The point of tax incidence is to separate who is legally responsible for paying a tax from who actually bears its burden. Those two answers are almost always different. This walkthrough computes the buyer and seller share of a per-unit tax, shows the elasticity rule that predicts the answer in advance, and finishes with the deadweight loss the tax creates.
Statutory vs. Economic Incidence
Statutory incidence is who the law tells to send the check to the government. Economic incidence is who actually has less money once prices adjust. They are not the same, and the central result of this topic is that they don't have to match.
Consider a $1 per-unit tax on, say, soda. Whether the law says the buyer pays it at the register or the seller remits it from each sale, the market moves to the same final outcome: the buyer pays a higher price, the seller receives a lower one, and the wedge between them is the $1. The split of how much of that $1 each side absorbs is the economic incidence.
How a Per-Unit Tax Shifts the Market
A specific tax of $t per unit drives a $t wedge between the price the buyer pays (P_b) and the price the seller keeps (P_s):
P_b − P_s = t
Graphically, a tax levied on sellers shifts the supply curve up by exactly t (every quantity now requires t more to motivate the same supply). A tax on buyers shifts the demand curve down by t (buyers are willing to pay t less to producers, since they owe t to the government). Either shift lands at the same new equilibrium quantity Q_t and the same buyer/seller prices — that is why statutory incidence does not matter.
The buyer's share of the burden is the rise from the original equilibrium price P up to P_b. The seller's share is the drop from P down to P_s. The two must sum to t.
A Worked Example With Numbers
Take the same kind of linear market from a price-control problem.
- Demand: Q_d = 100 − 2P
- Supply: Q_s = −20 + 2P (so the supply curve hits zero quantity at P = $10)
Equilibrium (no tax): set 100 − 2P = −20 + 2P → 120 = 4P → P = $30, Q = 40.
Now levy a per-unit tax of t = $8 on sellers. Replace P in the supply equation with the seller's received price P_s = P_b − 8, while demand is in terms of the buyer's price P_b:
- Q_d = 100 − 2P_b
- Q_s = −20 + 2(P_b − 8) = −36 + 2P_b
Set equal: 100 − 2P_b = −36 + 2P_b → 136 = 4P_b → P_b = $34. Then P_s = $34 − $8 = $26, and Q_t = 100 − 2(34) = 32 units.
Buyer share: $34 − $30 = $4 (half of t). Seller share: $30 − $26 = $4 (half of t). Both sides bear exactly half because, in this market, supply and demand have equal slopes (and so, at the equilibrium point, equal elasticities). That symmetry is the textbook special case, not the rule.
The Elasticity Rule: Who Bears More
The split between buyer and seller depends on relative elasticity, not on who writes the check. The rule:
The more inelastic side bears more of the tax.
Why: the less responsive your quantity is to a price change, the harder it is to escape the tax by buying or selling less. Insulin (highly inelastic demand) is taxed and the buyer absorbs almost the whole tax — buyers can't cut back, so sellers don't have to absorb much. Skilled labor with elastic supply (workers can leave the industry) is taxed and most of the burden falls on the buyer side (firms/consumers), because supply is what flees the tax.
A useful approximate formula for the buyer's share of a small tax:
Buyer share ≈ |E_S| / (|E_S| + |E_D|)
If supply elasticity equals demand elasticity (as in the worked example), each share is 1/2. If demand is twice as inelastic as supply, buyers bear two-thirds of the tax.
Tax Revenue and Deadweight Loss
Two consequences of any binding per-unit tax:
- Government revenue: t × Q_t. In the example, $8 × 32 = $256.
- Deadweight loss: the lost surplus from the units that no longer trade. For a linear market this is the area of a triangle with base = drop in quantity (Q* − Q_t) and height = t. In the example, base = 40 − 32 = 8, height = $8, so DWL = ½ × 8 × 8 = $32.
The deadweight loss grows with elasticity: when buyers and sellers can easily walk away from the taxed good, the tax destroys more trades and raises less revenue per dollar of harm. This is one reason governments tend to tax inelastic goods (gasoline, tobacco) — narrow excise taxes there raise revenue with a smaller deadweight loss. For more on what that DWL triangle measures, see the deadweight loss explainer.
Conclusion
Tax incidence answers a single question — who really pays? — and the answer follows the elasticities, not the statute. A per-unit tax of t opens a wedge between the buyer's price and the seller's price; the share each side absorbs is determined by relative elasticity, and the side that can least change its behavior pays the most. Add tax revenue (t times the new quantity) and deadweight loss (the triangle between the curves over the lost units) and you have everything a typical tax-incidence problem asks for.
