Miguel Ângelo Vilela Rodrigues (University of Minho), with Oliver Meza (CIDE) and Carmen Navarro (Universidad Autónoma de Madrid)

Gerrymandering is usually filed away as an American problem — a decennial ritual of oddly shaped congressional districts and litigation. But the practice does not require an American ballot. It requires only two things: a boundary that has to be redrawn, and a politician with an interest in where the line falls. Territorial reform supplies both, and it does so in countries that would never describe themselves as having a redistricting problem.

Portugal is a case in point. In our recent article in Local Government Studies, we look at the 2013 reform that amalgamated Portuguese sub-municipal units — the freguesias, roughly comparable to parishes — and ask a question that the amalgamation literature has largely left aside: not whether mergers deliver economies of scale or damage local democracy, but who decides which units merge with which, and what they get out of it.

A reform designed to be captured

The reform came out of the 2011 Memorandum of Understanding with the troika. Cutting the number of local entities was one of the commitments made, and central government initially tried to impose specific merger criteria from above. Local government rejected that outright. So the government softened its approach: it kept the targets — urban units to be reduced by 50–55 per cent, rural ones by 25–35 per cent — and handed the map itself to the mayors. Across 278 municipalities, the number of sub-municipal units fell from 4,269 to 3,092.

The design has an unusual feature. The units that had to merge were not the units that decided. Mayors drew the proposals; municipal assemblies approved them. And Portuguese municipal assemblies are hybrid bodies: alongside directly elected members, every sub-municipal president in the jurisdiction sits with a full vote. Sub-municipal presidents are also, in practice, part of a mayor’s electoral machinery — the people who campaign, mobilise and deliver a vote in their own patch.

That makes an amalgamation reform something more than an administrative exercise. Reducing the number of sub-municipal units aligned with the mayor’s party reduces the mayor’s support in the assembly that approves the budget and the investment plan, and thins out the voices in the field who will be needed at the next election. Conversely, a well-drawn map can pack opponents into a single unit or dilute them inside a larger friendly one. The pencil was handed to precisely the actors with most to gain from how it moved.

Vulnerability and feasibility

We modelled the mayor’s decision on two dimensions. The first is vulnerability: does the mayor have a majority of politically aligned sub-municipal units, or not? The second is feasibility: does the mayor command a majority of the elected seats in the municipal assembly, and so the ability to get a map approved without concessions?

Crossing them yields four situations. Where mayors are not vulnerable, redrawing is either irrelevant or indifferent — there is nothing to gain. Where they are vulnerable but lack an assembly majority, redrawing is crucial to survival but has to be negotiated, which means sharing the spoils. Where they are vulnerable and hold the assembly, redrawing is appealing: the incentive and the means coincide.

Using the 2009 sub-municipal election results projected onto the new 2013 map — the only information mayors actually had when drawing it — we then estimated multinomial logit models on all 278 affected municipalities.

The pattern is clear. Across the country, 51 per cent of municipalities ended up with more aligned sub-municipal units after a reform whose entire purpose was to reduce their number. In the appealing quadrant, the odds of an increase in aligned units rise more than nine-fold; in the crucial quadrant, where the map has to survive a coalition, they rise just over four-fold. The gap between the two is itself the finding: opposition parties will not sign off on a map that benefits the incumbent alone, so the gains get shared and diluted. Results hold when we substitute a vote-efficiency measure of the kind used in the US gerrymandering literature.

Why this should interest anyone outside Portugal

Two implications travel. The first is procedural. Structural reform is often delegated downwards on the sensible-sounding grounds that local actors know their own territory. They do — and they also know their own electoral arithmetic. Where the reform reshapes the boundaries that determine who sits in the body that scrutinises the reformer, delegation is not neutral. Anyone currently watching local government reorganisation in England, where boundaries, wards and the composition of new authorities are all in play, may find the logic familiar.

The second is about what happens next. Since 2013 there has been sustained pressure in Portugal to demerge some units, and the initiative for those petitions again rests with mayors. If the same rationality holds, demergers will be permitted where they are politically harmless and blocked where they are not — meaning the same territories get gerrymandered twice.

None of this requires bad faith or illegality. Every map was lawful and approved through proper procedure. That is rather the point: an effective opposition, not a rule about compactness, was the only thing that reliably constrained the outcome.

Miguel Ângelo Vilela Rodrigues is Associate Professor with Habilitation in the Department of Political Science, School of Economics and Management, University of Minho, and a researcher at CICP.

The full article — ‘Gerrymandering to survive: an explanation of the political conditions that shaped mayors’ decisions over an amalgamation process in Portugal’, by Miguel Rodrigues, Oliver Meza and Carmen Navarro — is published in Local Government Studies, 51(5), 993–1015. DOI: 10.1080/03003930.2024.2404104