Does Progressivity Raise Tax Capacity? Experimental Evidence from the D.R. Congo (Tourek, Laroche, Bergeron, Naritomi, Weigel & Mavungu Ngoma, 2026)

Summary

"Does Progressivity Raise Tax Capacity? Experimental Evidence from the D.R. Congo," by Gabriel Tourek, Arthur Laroche, Augustin Bergeron, Joana Naritomi, Jonathan Weigel, and Marina Mavungu Ngoma, circulated as NBER Working Paper 35536 in July 2026 (DOI 10.3386/w35536). It comes from the same research team behind the wiki's existing DRC state-capacity-ceiling study, and extends that agenda from the level of property-tax rates to their structure: in partnership with a provincial government, neighborhoods of a major Congolese city were randomly assigned to either a progressive or a proportional property-tax schedule.

The Experiment and Findings

• Progressivity raised revenue substantially. The progressive schedule increased revenue by 56% relative to the proportional one.
• The gains came from both ends of the schedule. At the top, higher statutory rates raised revenue despite some compliance losses; at the bottom, lower rates prompted compliance gains large enough to offset the reduced liabilities.
• Taxpayers responded to their own rates. Individuals reacted primarily to their personal tax rate, not to others' rates or to perceptions of the overall schedule's fairness.
• Progressivity plus targeted enforcement shifted the burden upward. When the progressive schedule was scaled citywide with enforcement targeted at high-value properties, the tax burden moved toward wealthier owners.

The authors' conclusion is that progressive property taxation can raise fiscal capacity in low-income settings when paired with strategic enforcement — it is not merely an equity choice purchased at a revenue cost.

Relation to the Georgist Case

The companion Econometrica study established that weak enforcement capacity puts a ceiling on the revenue-maximizing property-tax rate; this paper shows the shape of the schedule is a further, underused lever — and that the compliance responses which create the ceiling behave asymmetrically across the property-value distribution. For the Georgist program the notable result is that recurrent taxation of property can be made progressive without sacrificing revenue — indeed while raising it — in exactly the low-capacity settings where property registries and valuations are weakest. That bears on the standing debate over whether land-and-property taxation is regressive or progressive in incidence (see Land value tax can be progressive): here progressivity was not an incidence side-effect but a design choice that itself expanded fiscal capacity.

The usual caveats: this is a working paper, not yet peer-reviewed; the tax studied is a general property tax rather than a land-only tax; and the revenue result is one city in one campaign, with the burden-shifting result depending on the accompanying targeted-enforcement regime rather than the schedule alone.

Bears On

• Benefit (supports, with caveat): Higher property-tax rates raise welfare in developing countries — experimental evidence that schedule design, not just rate level, expands fiscal capacity in low-income states.
• Benefit: Land value tax can be progressive — an experimental data point that progressive recurrent property taxation is administrable and revenue-positive even in a very low-capacity setting.
• Research: Bergeron, Tourek & Weigel (2024), the state-capacity ceiling — the companion study on rate levels and enforcement from the same team and setting.

See Also

• Bergeron, Tourek & Weigel: The State Capacity Ceiling on Tax Rates
• Brockmeyer et al., Taxing Property in Developing Countries
• Property Tax in Africa (Franzsen & McCluskey)
• Higher property-tax rates raise welfare in developing countries

Sources

1. Gabriel Z. Tourek, Arthur Laroche, Augustin Bergeron, Joana Naritomi, Jonathan L. Weigel & Marina Mavungu Ngoma (2026), "Does Progressivity Raise Tax Capacity? Experimental Evidence from the D.R. Congo," NBER Working Paper 35536 (July 2026). nber.org/papers/w35536 — abstract fetched and read 2026-08-14 — used for the randomized neighborhood-level design, the 56%-revenue result, the top/bottom compliance asymmetry, the own-rate-response finding, and the citywide scaling with targeted enforcement. Scan depth: abstract only (B-claims); no claim below the abstract level is made.