Who Actually Pays a Tax? The Answer Is Not Who Hands Over the Money
By the BrainSnail editorial team. How these articles are written and checked, and how to tell us when one is wrong.
A tax collected from sellers can be paid almost entirely by buyers, and one collected from buyers can be paid by sellers. Who bears the cost depends on how each side responds to price, not on who writes the cheque.
The distinction that matters
Economics separates two things that are easily confused. Legal incidence is who is obliged to remit the tax to the authority, which is written into the statute and is visible on any receipt. Economic incidence is whose real income falls as a result, which is a consequence of how the market adjusts and is not written anywhere. The two need not coincide and frequently do not, because the party remitting the tax will change the price they charge or offer in response, shifting part or all of the burden to the other side. The result is that a government choosing which side to collect from is making an administrative decision about convenience and enforcement rather than a decision about who ends up paying, which is one of the most robust and least intuitive findings in the field.
What determines the split
The division follows from how responsive each side is to price:
- •The less responsive side bears more of the burden, because they have fewer alternatives
- •Buyers who will purchase regardless of price bear nearly all of a tax on that good
- •Sellers who cannot reduce supply bear nearly all of a tax on what they sell
- •Where both sides respond strongly, the tax collects little because the quantity traded falls sharply
- •Over longer periods both sides become more responsive, so incidence shifts with the time horizon
- •In a market where sellers have no alternative use for what they supply, the burden lands on them almost entirely
What this means in practice
The framework explains several policy outcomes that surprise people. Taxes on goods with few substitutes and strong habitual demand fall largely on consumers, which is one reason such goods are taxed, since the revenue is stable, and it is also why the tax reduces consumption less than its advocates hope. Payroll taxes divided nominally between employer and worker are borne largely by workers in the form of lower wages, because labour supply responds weakly to pay in aggregate, which most analyses conclude regardless of how the statute splits it. Property taxes on land fall on the landowner because land cannot be produced or withdrawn, which is the argument behind proposals to tax land specifically. Taxes on internationally mobile activity are difficult to make stick because the responsive side simply leaves.
Subsidies work the same way
The analysis applies unchanged to money flowing the other direction, which is less obvious and equally important. A subsidy paid to producers raises the price they receive and lowers the price buyers pay, and the split between those two follows exactly the same logic as a tax, so a subsidy intended to help buyers may largely raise producer revenue instead. Housing support paid to tenants can raise rents where supply is slow to respond, transferring much of the benefit to landlords, which is a well-studied and contested finding. Support for higher education interacts with what institutions charge. Agricultural payments can raise land values rather than farm incomes, since land is the fixed factor and absorbs the benefit. In each case the question is which side can adjust, and the answer determines whether the policy reaches the people it names.
Where the simple model breaks
The standard analysis assumes competitive markets and the real cases deviate in documented ways. A firm with market power may pass on more than the full tax, raising prices by more than the amount collected, which has been observed and which the competitive model does not predict. Prices frequently rise promptly when a tax increases and fall slowly or not at all when it decreases, an asymmetry documented in fuel and other markets. Salience matters, since a tax included in the displayed price affects behaviour more than one added at the till, which experiments have demonstrated and which the standard model ignores. Administrative costs fall on whoever remits regardless of economic incidence. And in markets with few sellers the tax can change strategic behaviour in ways that depend on the details rather than on responsiveness alone.
The takeaway
Who remits a tax and whose income falls are different questions, and the statute answers only the first. The less responsive side to price bears more of the burden, which is why payroll taxes fall largely on wages and land taxes on landowners. Firms with market power can pass on more than the full amount, and prices rise faster on an increase than they fall on a cut.