Case Study 2 — Three Goods, Three Elasticities: Insulin, Gasoline, and Restaurant Meals

This case study walks through three real-world goods with very different elasticities of demand. By comparing them side by side, you can see how the four determinants from §6.2 (substitutes, necessity, time horizon, budget share) translate into the actual numbers economists estimate from real data — and how those numbers shape policy and personal decisions.

The three goods: insulin, gasoline, and restaurant meals. Each has a different elasticity profile. Each has different policy implications. Each shows up in modern political debates in different ways. Comparing them tells you something about how the same analytical tool can be applied across very different markets.

Insulin: highly inelastic

Insulin is the canonical example of an extremely inelastic good. It is a hormone that type-1 diabetics need to survive. Without insulin, a type-1 diabetic dies — typically within days to weeks. There is no substitute. There is no alternative therapy. There is no way to consume "less" insulin and survive long-term. The price elasticity of demand for insulin among type-1 diabetics is essentially zero.

What does this imply?

1. Pricing power. Manufacturers of insulin can charge very high prices and the quantity demanded barely changes. This is exactly what has happened in the U.S. insulin market. Between 2002 and 2013, the price of a vial of insulin in the U.S. roughly tripled. The number of type-1 diabetics did not triple; in fact, it grew slowly. The manufacturers' revenue rose enormously while the underlying medical need was unchanged. This is the elasticity of zero in action: the manufacturers could raise prices because the patients had no alternative.

2. Tax incidence. A tax on insulin would fall almost entirely on patients, not on manufacturers. The reason: patients have nowhere else to go. Even if the tax were nominally collected from manufacturers, the manufacturers could pass it on as a higher price, and patients would still pay because they have to.

3. Welfare and policy implications. When demand is highly inelastic and the good is a necessity, the market produces outcomes that most people consider unjust. The Insulin Affordability Act, the Inflation Reduction Act's $35 cap on insulin for Medicare beneficiaries, the Eli Lilly $35 cap announced in 2023 — all of these represent attempts to override the market outcome through policy intervention. The economic justification is that the inelasticity makes the standard "competitive market" defense fail. There is no meaningful competitive pressure to keep insulin prices low when patients can't say no.

This is one of the cases where the standard market model breaks down most starkly, and where most economists support some form of policy intervention. The elasticity is the technical reason.

Gasoline: moderately inelastic in the short run, more elastic in the long run

Gasoline is a more complicated case. The empirical estimates of the price elasticity of demand for gasoline are roughly:

• Short run (within a few months of a price change): −0.1 to −0.3
• Long run (after several years): −0.5 to −1.0

In the short run, gasoline is fairly inelastic. When gas prices spike, people don't immediately change their commutes, sell their cars for hybrids, or move closer to work. They keep driving more or less the same amount, grumbling about the cost. The behavioral economist Daniel Kahneman would say they "feel" the price change but their behavior takes time to adjust.

In the long run, the elasticity is much larger. Over five or ten years, people can: - Replace their cars with more fuel-efficient ones - Move closer to work - Switch to public transit if it becomes available - Bike or walk more - Combine errands to reduce driving - Telecommute or change jobs

All of these are slow adjustments. Each takes months or years. Aggregate them across millions of people and you get a much larger long-run response than short-run response.

What does this imply for policy?

1. A gas tax doesn't reduce consumption immediately. When a gas tax is imposed, the short-run effect on consumption is small. Politicians who expect to "see" the impact of a gas tax in the first year are usually disappointed. The real effect takes years.

2. A gas tax does eventually work. Over a decade, the same tax produces meaningful reductions in consumption — perhaps 10–20% lower than what consumption would have been without the tax. This is consistent with how European countries (with higher gas taxes) have lower per-capita gasoline consumption than the United States.

3. The burden is borne mostly by consumers. Gasoline supply is fairly elastic at the global level (oil markets adjust over time), but in the short run with localized taxes, the supply curve faced by U.S. drivers is more elastic than demand. The tax burden falls mostly on consumers.

4. Gas taxes are regressive in the short run. Because gasoline is a larger share of low-income budgets than high-income budgets, the burden falls disproportionately on poorer households. This is one of the reasons gas tax proposals are politically difficult, even when their long-run benefits (reduced emissions, less driving, less infrastructure wear) are clear.

The elasticity story is essential to understanding why climate policy through gas taxes is harder politically than economic theory suggests it should be. The short-run inelasticity means the pain comes first; the long-run elasticity means the benefits arrive later. Voters experience the pain and don't always live long enough — politically — to see the benefits.

Restaurant meals: moderately elastic

Restaurant meals are a good with relatively elastic demand. Empirical estimates put the elasticity in the range of −1.0 to −2.0 for most types of restaurants — meaning that a 10% price increase causes a 10–20% reduction in quantity demanded.

Why is restaurant demand so elastic?

Substitutes are abundant. When restaurant prices rise, consumers can cook at home, eat at less expensive restaurants, or simply eat out less often. The substitutes are plentiful and immediately available.

Restaurant meals are luxuries (in the economic sense). They are not necessities — you can survive without ever eating at a restaurant. Demand for luxuries is more elastic than demand for necessities.

Time horizon barely matters. Unlike gasoline (where adjusting takes years), adjusting your restaurant habits takes a single decision — maybe a single conversation with your spouse. There's no infrastructure to change.

Budget share is moderate. Restaurant spending is a meaningful but not dominant part of most households' budgets, so the income effect of price changes is modest. The substitution effect (cooking at home, going to cheaper places) does most of the work.

What does this imply?

1. Restaurant prices are constrained by competition. Restaurants cannot raise prices very far without losing customers. The elastic demand keeps restaurant prices in line with what consumers are willing to pay.

2. Restaurants are vulnerable to economic downturns. When household incomes fall, restaurant spending falls disproportionately. This is one of the reasons restaurants suffered so much during the COVID recession — people stopped eating out, and the elasticity meant that even modest income declines could produce large reductions in spending.

3. Tax incidence falls mostly on producers. A tax on restaurant meals (like a state sales tax surcharge) falls mostly on restaurant owners, not on consumers, because consumers can easily substitute toward eating at home. The same tax on a more inelastic good would fall mostly on consumers.

4. Pricing flexibility. Restaurants can offer specials, discounts, and promotions to attract more customers — and these work because demand is elastic. A 10% discount can bring in 15–20% more customers, raising total revenue. Restaurants that don't understand their elasticity often miss this opportunity.

Three goods, three policy implications

Let's compare what these three elasticity profiles imply for tax policy:

The three cases also illustrate the four determinants from §6.2: - Insulin: no substitutes, life-saving necessity, no time horizon to adjust, large for affected patients - Gasoline: moderate substitutes (other transit, but takes time), necessity for most lifestyles, time horizon matters enormously, moderate-to-large budget share - Restaurant meals: abundant substitutes (cooking, cheaper restaurants), luxury, easy adjustment in days, moderate budget share

Each elasticity emerges from the interaction of all four determinants. None of the three goods is "elastic" or "inelastic" by some universal standard — each has its specific elasticity because of its specific market characteristics.

This is the practical lesson of Chapter 6: elasticity isn't a property a good has by nature. It's a property that emerges from the four determinants. Once you can identify the four determinants for a good, you can predict its approximate elasticity. And once you know its elasticity, you can predict how it will respond to policy interventions — including taxes, subsidies, price controls, and changes in supply or demand.

Discussion questions

1. Why is the U.S. insulin market a good example of where the standard "competitive market" defense fails? What policy would you propose, given that demand is essentially perfectly inelastic?

2. Gas taxes work in the long run but are politically difficult because of short-run pain. What's a way to design a gas tax that mitigates the short-run pain while still capturing the long-run benefits? (Hint: think about what to do with the revenue.)

3. Restaurant demand is elastic. How does this explain the survival strategies of restaurants — the menu pricing, the promotions, the loyalty programs, the seasonal specials?

4. Pick another good and try to estimate its elasticity using the four determinants. Justify your estimate. Then try to find an empirical estimate (from academic research or government reports) and see how close you got.

5. The case study claims that "elasticity isn't a property a good has by nature." Is this strictly true? Are there any goods whose elasticity is "fixed" regardless of context, or does context always matter?