The ordinal approach
Utility is ranked, not measured.
Do we need to measure satisfaction in utils to explain a consumer's choice?
The ordinal approach ranks bundles instead of counting utils, and finds the choice with an indifference curve and a budget line.
Ask a shopper how many utils a masala dosa gives her and you get a blank look. Ask whether she would rather have a dosa or a plate of idli, and she answers at once.
In plain wordsThe ordinal approach explains choice using only rankings: which of two bundles a consumer prefers, or that she likes them equally. It never asks by how much.
Exam definitionA standard textbook definition: in the ordinal approach, utility cannot be measured but can be ranked. The consumer can say whether she prefers one bundle to another or is indifferent between them.
J. R. Hicks and R. G. D. Allen set out the approach in 1934. They built on indifference curves, which F. Y. Edgeworth and Vilfredo Pareto had used earlier.
An example. A bundle is a combination of goods. For a student's week, one bundle is 3 samosas and 6 cups of chai; another is 2 samosas and 9 cups. She need only say "I prefer the first", "I prefer the second" or "either is fine".
The approach uses two tools: the indifference curve, which shows what she likes, and the budget line, which shows what she can afford. Her choice is where the two meet.
Its assumptions.
- Rational consumer: she seeks the most satisfaction her income allows.
- Ordinal utility: she can rank bundles but cannot measure satisfaction.
- Complete preferences: she can compare any two bundles.
- Transitivity: if she prefers A to B and B to C, she prefers A to C.
- Consistency: if she prefers A to B, she does not also prefer B to A.
- Non-satiation: more of a good is preferred to less.
- Diminishing marginal rate of substitution: the more she has of one good, the less of the other she will give up for it.
- Divisible goods, and no cardinal measurement of utility.
Don't confuse ranking with measuring. "I prefer A to B" is ordinal. "A gives twice the satisfaction of B" is cardinal. The ordinal approach needs only the first; the table under Consumer equilibrium compares the two approaches.
Where it breaks. Behavioural experiments find that choices shift with how options are presented, and are sometimes intransitive. The approach also assumes she can rank every bundle.
For exams. List the assumptions, name Hicks and Allen, and contrast the approach with the cardinal one.
Go deeper for practitioners
Revealed preference. Paul Samuelson (1938) went a step further. His revealed preference theory reads preferences from what consumers actually buy, without asking them to rank anything. If a buyer chooses bundle A when bundle B was also affordable, she has revealed that she prefers A.
For a firm, this is the difference between survey answers and purchase data: the first reports preferences, the second reveals them.
1. Which assumption says a consumer can compare any two bundles?
Not quite. The answer is “Complete preferences”. Completeness means every pair can be ranked or judged equal. Transitivity, the tempting choice, is about a chain of three: A over B and B over C means A over C.
2. A shopper prefers a masala dosa to a plate of idli but cannot say by how much. Which approach can still explain her choice?
Not quite. The answer is “The ordinal approach, which needs only rankings”. The ordinal approach asks only which bundle she prefers. The cardinal approach needs her to say by how much, in utils or in money, which she cannot.
3. A shopper prefers tea to coffee, coffee to juice, and juice to tea. Which assumption has she broken?
Not quite. The answer is “Transitivity”. Preferring tea to coffee and coffee to juice should mean preferring tea to juice. Completeness still holds: she has ranked every pair.
- Documented
Ranking preferences and the better and the pleasant
The Kaṭha Upaniṣad sets out a ranking, not a measurement: the good (śreyas) and the pleasant (preyas) both present themselves, and the wise person examines them and chooses. It is a classical statement of ordinal choice, where the question is which option is preferred.
श्रेयश्च प्रेयश्च मनुष्यमेतस्तौ सम्परीत्य विविनक्ति धीरः । śreyaś ca preyaś ca manuṣyam etas tau samparītya vivinakti dhīraḥ. Kaṭha Upaniṣad 1.2.2
What it addsA ranking, not a measurement: the good and the pleasant both present themselves, and the wise person examines them and chooses.
Where they differThe Kaṭha Upaniṣad concerns the choice between the good and the pleasant. The chapter's ordinal choice is between bundles of two goods.
“Rank three things you could spend a free evening on. Could you say by how much one beats another?”
Still unsure, or see it differently? Ask the community about thisYou’ll sign in first if you haven’t. Your note above stays private.
- The ordinal approach ranks bundles; it never measures satisfaction.
- Hicks and Allen set it out in 1934. Its tools are the indifference curve and the budget line.
- Assumptions: a rational consumer, ordinal utility, complete, transitive and consistent preferences, non-satiation, diminishing MRS and divisible goods.
- Revealed preference (Samuelson, 1938) reads preferences from what consumers actually buy.
Cardinal asks how much. Ordinal asks which is preferred.