Congo's Progressive Property Tax Experiment Raised Revenue by 56%

Why progressive property taxes are rare in poor countries

High-income countries lean on progressive taxation — income tax brackets, capital gains rates, estate taxes — as a matter of course. Developing-country governments mostly don't, and not because they've decided flat taxes are fairer. Progressivity requires knowing who owns what and what it's worth, then being able to enforce collection against owners who have both the means and the incentive to resist. Where property registries are outdated or missing and tax administrations are thin on staff and data, most public finance economists have assumed progressive schedules would collapse under their own administrative weight — coverage would be patchy, enforcement inconsistent, and richer owners would simply evade at higher rates than poorer ones, undoing the fairness gain and losing revenue in the process.

What the Congo experiment actually did

Six economists — Gabriel Z. Tourek, Arthur Laroche, Augustin Bergeron, Joana Naritomi, Jonathan L. Weigel, and Marina Mavungu Ngoma — partnered with a provincial government in a large Congolese city to test this directly, rather than model it theoretically. Neighborhoods across the city were randomly assigned to one of two property tax schedules: a flat, proportional rate applied uniformly regardless of property value, or a progressive schedule where the tax rate itself rose with assessed property value. Because assignment was random at the neighborhood level, differences in revenue and compliance between the two groups can be attributed to the schedule itself, not to pre-existing differences between richer and poorer parts of the city.

Where the extra revenue actually came from

The headline number is a 56% revenue increase under the progressive schedule relative to the proportional one. What makes the result more interesting than a single top-line figure is where that 56% came from: gains showed up across the entire property value distribution, not just at the top. Among the most valuable properties, the mechanical effect of a higher statutory rate raised more revenue per property even though compliance among wealthy owners dipped somewhat — some simply paid less of what they technically owed. Among lower-value properties, the opposite dynamic applied: a lower rate made the tax bill feel more worth paying, and the resulting jump in how many people actually paid more than offset the smaller amount owed per compliant owner.

What this means beyond one city in Congo

The immediate implication is narrow and empirical: progressive property taxation is not automatically self-defeating in a low-capacity administrative environment, at least not in this setting. The broader implication is about how governments think about the trade-off between fairness and feasibility. Finance ministries and municipal authorities across the developing world routinely default to flat-rate instruments — VAT, flat property levies, uniform license fees — partly out of a belief that progressive alternatives are administratively unworkable outside rich-country tax agencies. This study is direct evidence, from an actual government partnership rather than a lab or survey, that progressive design can outperform flat design on revenue grounds alone, before any fairness argument is even invoked. It also matters for how enforcement resources get allocated: the researchers found that when the city later scaled the progressive schedule citywide and paired it with enforcement specifically targeted at high-value properties, the gap between statutory and effective tax rates at the top closed — the system stopped being regressive in effective terms, not just in the rate schedule written into law.

How this fits with what we already knew

Prior research on taxation in low-income settings has mostly documented the opposite problem: attempts at progressive taxation — on income, on high-value assets — running aground on weak third-party information (no reliable employer wage reporting, no bank data-sharing, no digitized land registries) that rich-country tax agencies take for granted. Property tax is unusual in this literature because the tax base is physically observable — a building doesn't move, and its relative size and location are visible without any digital paper trail. That may be exactly why progressivity worked here when it has struggled elsewhere: assessors could rank properties by observable characteristics well enough to apply different rates credibly, even without the kind of granular financial data an income tax would require.