Tax competition

Why do countries keep cutting the same tax when none of them ends up ahead?

Over four decades, statutory corporate income tax rates fell almost everywhere. The worldwide average was around 40 percent in 1980 and sits near 23 percent today. No single treaty ordered this, no ideology swept every government at once, and the direction never reversed for long.

What is odd is not that rates fell. It is that they fell in a group whose members were, by their own account, competing with each other for the same thing. If everyone runs faster in the same race, nobody moves up the field. So why does each government keep cutting a tax that, once the others match it, leaves its relative position exactly where it was — and its treasury smaller?

Start with one finance minister and hold every other country fixed. She is choosing a rate. What does she actually gain by cutting?

Not more tax from the firms already there — she gets less from them. She gains only whatever arrives because of the cut: profit booked in her jurisdiction rather than a neighbour's, a factory sited on her side of a border. So the cut pays off exactly to the extent that the tax base is mobile. Ask yourself which base is most mobile, and you get the shape of the whole problem: a patent's royalty income can be relocated with a signature, a smelter cannot.

Now notice what this makes the decision. Given what the others have chosen, cutting is her best move — she captures base. And if the others cut first, cutting is still her best move, because now she is defending base she already has. There is no configuration of other countries' rates at which she would rather sit above them. That is the signature of a collective-action problem: the individually optimal choice does not depend on what anyone else does, and it is the same choice that makes everyone worse off when all of them make it.

Follow it to the end state. Every country cuts, every country's relative attractiveness is restored to roughly where it started, and the mobile base sits where it would have sat anyway. What has changed is the level of revenue collected from it. The competition transferred money from treasuries to capital without moving much capital.

So why does nobody stop? Here the reasoning has to be careful. It is not that ministers cannot see this — they can, and they have said so for thirty years. It is that seeing it changes nothing about the individual choice. Unilateral restraint costs you base and buys nobody's cooperation. Even an agreement among ten countries is undone by the eleventh, because the base is mobile enough to find whoever stayed out. A promise not to defect is worth nothing when defection is unobservable and profitable.

What would actually work, then? Not a pledge — a change in the payoff. That is what the 2021 OECD and G20 agreement on a global minimum tax attempts. Its core device is a top-up: if profits in a jurisdiction are taxed below 15 percent, another jurisdiction with a claim on that group collects the difference. Think about what that does to the incentive. A country offering 5 percent no longer offers the firm a 5 percent rate; it offers 5 percent plus a 10 percent top-up somewhere else. The undercutting stops paying, so the reason to undercut dissolves.

Whether it holds is genuinely unsettled. Implementation is uneven, the rules are complex enough to invite new forms of arbitrage, and countries have already begun competing on subsidies and credits instead — which is the same race with a different baton.

Falling rates did not produce proportionally falling revenue. Statutory rates roughly halved, yet corporate tax revenue across the OECD has hovered near 3 percent of GDP for decades. Governments broadened bases while cutting rates, and more business income shifted into corporate form. So the treasury damage is real but smaller than the headline suggests — and anyone arguing from the rate alone is overstating the case.

"Nobody ends up ahead" is not quite true. Very small jurisdictions genuinely do win. A country whose domestic base is tiny loses almost nothing by cutting and can gain booked profits worth many times its own economy. Their position is not irrational; it is a different position on the same board. The symmetric losing outcome applies to large economies competing with each other.

The competition is mostly over paper, not plant. Evidence consistently finds reported profit far more responsive to tax rates than real investment is. That matters for the diagnosis: if what moves is where income is declared, then the cuts are buying accounting rather than factories, and the case for them is much weaker than the case usually made.

The falsification test. If this is a collective-action trap rather than a shared change of economic belief, then removing the gain from undercutting — as a binding minimum with a top-up does — should stop the downward drift without any government changing its view about taxes. If rates keep falling in participating countries despite the top-up, the mechanism is wrong and something else was driving the trend.

• Whether subsidy and credit competition simply replaces rate competition once a minimum binds.
• The same structure in other arenas: financial regulation, environmental standards, minimum wages across neighbouring states.
• Why the winners' curse for small jurisdictions may be temporary — what happens to a haven when the top-up arrives.

Every term the collection defines is gathered in the glossary.