The Tax Nobody Argues About

Value added tax was invented by a French official in the 1950s and has since spread to more than 160 countries. It raises roughly a fifth of all tax revenue across the OECD. It attracts a tiny fraction of the political attention devoted to income tax, for the straightforward reason that it arrives buried inside prices rather than as a line on a payslip.

Its design deserves more admiration than it gets. Each firm in a chain charges tax on its sales and reclaims the tax it paid on its purchases, remitting only the difference: the tax on the value it added. Two consequences follow. The total collected across the whole chain equals the rate applied once to the final price, no matter how many businesses the good passed through, which avoids the cascading that made older turnover taxes reward vertical integration for no productive reason. And every firm has a financial motive to obtain documentation from its supplier, because that documentation is what it reclaims against. The enforcement is built into the incentives rather than bolted on.

Compare a retail sales tax, which has no such chain. It relies on the final seller to remit, and it collapses at exactly the points where both parties to a transaction would prefer no record: services, cash trades, cross-border sales, and any purchase where the buyer can claim to be a business. The United States is the only large rich economy that still runs on this, and proposals for a federal alternative surface every decade and fail on the same objection: that it would be an extremely effective machine for raising revenue.

Incidence is where the argument usually starts. Measured against income in a single year, a consumption tax is clearly regressive: poorer households spend a larger share of what they earn, so they pay a larger share of income in tax. Measured against lifetime spending it flattens considerably, since savings are eventually spent or bequeathed and taxed at that point. Both measurements are correct and they answer different questions. If you care about hardship this year, the annual figure is the relevant one. If you care about the burden over a life, it is not.

Which brings us to the reflex most governments reach for, and the evidence against it. Zero-rating food, children's clothes and domestic fuel is intended to protect poor households, and it is a poor instrument for doing so. Rich households buy more food in absolute terms than poor ones, so most of the revenue forgone goes to people who did not need help. The relief costs several times what a targeted transfer of equivalent benefit would cost, and the calculation has been done repeatedly with the same result.

It persists anyway, and the reason is not stupidity. A zero rate is visible, permanent, and requires no one to apply for anything or prove their circumstances to an official. A transfer has to be designed, administered, claimed and defended at every budget. The economics favours the transfer and the politics favours the exemption, and the politics has won nearly everywhere.