Measuring development

• Explain why economic development has several dimensions and cannot be measured by income alone.
• Explain what each of the five groups of single indicators measures, with an example of each, and what it leaves out.
• Explain how the Human Development Index, the Inequality-adjusted HDI, the Gender Inequality Index and the Happy Planet Index are constructed and how a value of each is read.
• Evaluate the strengths and limitations of single and composite indicators as measures of economic development.
• Discuss the possible relationships between economic growth and economic development, using indicator values for named countries.

Economic development is a broad and lasting improvement in the living standards of a population. It includes higher incomes, and also lower poverty, better health, more and better education, and greater freedom for people to choose how they live and work. It is not the same as Economic growth, which is an increase in real GDP: growth describes how much is produced, while development describes what that output does for people's lives and how it is shared.

Development is called multidimensional because Economic well-being has many parts that do not move together automatically. Income matters, because it buys food, housing and security. A long and healthy life, knowledge, a clean and safe environment, equal treatment of women and men and a say in decisions matter in themselves, and none of them follows mechanically from income. A country's average income can rise while most households see none of the gain, while its air becomes unsafe to breathe, or while girls stay out of school.

The consequence for measurement is that no one statistic can capture development. Economists use two approaches. A Single indicator measures one aspect with one statistic, such as life expectancy at birth or GNI per capita. A Composite indicator combines several single indicators into one index number, such as the Human Development Index. Single indicators are precise about one thing; composite indicators summarise several.

Measurement is not neutral. Deciding what to count, and how much weight to give each part, is a judgement about what a good life requires. Two measures built on different judgements can rank the same countries in different orders, and neither ranking is the one true level of development. A figure for development should therefore be read together with what it includes and what it leaves out.

A single indicator reports one statistic about one aspect of development. Single indicators fall into five groups: income per person; health and education; economic and social inequality; energy; and the environment.

Income per person. GDP per capita is the value of the output produced within a country in a year divided by its population. GNI per capita is the income earned by the country's residents divided by its population, where Gross national income equals GDP plus net income from abroad. Where the two differ, GNI per capita is the better guide to residents' living standards. If much of the profit made in the country flows out to foreign owners, GDP overstates what residents receive; if residents earn substantial income from work or investments abroad, GDP understates it.

To compare countries, incomes have to be expressed in one currency, and the market exchange rate does this badly. Exchange rates are set by trade and financial flows, while much of what households buy, such as housing, transport and personal services, is never traded and is far cheaper in low-income countries. A dollar therefore buys more there than the exchange rate suggests. Conversion at Purchasing power parity (PPP) corrects for this: it uses the exchange rate at which a common basket of goods and services would cost the same in each country. GNI per capita at PPP is the standard single indicator of material living standards. For comparisons over time the figures must also be real, that is, adjusted for inflation.

Income per person shows residents' average command over goods and services, and it is closely related to most other indicators, because income pays for food, clinics and schools. What it cannot show is as important. It is an average, so it hides the distribution. It leaves out production that is not sold in markets, such as subsistence farming and unpaid work at home, which is a large part of activity in low-income countries. And it does not show what was produced, what environmental damage the production caused, or whether people are healthy and educated.

The World Bank uses GNI per capita to sort countries into four income groups, which is where terms such as 'low-income country' and 'middle-income country' come from.

Health indicators measure how long and how well people live. Life expectancy at birth is the summary measure. The Infant mortality rate (deaths before the age of one per 1,000 live births), the under-five mortality rate and the maternal mortality ratio (deaths of mothers per 100,000 live births) are sensitive to the conditions of the poorest households, because most of these deaths can be prevented with clean water, adequate food and basic care. Other health indicators measure access: doctors per 1,000 people, the share of births attended by skilled staff, and access to safe water and sanitation.

Education indicators measure knowledge and skills. The Adult literacy rate reports the share of adults who can read and write. Mean years of schooling reports how long today's adults spent in school, and expected years of schooling how long a child starting school now can expect to stay. Enrolment and completion rates show how many children are in primary and secondary school and how many finish. Outcome indicators such as life expectancy and literacy show results; input indicators such as spending, numbers of doctors and enrolment show effort. Neither kind measures quality: a year in school counts the same whatever was learned in it.

Inequality indicators show whether the average is typical. Economic inequality is measured by the Gini coefficient, which runs from 0 (everyone has the same income) to 1 (one person has it all) and is derived from the Lorenz curve; by the share of income received by each fifth or tenth of the population; and by the share of people living below a poverty line. The World Bank's International poverty line for extreme poverty has been $3.00 per person per day at 2021 PPP since June 2025. Social inequality is measured by gaps between groups, for example in literacy, school enrolment, life expectancy or participation in paid work between women and men, between town and countryside, or between ethnic groups. These indicators say nothing about the level of living standards: a country can be equal and poor.

Energy indicators include the share of the population with access to electricity, the share with access to clean cooking fuels, energy consumption per person and the share of energy that comes from renewable sources. Access to energy indicates development because of what it makes possible: light to study by, refrigeration for food and vaccines, machinery, pumped water and communications. Households without it burn wood and charcoal, which damages their health and takes hours of collecting, mostly by women and girls. Energy consumption per person is harder to read: high use signals high output, and also high emissions unless the energy is clean.

Environmental indicators include carbon dioxide emissions per person, air pollution (the concentration of fine particles), forest cover and the rate of deforestation, the share of land that is protected, and the Ecological footprint per person. They show whether development is sustainable, which no income, health or education indicator does. They often move against the other groups: as an economy industrialises, income and energy use rise while air quality and forest cover can fall.

Single indicators can point in different directions for the same pair of countries, so the choice of indicator shapes the conclusion. In the table below Nigeria is ahead on income and schooling, and Rwanda is ahead on life expectancy, maternal survival and women's representation.

For this reason single indicators are best read as a dashboard: several together, each compared over time and beside a measure of its distribution, so that a strength in one does not hide a weakness in another.

The Human Development Index (HDI) is the best-known composite indicator. It is published by the United Nations Development Programme (UNDP) and summarises three dimensions of development with four indicators.

1. Each indicator is turned into an index between 0 and 1 by comparing it with its goalposts: index = (actual value − minimum) ÷ (maximum − minimum). A life expectancy of 72 years would give (72 − 20) ÷ (85 − 20) = 0.8; the 72 is an invented figure used to show the method.
2. The education index is the arithmetic mean (the simple average) of the two schooling indices.
3. For income, the natural logarithm of GNI per capita is used, so each extra dollar adds less to the index as income rises. This reflects the diminishing contribution of income to well-being. Income above the maximum of $75,000 adds nothing; nine countries were above it in 2023.
4. The HDI is the geometric mean of the three dimension indices, that is, the cube root of their product.

A geometric mean is used so that a weak dimension cannot be fully offset by a strong one. Suppose the three indices were 0.9, 0.9 and 0.3 (invented values). A simple average would give 0.70, while the geometric mean gives about 0.62. The low score pulls the index down further, so a country cannot make up for poor health or schooling with high income alone.

An HDI lies between 0 and 1, and a higher value means higher human development. In the data for 2023 Iceland ranked first with 0.972 and South Sudan last, 193rd, with 0.388; the value for the world was 0.756. UNDP also publishes each country's GNI per capita rank minus its HDI rank. A positive number means that the country does better on human development than its income alone predicts, and a negative number the reverse.

The HDI leaves a good deal out. By UNDP's own account it does not reflect inequality, poverty, human security or empowerment, and it has no environmental component. Each of the other three composite indicators supplies one of the missing pieces.

The Inequality-adjusted Human Development Index (IHDI) (IHDI) deals with inequality. The HDI is built from national averages, so two countries with the same averages get the same HDI even if, in one of them, health, schooling and income are concentrated in a small group. The IHDI discounts each of the three dimension indices according to how unequally that dimension is distributed across the population, and then combines the three adjusted indices, again as a geometric mean.

The IHDI equals the HDI when there is no inequality and falls further below it as inequality rises. The percentage gap, (HDI − IHDI) ÷ HDI, is called the loss in human development due to inequality. The HDI can be read as the level a country could reach if everyone shared the average, and the IHDI as the level actually achieved.

South Africa's HDI is not far below Brazil's, but more than a third of it disappears once the distribution is counted, while Iceland loses a twentieth. The IHDI has limits of its own. It adjusts each dimension separately, so it cannot show whether the same people are deprived in all three. It also needs household-survey data, which come from different years and are missing for many countries.

The Gender Inequality Index (GII), also published by UNDP, measures inequality between women and men in three dimensions: reproductive health, empowerment and the labour market. It runs from 0, where women and men fare equally, to 1, where one gender fares as poorly as possible in every dimension. A higher value is therefore worse, the opposite of the HDI and the IHDI.

Reproductive health is measured by outcomes that affect only women, such as the maternal mortality ratio. Empowerment is measured by women's place in public life, for example the share of seats in parliament that women hold. The labour-market dimension compares how far women and men take part in paid work. The index measures the gap, not the level of development: a poor country in which women and men fare alike can have a better GII than a rich one in which they do not.

Gender inequality is measured separately because national averages conceal it and because it holds development back. When girls leave school early and women are kept out of paid work and out of decisions, the human capital of half the population is underused, so output and incomes are lower than they could be. A mother's education and income also carry forward: her children are more likely to be well fed, healthy and in school, so inequality today lowers the development of the next generation.

The data for 2023 show how different the picture can be from the HDI's. Denmark ranks first on the GII with 0.003 and Yemen 172nd with 0.838. Nigeria ranks 171st with 0.677: its maternal mortality ratio is 1,047 deaths per 100,000 live births, and women hold 3.6% of the seats in its parliament. Rwanda ranks 159th on the HDI but 99th on the GII, with 0.394, and women hold 54.7% of its parliamentary seats.

The GII has limits. A high share of parliamentary seats does not by itself show that women have economic power, yet it can move a country's rank a long way, as it does for Rwanda. The index has no measure of unpaid care work or of violence against women. And because part of it records outcomes for women only, it mixes absolute conditions with gaps between women and men.

The Happy Planet Index (HPI) is published by the Hot or Cool Institute, a think tank, and not by the United Nations. It asks a different question from the other three: how efficiently a country turns its use of the planet's resources into long and satisfying lives. It is a measure of sustainable well-being, not of output.

In the 2026 edition the index is approximately life expectancy multiplied by experienced well-being, divided by the Ecological Footprint per person, with statistical adjustments. Life expectancy at birth comes from United Nations data. Experienced well-being is the average answer to a survey question in the Gallup World Poll, which asks people to place their own life on a ladder from 0 to 10. The Ecological footprint per person is the area of productive land and water needed to supply what the average resident consumes and to absorb the waste. A country scores well by achieving long, satisfied lives with a small footprint.

The method has changed between editions. The 2016 edition also adjusted for inequality, and the 2024 edition is reported to have used a carbon footprint in place of the Ecological Footprint. Older descriptions of the index, including those that list four elements, are therefore out of date, and scores cannot be compared from one edition to the next.

The results differ sharply from a ranking by income or by HDI. The 2026 edition covers 134 countries. Costa Rica ranks first, although it is 62nd on the HDI; Tajikistan is third; and Spain, in fourth place, is the highest-scoring high-income country. Wealthy countries such as Singapore, Luxembourg and the USA perform poorly because of their environmental impact, and no country achieves high life expectancy and high well-being within environmental limits.

The HPI adds what the other three lack, sustainability and people's own assessment of their lives, and it omits what they contain. It has no income or education component and says nothing about inequality, political freedom or human rights, so a high rank does not mean that a country has no serious problems. Reported well-being is influenced by culture and expectations. And one footprint figure cannot capture every kind of environmental damage.

One further composite measure is often confused with these four. The Multidimensional Poverty Index counts the people who are deprived in several aspects of health, education and living standards at once. It measures multidimensional poverty, not development as a whole, and it belongs with inequality and poverty (topic 3.4).

Single and composite indicators are two approaches to the same task, and each is strong where the other is weak. Which is the better measure depends on the purpose: a broad comparison between countries, or a diagnosis that can guide policy.

A single indicator is simple, precise about one thing and widely available, often every year. It shows exactly where a problem lies, so it is the right tool for aiming a policy and judging whether it worked: a programme to reduce deaths in childbirth is assessed with the maternal mortality ratio, not with the HDI. Its weakness is that it is partial. A country can do well on one indicator and badly on another, as Rwanda and Nigeria each do, so a conclusion drawn from one statistic turns on which statistic was chosen.

A composite indicator reflects several dimensions of development in one number that can be ranked, tracked over time and reported, which is why the HDI draws more public attention than any table of statistics. Its greatest strength is that it shows what income conceals.

The same case shows the main limitation. Sri Lanka and Guyana share an HDI of 0.776, but Guyana reaches it with almost four times the income, a life expectancy more than seven years shorter and about two fewer years of schooling. The index hides which component is weak, so it cannot guide policy on its own. Two further limitations follow from the way composites are built. The choice of components and their weights is a value judgement: the HDI gives health, education and income equal weight, and a different weighting would give a different ranking. And every index omits something, so different composites rank the same country differently: Costa Rica is 62nd on the HDI and first on the Happy Planet Index.

Some limitations apply to both approaches. The data are weakest where they matter most: many low-income countries hold censuses and surveys infrequently and have large informal sectors, so their figures are estimates. Figures arrive late: the HDI published in 2025 uses data for 2023. PPP conversion is imprecise. National averages conceal differences between regions, ethnic groups, women and men, and town and countryside, which the IHDI and the GII only partly correct. And both approaches measure what is easy to count, so the quality of schooling, of health care and of institutions, and political freedom, are largely missing. Small differences in a score or a rank therefore mean little.

Who uses the measure, and over what period, also matters. A health or education ministry needs single indicators, updated every year, to direct its spending. International organisations and aid donors comparing countries need a composite, and they need distribution-sensitive measures beside it if help is to reach the people who are poor and not only the countries that are poor on average. Over a few years only single indicators such as school enrolment or access to electricity move enough to show whether a policy is working; composite indices built on life expectancy and years of schooling show change over decades.

Neither approach is sufficient alone. A composite indicator is the better summary of how developed a country is; single indicators are the better guide to why, and to what should be done. The strongest measurement uses a composite for the overview, its components and other single indicators for the detail, and inequality-adjusted measures beside the averages.

Economic growth and economic development are related, but neither guarantees the other. Three relationships are possible: growth can promote development, growth can occur without development, and development can itself cause growth.

A production possibilities curve shows why. Put health care and education on the vertical axis and all other goods on the horizontal axis. Growth shifts the curve outward, from PPC1 to PPC2, so that more health care and education and more of other goods are attainable together. The shift does not decide which point on PPC2 the economy chooses: it can move to a point with more of both, or to a point with more of other goods and no more health care and education than before. Growth widens the choice; how income is distributed and how the government spends decide where the economy ends up.

Growth can promote development through three channels. Higher household incomes let families buy more and better food, housing, schooling and health care. Higher incomes and profits raise tax revenue, which lets the government pay for schools, clinics, clean water, infrastructure and safety nets; this channel needs a tax system that collects the revenue and a government that chooses to spend it on those things. And growth in sectors that employ many people creates jobs, which is the most direct way out of poverty.

Growth may fail to bring development for four reasons. If the gains go mainly to those who are already well off, average income rises while the poor see little change. If growth comes from a capital-intensive sector with few links to the rest of the economy, such as mining or oil, it creates few jobs and the revenue may be captured by a narrow group. If production creates negative externalities, pollution and resource depletion damage health now and reduce what future generations can produce. And where institutions are weak or corrupt, government revenue does not reach public services. Growth is therefore not sufficient for development.

Development can run ahead of growth for a time. Moving existing resources towards primary health care, basic schooling, clean water and the education of girls raises health and education outcomes at a given income, which is how some countries reach human development well above what their income predicts. This is hard to sustain for long, because public services have to be paid for, and without rising output the resources run out. Over long periods growth is close to necessary for development.

The relationship also runs from development to growth. Healthier and better-educated people are more productive, so investment in Human capital raises potential output. Lower inequality widens the domestic market and reduces the risk of social conflict. An economy that develops its people may therefore grow faster later.

The UNDP data for 2023 illustrate each relationship. Sri Lanka and Equatorial Guinea have almost the same GNI per capita, yet Sri Lanka's HDI is 0.776 and Equatorial Guinea's 0.674: similar income has been turned into very different development. Nigeria's GNI per capita of 5,569 (2021 PPP $) is nearly double Rwanda's 2,971, yet Rwanda has a life expectancy of 67.8 years against 54.5 and a slightly higher HDI, 0.578 against 0.560. Guyana shows the role of time: its HDI rose from 0.722 in 2020 to 0.776 in 2023 as its income jumped, but its GNI per capita rank minus its HDI rank is −54, because health and schooling have not caught up with income. Qatar (−39) and Kuwait (−29) also have HDI ranks well below their income ranks.

The evidence supports a clear position: over the long run growth is generally necessary for development, and it is never sufficient. Whether growth becomes development is decided by how the income is distributed, what the government does with the revenue, what kind of production generates it and whether it damages the environment.