easier to secure.

Sen (2000), chapter 2

The debate decides what agencies actually do

Can you write policy, or fund a programme, without a target list?
Can you defend human rights without imposing values?
Who is accountable when the list is only a method?

The HDI is what this compromise looks like once it has to be published every year.

Capabilities in practice: the HDI

The HDI was built to beat GDP in the newspaper

Created by Mahbub ul Haq with Sen; first Human Development Report 1990
The goal was political: an alternative to GDP simple enough to be quoted
Sen’s own objection: any single number is crude. He helped build it anyway
The task: one number out of an irreducibly plural theory

Three dimensions — chosen where data exist for nearly every country.

Three dimensions, four indicators, one number

Dimension	Indicator	Goalposts
A long and healthy life	Life expectancy at birth	20 – 85 years
Knowledge	Expected years of schooling	0 – 18 years
	Mean years of schooling	0 – 15 years
A decent standard of living	GNI per capita (PPP $)	100 – 75,000

Income enters as a logarithm — diminishing returns, by construction

The goalposts are conventions. Change them and every country’s score moves.

The geometric mean says capabilities cannot substitute

Country A

Health 0.70 · Education 0.70 · Income 0.70

Arithmetic: $\frac{0.70+0.70+0.70}{3} = 0.700$

Geometric: $\sqrt[3]{0.70 \cdot 0.70 \cdot 0.70} = \mathbf{0.700}$

Country B

Health 0.95 · Education 0.95 · Income 0.20

Arithmetic: $\frac{0.95+0.95+0.20}{3} = 0.700$

Geometric: $\sqrt[3]{0.95 \cdot 0.95 \cdot 0.20} = \mathbf{0.565}$

Same arithmetic mean, 0.135 apart on the geometric one — income cannot buy off dying young.

The computation in practice

Benin — a country of about 14 million on the West African coast, HDI rank 173 of 193 — for 2023.

Dimension	Indicator	Value	Index
Health	Life expectancy at birth (years)	60.8	0.627
Education	Expected years of schooling	10.4	0.580
Education	Mean years of schooling	3.2	0.213
Standard of living	GNI per capita (2021 PPP $)	3,806	0.550

Each indicator is first rescaled to $[0,1]$ against the goalposts on the previous slide
The two schooling indices are then averaged: $(0.580 + 0.213)/2 = 0.397$

\[\text{HDI} = \sqrt[3]{0.627 \cdot 0.397 \cdot 0.550} = 0.515\]

Own calculation from UNDP, Human Development Report 2025, statistical annex, table 1 (2025).

Income predicts human development — with diminishing returns

GDP = (important) means, not measure of capabilities
Same GDP can hide capability losses: distribution, public provision, conversion
Growth-oriented policies can shrink capabilities
- Less leave, rest, care, time autonomy,… (e.g. holidays, children’s sick days)

The same countries, ranked twice, do not agree

UNDP, Human Development Report 2025, statistical annex, table 1 (2025).

What that gap is actually telling you

Cuba ranks 30 places higher on HDI than on income
Qatar 39 places lower
Same countries, same year — two orderings
Theory-ladenness with a number attached

The gap is not measurement error: capability conception ≠ income conception.

The standard criticisms stick

Equal weights on the three dimensions are a value judgement dressed as arithmetic
Ravallion (2012): the implied trade-offs are indefensible — the HDI prices a year of life far lower in poor countries
The goalposts are conventions, and the country set changes when they move

Conventions presented as measurement. Pragmatism is fine; leaving it implicit is not.

The UNDP’s own answer is a dashboard

IHDI

Discounts each dimension by its inequality

GDI · GII

Female vs. male HDI; reproductive health, empowerment, labour

MPI

Overlapping deprivations, measured on households

PHDI

HDI discounted by CO₂ and material footprint per capita

Every addition is an admission about the original.

Example for where good theory facilitated new and better measurement

Expand every abbreviation out loud. The boxes carry the acronym only, and this is the slide where the room quietly loses the thread.

IHDI — Inequality-adjusted HDI (2010). Each of the three dimension indices is discounted by how unequally that dimension is distributed within the country. With no inequality IHDI = HDI, so the gap between the two is the price of the inequality.
GDI — Gender Development Index. The female HDI divided by the male HDI: the same three dimensions, computed twice. 1.000 is parity.
GII — Gender Inequality Index. Neither a ratio nor on the HDI scale: a composite of maternal mortality and adolescent births (reproductive health), parliamentary seats and secondary education (empowerment), and labour-force participation. 0 is no inequality, 1 is maximal.
MPI — Multidimensional Poverty Index, built with OPHI in Oxford. Ten indicators across health, education and living standards, counted per household in survey microdata. You count as poor if deprived in at least a third of them, weighted.
PHDI — Planetary pressures-adjusted HDI (2020). The HDI discounted by CO₂ emissions and material footprint per capita.

What distinguishes them — the part worth drilling, and the part the exam catches people on:

Direction and scale. IHDI and PHDI sit on the HDI scale and collapse back onto the HDI once the problem disappears; GDI is a ratio around 1. For all three, higher is better. GII and MPI are their own scales on which higher is worse. So most of these numbers cannot be compared with each other at all — only with the HDI they were derived from.
GDI against GII is the trap. Same subject, incompatible construction. GDI asks: one index, how do the two groups score on it? GII asks: how much is lost to gender inequality — and builds that from indicators with no male counterpart, since nobody has a male maternal mortality ratio. A country can therefore look respectable on one and poor on the other, and that is not an error in the data.
MPI is the only one built from household microdata. The other four are assembled from country aggregates, so the MPI alone can see two deprivations landing on the same person. That is what “overlapping” means, and it is the thing an average can never recover.
PHDI is the odd one out. The other four ask who got the development. PHDI asks what producing it cost — which is why it, and not the others, comes back in session 10.

Gender is where the capability approach bites hardest

What the approach sees

Adaptive preferences — internalised subordination is not a preference
Intrahousehold inequality — household income hides who eats
Reproductive autonomy — control over one’s body as a capability
Care work — unpaid work costs capabilities elsewhere

What the GII measures

Reproductive health — maternal mortality, adolescent births
Empowerment — parliamentary seats, secondary education
Economic status — labour force participation

Three of the four concerns on the left have no column on the right.

What we learned

Where capability thinking already sits

The SDGs — multidimensional by construction
Social protection and impact evaluation — participatory designs ask what people value
Constitutional rights — South Africa and India use capability language
Far more influential internationally than in German domestic policy

You will meet the vocabulary again in many development documents you read.

What to take away

Resources are means — conversion factors decide what they are worth
Capability, not functioning — the freedom, not only the outcome
Agency — development that is done to people is not development
The HDI is a compromise — useful, incomplete, and honest about neither unless you make it so

Before next week

Do the exercises for this session — course website → session 3
Read the Overview of the World Bank’s World Development Report (2024) — the mandatory reading for session 4

Next: growth models. Income is not the end of development — but it is still important and at the centre of many theories.

References

Muradian, R. (2019) “Frugality as a choice vs. Frugality as a social condition. Is de-growth doomed to be a Eurocentric project?” Ecological Economics 161: 257–260.

Nussbaum, M. C. (2013) Creating capabilities: The human development approach, Cambridge, Massachusetts London, England: Harvard University Press.

Ravallion, M. (2012) “Troubling tradeoffs in the Human Development Index,” Journal of Development Economics 99(2): 201–209, doi:10.1016/j.jdeveco.2012.01.003.

Sen, A. (1990) “Development as Capability Expansion,” in K. Griffin and J. Knight (eds.). Human development and the international development strategy for the 1990s. Basingstoke, Hampshire: Macmillan [u.a.], pp. 41–58.

Sen, A. (2000) Development as Freedom, New York: Alfred Knopf.

UNDP (ed.) (2025) A matter of choice: People and possibilities in the age of AI, New York: United Nations Development Programme (UNDP), available at https://hdr.undp.org/content/human-development-report-2025.

World Bank (2024) World Development Report 2024: The Middle-Income Trap, Washington, DC: World Bank, available at https://doi.org/10.1596/978-1-4648-2078-6.