The Institutions That Survive Autocrats — and How to Build Them

Democracies, it turns out, die in ways that are boringly predictable. The playbook is now familiar: pack the courts, capture the electoral commission, starve independent media of advertising, prosecute opposition figures on corruption charges that are technically legal and substantively absurd. What is less well understood is how democracies recover. The literature on backsliding is vast; the literature on reconstruction is thin. This is, to put it mildly, a problem. For if 72 countries have experienced democratic erosion since 2010, according to V-Dem's most recent data, it would be useful to know which institutions, once rebuilt, give democracies the best chance of staying rebuilt.

The evidence from the twentieth century's democratic recoveries — Spain after Franco, Chile after Pinochet, Central Europe after communism, South Korea after military rule — suggests three institutions matter above all others. Get these right, and the odds of sustained recovery improve dramatically. Get them wrong, and the next strongman merely has to wait.

The Numbers

The pattern is striking. Countries that preserved some measure of judicial independence during authoritarian rule — even if only in commercial or administrative law — had markedly better prospects for democratic consolidation afterwards. Spain's commercial courts continued to function with reasonable professionalism under Franco; Chile's lower courts retained surprising autonomy even as Pinochet controlled the Supreme Court. These pockets of legal normalcy became the institutional memory around which democratic judiciaries could be reconstructed.

Civil service autonomy matters nearly as much. Where authoritarian regimes completely captured the bureaucracy — replacing professional administrators with party loyalists from top to bottom — reconstruction proved agonisingly slow. Poland's lustration laws, which banned communist-era officials from public service, are often cited as a success. Less remarked is that Poland retained a cadre of technically competent civil servants who had joined the bureaucracy in the 1970s for apolitical reasons. Hungary, by contrast, purged and counter-purged its civil service so thoroughly that no institutional knowledge survived. Three decades later, Viktor Orbán found a bureaucracy with no memory of independence to resist him.

A Familiar Pattern

The third institution — local government — is perhaps the most overlooked. Scholars of democratic breakdown focus, understandably, on national politics: the capture of parliaments, presidencies, and supreme courts. But the historical record suggests that where local elections survive in some form, democratic muscles continue to be exercised. Voters learn, or remember, the habits of accountability. Politicians learn, or remember, the skills of coalition-building. When national democracy becomes possible again, these skills transfer upward.

Spain's transition illustrates the pattern vividly. Franco permitted municipal elections in small towns, calculating — correctly — that national parties would struggle to organise at that level. But by 1977, when national elections returned, Spain had thousands of citizens who had served as mayors, councillors, and local administrators under competitive (if constrained) conditions. These formed the backbone of democratic governance.

The Mechanism

Why do these three institutions matter more than others? The answer lies in what political scientists call "institutional stickiness" — the tendency of organisational cultures to persist even when formal rules change. Courts, civil services, and local governments are characterised by dense networks of personnel who spend entire careers within the institution. They develop norms, practices, and informal codes that can survive changes in political leadership.

Electoral commissions and press freedom laws, by contrast, are more easily captured and recaptured. An autocrat can replace a five-member electoral board in an afternoon. He cannot replace 5,000 municipal administrators, or 500 judges scattered across provincial courts, without massive disruption. The very diffuseness of these institutions protects them. Their capture requires sustained effort over years, which most autocrats either lack the patience for or consider unnecessary.

There is also a psychological mechanism at work. Citizens who have never experienced accountable government — who have no memory of courts ruling against the state, of civil servants saying no to politicians, of local elections producing peaceful transfers of power — find it harder to demand these things. Institutional memory is not merely bureaucratic; it is cultural. Where institutions of accountability survive, even in diminished form, they keep alive the idea that such things are possible.

What Is Being Done

International efforts to support democratic resilience have, in recent years, focused heavily on the wrong targets. The European Union's rule-of-law mechanism, triggered against Hungary and Poland, concentrated on constitutional courts and media independence — important, but not the institutions with the highest predictive value for recovery. USAID's democracy programming has emphasised election monitoring and civil society support, activities that matter during transitions but contribute less to long-term resilience.

The World Bank and regional development banks have begun to recognise the importance of civil service professionalisation, but their programmes focus overwhelmingly on efficiency and anti-corruption rather than institutional autonomy. A well-trained bureaucrat who answers to an autocrat is not a bulwark of democracy; he is merely a competent instrument of its destruction.

Domestic reformers in backsliding democracies face an agonising dilemma. The institutions most worth protecting — lower courts, local governments, professional civil service — are precisely the ones that attract the least public attention. Defending a provincial administrative court lacks the drama of defending a supreme court justice. Campaigns to protect municipal election authority generate fewer headlines than campaigns against media takeovers. Yet the historical evidence suggests these quieter battles may matter more.

What Should Be Done

The implications for policy are reasonably clear, if difficult to implement. First, constitutional designers should decentralise judicial appointments and civil service hiring as much as possible. Hungary's constitution concentrated judicial appointments in a single council that Orbán captured in 2012; Germany's Länder-based judicial appointments have proven far more resilient to federal interference. The principle extends beyond courts: any institution that can be captured by controlling a single node is a bad design for democracy.

Second, defenders of democracy should invest heavily in local government capacity. This means resisting the centralising tendency of modern governance — the urge to standardise, nationalise, and consolidate. Local government is messy and inefficient, but it is also democracy's fallback position. When national institutions fail, local institutions can preserve democratic practice until recovery becomes possible.

Third, international democracy assistance should reorient toward supporting institutional diffusion rather than institutional excellence. It is better to have 500 mediocre courts that an autocrat cannot easily capture than five excellent courts that he can. Democracy assistance has long prized "capacity building" and "best practices"; it should prize resilience and redundancy instead.

The Long Game

None of this offers comfort to citizens living under authoritarian rule today. The time horizon for democratic recovery, even under favourable conditions, is measured in decades. Spain's transition took seven years from Franco's death to consolidated democracy; Chile's took fifteen; South Korea's took twelve. These were success stories. Argentina has been cycling through democratic breakdowns and recoveries for a century.

But the historical record offers something better than comfort: it offers a strategy. The institutions that autocrats find hardest to capture — the diffuse, the local, the unglamorous — are the institutions most worth defending. And where they survive, democracy has a chance to return. Franco ruled Spain for 36 years. The municipal elections he permitted as a harmless concession became the seedbed of his regime's replacement. Autocrats, it seems, are not always good at identifying what threatens them. Their opponents should be.