Understanding the Limitations of Utility Analysis in Consumer Demand Theory

Before economists started drawing curves on graphs, they tried something more straightforward: measuring satisfaction directly, in numbers. This is the utility analysis approach, and it sits at the heart of classical demand theory. It sounds neat in principle, but the moment you try to apply it to real consumer behaviour, cracks start to show. Understanding exactly where and why utility analysis falls short is what eventually pushed economists toward indifference curves, and it is also a favourite exam topic because it tests whether you actually understand the assumptions behind the model, not just the formulas.

A quick recap of the utility approach 🔗

Utility analysis, developed largely by Alfred Marshall, treats satisfaction as a measurable quantity, expressed in imaginary units called utils. A consumer buying successive units of a good is assumed to experience diminishing satisfaction from each additional unit, which is the law of diminishing marginal utility. Consumer equilibrium is reached when the utility gained per rupee spent is equalised across all goods purchased. It is a clean, mathematically convenient model. The trouble is that convenience and realism are not the same thing, and this approach makes several assumptions that do not survive contact with how people actually behave.

Where the utility approach starts to crack 🔗

1. You cannot really put a number on satisfaction 🔗

The biggest objection to utility analysis is also the most obvious one: satisfaction is a mental, subjective experience, not a physical quantity like weight or length. There is no scale you can use to say a plate of biryani gives you exactly 20 utils while a movie gives you 15. Two people eating the same meal will feel entirely different levels of satisfaction, and even the same person’s satisfaction from an identical good changes with mood, health, or context. This is precisely why later economists moved toward ranking preferences instead of quantifying them. Under the ordinal approach that indifference curves rely on, a consumer only needs to say which bundle of goods they prefer, not by how much, since utility is treated as a psychological state that resists precise measurement.

2. The constant marginal utility of money assumption doesn’t hold up 🔗

Marshall’s model works only if the marginal utility of money stays constant while a consumer spends it, which lets money act as a stable yardstick for measuring the utility of other goods. But this directly contradicts the very law of diminishing marginal utility that the theory is built on. If satisfaction from every other commodity diminishes with more consumption, there is no logical reason for money, itself a claim on all those commodities, to behave differently. Marshall’s own framework depended on this constant marginal utility of money to translate satisfaction into a monetary demand price, and economists have long pointed out that this assumption only simplifies the analysis rather than reflecting reality. In fact, later work in demand theory shows that Marshall’s own caution about this assumption became unnecessary only once economists developed a more rigorous definition of consumer surplus using duality theory, decades after his original formulation.

3. It ignores that no two people feel satisfaction the same way 🔗

Utility analysis also assumes that satisfaction can be compared across individuals, which creates serious problems the moment you try to use it for policy. If one person claims to get more utils from an additional thousand rupees than another person does, is that a fact or just a value judgement? Economist Lionel Robbins famously argued that every mind is fundamentally inscrutable to another, making interpersonal utility comparisons scientifically indefensible, a critique that pushed mainstream economics firmly toward the ordinalist, preference-ranking approach from the 1930s onward. This matters beyond the classroom too: it directly complicates arguments like taxing the rich because a rupee supposedly matters less to them, since there is no objective way to verify that claim.

4. It cannot explain Giffen goods 🔗

Ordinary demand theory says that as the price of a good falls, people buy more of it. Giffen goods break this rule. Named after the economist who first observed the pattern, a Giffen good is one where the negative income effect from a price change is strong enough to override the usual substitution effect, causing demand to move in the same direction as price rather than the opposite direction. The textbook example is a staple food like coarse grain during a famine: when its price rises, poor households, unable to afford meat or better alternatives, end up buying even more of it because it is still the cheapest way to fill their stomachs. Utility analysis, which never separates the income and substitution components of a price change, has no tools to explain this. It simply predicts falling demand as price rises, full stop, and has no room for the paradox.

5. It cannot separate the income effect from the price effect 🔗

This limitation is really the root cause of the Giffen goods problem above, but it deserves attention on its own because it affects almost every prediction utility analysis makes about price changes. Whenever a price falls, two things happen simultaneously: the good becomes relatively cheaper compared to substitutes (the substitution effect), and the consumer’s real purchasing power rises (the income effect). Whether demand ultimately rises or falls, and by how much, depends on how these two effects interact for a given good, and this interaction is what determines whether a good behaves normally, like an ordinary inferior good, or like a true Giffen good, where the income effect is large enough to completely outweigh the substitution effect. Because utility analysis bundles the total price effect into a single number, it cannot tell you which force is doing the work, which makes it a poor tool for predicting how consumers respond to subsidies, taxes, or price controls.

6. It treats every good as if it exists in isolation 🔗

A subtler but equally important assumption is that the utility derived from one good is independent of how much of other goods a person consumes. In reality, goods are rarely independent. Tea and coffee compete for the same craving, so buying more of one reduces the marginal utility of the other, since they are substitutes. Bread and butter, on the other hand, are typically consumed together, so having more bread can actually raise the satisfaction gained from butter, since they are complements. Classical utility theory assumes marginal utility changes only with a good’s own consumption, when in practice the marginal utility of a good shifts depending on how much of its substitutes or complements a consumer already holds. By ignoring these cross-effects, utility analysis oversimplifies a shopping basket that, in reality, is full of interconnected choices.

A quick side-by-side view 🔗

Why these gaps matter 🔗

None of this means utility analysis was a wasted effort. It gave economics its first systematic language for consumer behaviour and introduced ideas, like diminishing marginal utility and consumer equilibrium, that remain useful even today. But its limitations are exactly why John Hicks and R.G.D. Allen developed the ordinal, indifference curve approach, which drops the need to measure utility in absolute terms and instead relies only on ranking preferences. That shift let economists explain Giffen goods properly, cleanly separate income and substitution effects, and build demand theory on far fewer, more defensible assumptions. If you are studying indifference curve analysis right after this topic, keep these six limitations in mind. Every strength of the ordinal approach you will study next exists specifically because it was designed to fix one of these gaps.

What do you think? If satisfaction cannot truly be measured in numbers, does ranking preferences through indifference curves fully solve the problem, or does it just sidestep it? And can you think of a good in your own daily life, besides food staples, that might behave like a Giffen good under the right conditions?

References

1. https://www.sciencedirect.com/topics/social-sciences/marginal-utility
2. https://link.springer.com/rwe/10.1057/978-1-349-95121-5_954-2
3. https://utilitarianism.net/guest-essays/welfare-economics/
4. https://en.wikipedia.org/wiki/Giffen_good
5. https://socialsci.libretexts.org/Bookshelves/Economics/Microeconomics/Intermediate_Microeconomics_with_Excel_(Barreto)/04%3A_Compartive_Statics/4.06%3A_Income_and_Substitution_Effects
6. https://libjournals.mtsu.edu/index.php/jfee/article/download/1496/1075/4113
7. http://www.eagri.org/eagri50/AECO141/lec03.pdf