Transitivity and Consistent Bundle Rankings

Why Rankings Must Fit Together

A consumer may compare bundles two at a time: one bundle is at least as good as another, or perhaps strictly better. But pairwise judgments should also fit together. If bundle A is at least as good as bundle B, and B is at least as good as C, a consistent ranking cannot place C strictly above A. Transitivity is the axiom that formalizes this requirement.

This is a logical condition on preferences, not a claim that every bundle has an obvious numerical score. The bundles below can contain different quantities of two goods. Unless a comparison is stated or follows from an axiom, their coordinates alone do not tell us which bundle is preferred. In particular, this tutorial does not assume that having more of one good automatically makes a bundle better.

Definition: A preference relation is transitive if, for any three bundles A, B, and C, whenever A is weakly preferred to B and B is weakly preferred to C, A is weakly preferred to C. In notation: $$ A \succeq B \ \text{and}\ B \succeq C \quad\Longrightarrow\quad A \succeq C. $$ Here, $A \succeq B$ means that A is at least as good as B.

The weak preference symbol matters. It allows the first comparison to be a tie: $A \succeq B$ does not necessarily mean that A is strictly better than B. As in the earlier material on weak preference, strict preference is written $A \succ B$, and indifference is written $A \sim B$. Specifically, $A \succ B$ means $A \succeq B$ but not $B \succeq A$; $A \sim B$ means both $A \succeq B$ and $B \succeq A$.

What Transitivity Does—and Does Not—Say

Transitivity carries a ranking through a chain. If A is at least as good as B, and B is at least as good as C, then A must be at least as good as C. This implication applies regardless of whether one or both of the given comparisons are ties. It does not say that A must be strictly better than C.

Transitivity is distinct from completeness, the other ranking axiom introduced in the earlier material on completeness. Completeness says that any two bundles can be compared: for any A and B, either $A \succeq B$, or $B \succeq A$, or both. Transitivity says how comparisons must fit together once they are known. A transitive preference relation need not rank every possible pair.

For example, suppose we know only $A \succeq B$ and $B \succeq C$. We can infer $A \succeq C$, but we cannot conclude that $A \succ C$. All three bundles could be indifferent, or A could be strictly preferred to B and B strictly preferred to C. The premises do not tell us which of these possibilities applies.

Worked Example: Carrying a Weak Ranking Across Three Bundles

A shopper compares three hypothetical weekly grocery bundles. Let A contain 2 prepared meals and 1 bag of fruit, B contain 1 prepared meal and 3 bags of fruit, and C contain 3 prepared meals and 2 bags of fruit. The shopper reports $A \succeq B$ and $B \succeq C$. What comparison must hold between A and C? Does the information establish that A is strictly preferred to C?

Step 1: Identify the applicable method. Use transitivity of weak preference. The stated comparisons form a chain: A is weakly preferred to B, and B is weakly preferred to C.

Step 2: Apply the axiom. The transitivity implication is $$ A \succeq B \ \text{and}\ B \succeq C \quad\Longrightarrow\quad A \succeq C. $$ Therefore, A must be at least as good as C.

Step 3: Check what is not established. The conclusion is $A \succeq C$, not necessarily $A \succ C$. For example, the stated comparisons would also hold if $A \sim B$, $B \sim C$, and $A \sim C$. They would also hold under a strict chain $A \succ B \succ C$, which implies $A \succ C$.

Conclusion and common mistake avoided. The required ranking is $A \succeq C$. Do not replace “at least as good as” with “strictly better than” unless the information supports that stronger claim. The bundle quantities are not enough to make a different ranking without an additional preference assumption.

Strict Preference and Transitivity

The weak-preference axiom also implies that strict preferences are transitive. If A is strictly preferred to B, and B is strictly preferred to C, then A must be strictly preferred to C. This result is useful for detecting a strict preference cycle, where each of three bundles is said to be strictly better than the next.

Result: If weak preferences are transitive and $A \succ B$ and $B \succ C$, then $A \succ C$. A cycle of strict preferences $A \succ B$, $B \succ C$, and $C \succ A$ cannot satisfy transitivity.

The strict-preference result follows from the definitions. First, $A \succ B$ gives $A \succeq B$, and $B \succ C$ gives $B \succeq C$. Weak transitivity therefore gives $A \succeq C$. To prove the comparison is strict, suppose instead that $C \succeq A$. Since $A \succeq B$, transitivity would then imply $C \succeq B$. But $B \succ C$ means that $C \succeq B$ is false. This contradiction shows that $C$ is not weakly preferred to A. Thus $A \succeq C$ and not $C \succeq A$, which is exactly $A \succ C$.

A \succ B,\ B \succ C \ \Longrightarrow\ A \succeq C \ \text{and not } C \succeq A \ \Longrightarrow\ A \succ C.

Worked Example: Finding a Strict Preference Cycle

An evaluator compares three hypothetical travel bundles, each combining a travel pass and a meal voucher. The reported rankings are $R \succ S$, $S \succ T$, and $T \succ R$. Determine whether these rankings can satisfy transitivity.

Step 1: Follow the first two comparisons. Since $R \succ S$ and $S \succ T$, strict transitivity requires $R \succ T$.

Step 2: Compare the implication with the third report. The report $T \succ R$ means $T \succeq R$ and not $R \succeq T$. But $R \succ T$ requires $R \succeq T$. These conditions conflict.

Step 3: State the diagnosis. The three strict rankings form a cycle. They cannot all be true of a transitive preference relation.

Conclusion and common mistake avoided. The rankings violate transitivity. Do not treat this as merely a missing comparison that can be filled in arbitrarily: the first two comparisons already require $R \succ T$, while the third states the reverse strict ranking.

Weak Cycles Are Not Automatically Violations

A cycle involving weak preferences requires more care. Suppose $A \succeq B$, $B \succeq C$, and $C \succeq A$. The first two comparisons imply $A \succeq C$. Together with $C \succeq A$, this means $A \sim C$. The other pairwise comparisons also become ties: $A \succeq B$ and $B \succeq A$, and $B \succeq C$ and $C \succeq B$. Thus all three bundles are indifferent.

That weak cycle is consistent with transitivity. The conclusion differs from a strict cycle because weak preference allows a tie. A violation requires a contradiction with the axiom, not just a circular-looking list of weak comparisons.

Worked Example: Interpreting a Cycle of Weak Rankings

A customer compares three hypothetical home-delivery plans. The reported comparisons are $M \succeq N$, $N \succeq L$, and $L \succeq M$. Decide whether the reports necessarily violate transitivity, and identify what they imply if the preference relation is transitive.

Step 1: Use the first two comparisons. From $M \succeq N$ and $N \succeq L$, transitivity implies $M \succeq L$.

Step 2: Combine the implication with the third comparison. The reported $L \succeq M$, together with the implied $M \succeq L$, gives $M \sim L$.

Step 3: Establish the other ties. Since $M \succeq L$ and $L \succeq M$, and $M \succeq N \succeq L$, transitivity gives $L \succeq N$. Combining $M \succeq N$ with $N \succeq M$ (which follows from $N \succeq L$ and $L \succeq M$) gives $M \sim N$. The same reasoning gives $N \sim L$.

Conclusion and common mistake avoided. The reports do not necessarily violate transitivity; they imply that all three plans are indifferent. A common error is to label every cycle a violation without checking whether the comparisons are weak or strict.

A Reliable Test for Three-Bundle Rankings

When a question gives several comparisons among three bundles, write down the chain before making a judgment. This keeps the direction of each ranking clear and helps separate a forced conclusion from a conclusion that merely seems plausible. The method applies to any bundle labels; it does not depend on the names of the goods.

Record each comparison in its stated direction.
For example, $A \succeq B$ points from A to B. Do not reverse it unless the reverse comparison is also given or can be proved.

Find a connected chain.
Look for a middle bundle that links two comparisons, such as $A \succeq B$ and $B \succeq C$.

Apply transitivity at the correct strength.
The two weak comparisons imply $A \succeq C$. If both are strict, strict transitivity implies $A \succ C$.

Check every additional report against the implication.
A report such as $C \succ A$ conflicts with $A \succeq C$. A report $C \succeq A$ may instead imply indifference between the endpoints.

Do not infer comparisons that have no logical support.
If the premises do not connect two bundles through a chain, transitivity alone may say nothing about their ranking.

Common Mistakes and Why the Axiom Matters

Transitivity matters because a ranking that reverses around a strict cycle cannot provide a stable ordering of the options being compared. In applications, the axiom helps make consumer rankings internally consistent. It does not prescribe which bundle a consumer should prefer; it restricts how a set of reported preferences can fit together.

Making a weak conclusion strict. From $A \succeq B$ and $B \succeq C$, infer $A \succeq C$, not automatically $A \succ C$.
Reversing a comparison without justification. $A \succeq B$ alone does not imply $B \succeq A$. Both directions are needed for indifference.
Calling every cycle inconsistent. A cycle of weak comparisons can represent indifference among the bundles. A strict cycle is the decisive contradiction.
Using bundle quantities as an unstated preference rule. Different quantities do not establish a ranking unless a relevant assumption is supplied.
Confusing transitivity with completeness. Transitivity extends a known chain; it does not guarantee that every pair can be compared.

Key takeaway: Transitivity requires $A \succeq C$ whenever $A \succeq B$ and $B \succeq C$. It also rules out a strict cycle $A \succ B \succ C \succ A$. A cycle of weak comparisons, by contrast, may simply mean that the bundles are all indifferent.

Check Your Understanding

For each problem, distinguish what transitivity requires from what the stated comparisons do not establish.

For three hypothetical meal bundles, $X \succeq Y$ and $Y \succeq Z$. State the comparison required between X and Z, and say whether strict preference is established. (Answer: $X \succeq Z$; strict preference is not established.)
A traveler reports $A \succ B$, $B \succ C$, and $C \succ A$ for three fare-and-lodging bundles. Use the first two reports to infer a ranking, then determine whether all three reports can satisfy transitivity. (Answer: The first two imply $A \succ C$; the report $C \succ A$ contradicts this, so the cycle violates transitivity.)
For three subscription bundles, $R \succeq S$, $S \succeq T$, and $T \succeq R$. If preferences are transitive, what do these reports imply about the bundles? (Answer: $R \sim S$, $S \sim T$, and $T \sim R$; all three bundles are indifferent, so the weak cycle is consistent with transitivity.)
A customer reports $K \succeq L$, $L \succeq M$, and $M \succ K$. Derive the comparison between K and M from the first two reports, and explain the conflict. (Answer: The first two imply $K \succeq M$; $M \succ K$ implies that $K \succeq M$ is false, so the reports are inconsistent with transitivity.)
For three hypothetical mobile plans, the only reported comparison is $P \succeq Q$. Can transitivity alone determine whether P is preferred to R, R is preferred to P, or they are indifferent? Explain. (Answer: No. There is no comparison chain connecting P and R, so transitivity alone determines no ranking between them.)

What You'll Learn