Wealth Ranking
- Introduced by
- Participatory Rural Appraisal tradition (Robert Chambers and colleagues)
- Year
- 1994
- Type
- Participatory stratification of households by locally defined wealth or wellbeing
- Family
- Participatory Rural Appraisal (PRA)
What does it do?
Wealth ranking is a participatory rural appraisal technique in which knowledgeable community members sort cards representing local households into a set of wealth or wellbeing strata that they themselves define.
Several informants each perform the sort independently, and because they may use different numbers of piles, their placements are converted to a common scale and averaged into a relative wealth score for every household. The procedure produces both a stratification of the community and, crucially, the local (emic) criteria people actually use to judge who is poor and who is well off.
The intuition behind the method
Outsiders tend to measure wealth in income or assets, but communities reckon it through their own signs — the kind of roof, the number of cattle, whether children go to school, who must sell labour to eat. Wealth ranking taps that local theory directly: by asking residents to sort their neighbours into 'piles' of similar standing and then explain what each pile has in common, you recover both a ranking and the reasoning behind it. Using several independent sorters and averaging their placements turns a set of subjective judgments into a stable relative score, while disagreements between sorters become informative rather than a nuisance.
History
Wealth ranking became a signature tool of participatory rural appraisal in the late 1980s and early 1990s, popularized through the work of Barbara Grandin and taken up in the PRA movement described by Robert Chambers in his 1994 account of its origins and practice. It answered a recurring problem in development work — how to identify the poor and stratify a community without an expensive income survey — by handing the judgment to local people and recording the criteria they used. Bernard's Research Methods in Anthropology treats it within the broader family of ranking and card-sorting techniques used to elicit cultural judgments.
When to use it
Use wealth ranking when you need to stratify a community by socioeconomic standing on local terms — for example to sample across wealth groups, to target or evaluate a programme, or to study inequality and poverty as the community understands it. It works best in small, bounded communities where informants know most households and where wellbeing is socially legible. It assumes a reasonably complete household roster and informants willing to judge their neighbours. It is less suitable for large anonymous populations, for settings where ranking neighbours is socially explosive, or when you require absolute income figures rather than relative standing.
Limitations and considerations
- Gives relative standing, not absolute income or asset values, so it cannot be compared directly across communities.
- Depends on informants knowing the households well; it breaks down in large or anonymous populations.
- Sorters apply different and partly idiosyncratic criteria, so aggregation can blur genuinely different dimensions of wellbeing.
- Asking residents to rank neighbours is socially sensitive and can be intrusive, biased by rivalry, or simply refused.
Source conditions do not replace independent scientific review or validation by an analysis engine.
Strengths
- Recovers emic criteria of wealth and poverty, revealing the community's own model of deprivation rather than an imposed metric.
- Produces a relative ranking of every household quickly and cheaply, useful for sampling, targeting, and stratified analysis.
- Using several independent sorters yields a stable averaged score and turns disagreement into a measure of contestation.
- Low-literacy and card-based, making it accessible and engaging for informants across very different contexts.
Applications
- Stratifying a community for representative or wealth-targeted sampling in surveys and ethnographic studies.
- Targeting and later evaluating poverty programmes, microfinance, or social transfers by relative standing.
- Eliciting local definitions and indicators of poverty and wellbeing for participatory poverty assessments.
- Studying social and economic differentiation and its links to behaviour, health, or resource access.
- Cross-checking household survey wealth measures against the community's own perceptions.
How it works
How it works
Working from a household listing or census of the community, the facilitator prepares one card per household, each bearing a name or identifier known to local informants. The card set defines the universe to be ranked and should cover every household so the resulting stratification is complete rather than partial.
Each of several informants, chosen for their broad knowledge of the community, independently sorts the full deck of cards into piles representing different levels of wealth or wellbeing, using as many piles as they see fit. They are not told how many strata to make; the number and meaning of the piles are theirs, capturing how that person carves up the social hierarchy.
After sorting, each informant is asked to describe what distinguishes the piles — what makes a household belong to the top, middle, or bottom group. These explanations yield the local criteria of wealth and poverty (land, livestock, labour status, dependants, food security, housing) that are often the most valuable output, revealing the community's own model of deprivation.
Because informants use different numbers of piles, each pile is rescaled to a common range and a household's placements are averaged across the sorters who ranked it: w̄_h = (1/n_h) Σ s_ih, where s_ih is the rescaled stratum that informant i assigned to household h and n_h is the number of informants who placed it. Households are then ordered by this average, and the spread of placements indicates how much sorters agreed.
Formula
\bar{w}_h = \frac{1}{n_h}\sum_{i} s_{ih}Common pitfalls
- Imposing a fixed number of strata instead of letting each informant choose, which suppresses their own model of the hierarchy.
- Recording only the final piles and skipping the criteria, discarding the richest and most defensible part of the data.
- Averaging across informants without rescaling piles to a common range, so a three-pile and a five-pile sort are wrongly equated.
- Ignoring the spread of placements and treating a contested household with high disagreement as if it were cleanly ranked.
Frequently asked questions
How many informants should do the sorting?
Typically several independent sorters — often around three to six knowledgeable residents — are enough to produce a stable averaged ranking and to reveal where judgments diverge. More sorters tighten the estimate and expose contested households, but each must genuinely know most of the households, so quality of knowledge matters more than sheer numbers.
How are different sorters' piles combined when they use different numbers of strata?
Each informant's piles are rescaled onto a common range (for example 0 to 100, top to bottom) so that a three-pile sort and a five-pile sort become comparable. A household's rescaled placements are then averaged across the informants who ranked it, w̄_h = (1/n_h) Σ s_ih, giving a relative wealth score; the variance of those placements indicates agreement.
Is wealth ranking ethical given that people are judging their neighbours?
It can be, but it demands care. Cards and results must be kept confidential, informants must consent and be free to decline, and in communities where ranking neighbours would cause conflict the method may be inappropriate. Done sensitively in small communities, it is often less intrusive than a detailed income survey because it builds on judgments people already make informally.
Related methods
Sources
Chambers, R. (1994). The origins and practice of participatory rural appraisal. World Development, 22(7), 953–969.
Open DOI referenceCopied.Bernard, H. R. (2017). Research Methods in Anthropology: Qualitative and Quantitative Approaches (6th ed.). Lanham, MD: Rowman & Littlefield.
ISBN: 9780759112421Copied.