The Shift

The wedge a tax drives between the buyer's price and the seller's has a size — the tax — but the wedge must be divided: the buyer pays somewhat more, the seller receives somewhat less, and the two shares add to the tax. How the wedge divides — who bears more of it — is the question of tax incidence, and its answer is the volume's central surprise. The division is not decided by the law that levies the tax, nor by which party the law names as payer; it is decided by the market, by a single property of the two sides: how readily each can escape the taxed trade.

Begin with the fact that astonished the first analysts: it makes no difference, to who bears the tax, on which side the law lays it. Suppose a tax on bread. The law may levy it on the baker, who must remit it, or on the customer, who must remit it — and one's intuition is that a tax on the baker falls on bakers, a tax on the customer on customers. But this intuition is wrong. Whichever side the law names, the price adjusts to redistribute the burden in exactly the same way. Tax the baker, and bakers, to recoup the tax, raise the price of bread, passing some of the burden to customers; tax the customer, and customers, facing the tax, will pay bakers less for the untaxed bread, passing some of the burden to bakers. When the prices have settled, the division of the burden between baker and customer is identical, whichever side the law taxed. The legal incidence — whom the law names — does not determine the economic incidence — who bears the burden. The law can choose the collection point; it cannot choose the bearer.¹

If the law does not decide the division, what does? The answer is elasticity — the responsiveness of each side to price, which measures how readily each can escape the taxed trade. The burden falls more heavily on the side that is less elastic, less able to change its behavior, more stuck. Consider why. When the tax pushes the price up for buyers and down for sellers, each side responds by reducing its participation — buyers buy less, sellers sell less — and the side that responds more, the more elastic side, sheds more of the burden, because it can more easily retreat from the taxed trade, while the side that responds less, the more inelastic side, is stuck in the trade and absorbs more of the tax. A buyer who can easily do without the good, or substitute another, escapes the tax by buying less; a buyer who must have the good, whatever the price, is stuck, and bears it. So the burden settles on the inelastic: on whoever cannot readily escape — the necessity no one can forgo, the labor a worker must sell, the land that cannot move. Elasticity is a property of the market, of tastes and technologies and alternatives, not of the law; and so incidence, governed by elasticity, is decided by the market, not the legislature.²

Fleeming Jenkin, a Scottish engineer who turned his diagrammatic mind to economics, was the first to render the division visible, drawing the supply and demand curves and showing how a tax splits between them according to their slopes. His diagrams made the extremes vivid. If demand is perfectly inelastic — buyers will take the same quantity at any price, because the good is an absolute necessity — then buyers bear the entire tax: the seller raises the price by the full tax, and buyers, unable to retreat, pay it all. If supply is perfectly inelastic — sellers will offer the same quantity at any price, because the thing cannot be reduced, as with land — then sellers bear the entire tax: the price to buyers does not rise, and the seller absorbs it all. Between these extremes, the burden divides in proportion to the relative inelasticities, the more inelastic side bearing the larger share. Jenkin's geometry showed that the division is a matter of the curves — of how steeply each side's quantity responds to price — and that the legislator, who can draw no curves, has no hand in it. The picture made plain what the algebra confirmed: incidence is read off the elasticities, and the elasticities are facts about the market.³

The principle, applied, dissolves the surface of real taxes. The payroll tax, levied half on the employer and half on the worker by law, is in economic fact borne almost wholly by the worker: labor supply is quite inelastic, since people must work, and work about as much whatever the wage, so the employer passes nearly the whole tax back in lower wages, and the legal split is a fiction. A tax on a staple food, levied on grocers, is borne mostly by consumers, for whom the staple is a necessity, its demand inelastic; a tax on a luxury, easily forgone, is borne more by its sellers, who must cut their prices to keep buyers, its demand elastic. A tax on apartments may fall on tenants where housing is scarce and they are stuck, or on landlords where tenants can readily move. In each case the rule is the same: find the inelastic side, and there the burden settles, whatever the law's nomination. And the policy lesson follows: to know who really bears a tax, ignore the law's naming and ask only who is stuck — who cannot escape the taxed trade.⁴

The shift, then, is the heart of incidence: the burden of a tax, levied at one point, slides through the price to settle at another — on the inelastic side, whoever cannot escape the taxed trade — and the law that names the payer has no power over this settling, which the market governs through elasticity alone. This already overturns the surface of tax debate: the payroll tax does not fall on employers, the corporate tax not on corporations, the tax on the rich not necessarily on the rich. But the full depth of the matter — and its meaning for the state's power — waits for the flagship. For so far we have watched the burden divide between the two parties to a single taxed trade, the buyer and the seller. The flagship will show that the burden does not stop there but ripples through the whole economy, settling on people far from the taxed transaction, whom the legislator never imagined; and that this exposes a deep limit on the state's power — that the state can compel the payment of a tax but cannot place its burden, can tax but cannot burden. Before that, two plates take up the design of taxes given their incidence: how to tax well, and the rare tax that does good. The next plate is the optimum.⁵

The shift is the mechanism of incidence: a tax divides between buyer and seller not as the law dictates but as the market does, the burden settling on the side less able to escape the taxed trade — the inelastic side, the stuck side. It makes no difference whom the law names as payer; the price adjusts to the same division either way, a division read off the elasticities, which are facts about the market and not the law. Jenkin first drew the picture; the extremes show buyers bearing all when demand cannot retreat, sellers bearing all when supply cannot, and every case between divided by the relative inelasticities. So the payroll tax falls on the worker, the staple tax on the consumer, and the law's naming is, for the question of who bears, an illusion. But the burden's journey does not end with the two parties to the trade; it ripples outward through the whole economy, as the flagship will show, settling where the legislator never looked. First, though, the design of taxes: given that a tax distorts and that its burden shifts, how should a tax be laid? The next plate is the optimum.