Péter Magyar’s Tisza government says it is dismantling the system Viktor Orbán built over sixteen years. Several of its tools are the ones this course treats as warning signs of democratic erosion: constitutional amendments that removed Orbán-era officeholders, rewrote tenure rules for judges and MPs, and barred Orbán from returning as prime minister, followed this week by the arrest of two former ministers. The question is not whether these moves are legal. Most are. It is whether Magyar is building institutions that can constrain whoever governs Hungary next, or swapping one partisan system for another. Three tests will tell: whether courts can rule against his government, whether accountability runs in both directions, and whether the opposition can compete in 2030. On the evidence so far Hungary passes the first two. The risk sits in the third.

The ambiguity began with the scale of Tisza’s win. Orbán conceded in April after sixteen years, and the final count gave Tisza 141 of 199 seats, above the two-thirds threshold needed to amend the Fundamental Law. The alternation itself matters. Przeworski’s “Botswana rule” treats an incumbent losing an election and leaving office as proof that competition is real, and by 2022 the European Parliament had labeled Hungary a “hybrid regime of electoral autocracy.” The exit was still open. But the same vote handed Magyar the constitutional power that let Fidesz reshape Hungary’s institutions in the first place.

Magyar used that power within weeks. In June parliament imposed an eight-year lifetime cap on service as prime minister, counting past service, which bars Orbán from the premiership for good. In July a second amendment, passed 139 to 6 with Fidesz boycotting, ended President Tamás Sulyok’s term, removed the president of the Constitutional Court, reinstated a retirement age of 70 for its justices (removing four of 15 on September 1), introduced a 12-year tenure cap on MPs that hits more than half of Fidesz’s current deputies, and created an asset-recovery office that can take cases away from prosecutors. That is four of the five pathways to constitutional retrogression that Huq and Ginsburg describe (amendment, weakened checks, politicized executive power, restricted competition), in two bills. Yet the same July amendment restored the Constitutional Court’s full jurisdiction and its right to elect its own president, powers Orbán had stripped. One package removes officials, constrains opposition politicians and hands autonomy back to institutions Orbán weakened. It supports both readings at once.

This is why legality can’t answer the question. Varol’s “stealth authoritarianism” works because it runs through ordinary legal mechanisms whose legitimate and abusive uses look alike, and Levitsky and Ziblatt’s institutional forbearance exists because constitutions permit far more than democrats should do with them. The sharper test, drawn from Varol, is whether institutions are being used in ways that raise the cost of removing the incumbent. The double standard matters too: if Orbán had barred his main rival for life, removed a president by amendment and arrested two opposition MPs in one week, this blog would call it stealth authoritarianism.

The first test is whether Hungary’s courts can constrain Magyar. Ending mandates by amendment sets a precedent any future majority can reuse, and the process was rushed: the Hungarian Helsinki Committee’s lawyers note a five-day public comment window and a one-sentence provision removing the president. The same analysts call other parts forward-looking, because jurisdiction and self-governance went back to the court. So far the courts are acting like courts: a district court in Kecskemét, not the government, ordered pre-trial detention for one of the two former ministers arrested this week. The test is whether judges can rule against Tisza and have those rulings respected. A move against the remaining justices, or against the prosecutor general, would flip this reading.

The second test is whether accountability runs in both directions. On September 28 parliament lifted the immunity of former ministers Miklós Seszták and Balázs Hankó, who were detained the same day over roughly €30 million in alleged bus-tender bribes and €46 million in cultural grants. Both deny wrongdoing; Fidesz calls the cases political revenge. Yet parliament also lifted Magyar’s own immunity in a theft case, at his request. That doesn’t make the justice system neutral, but it cuts against a story in which the law only runs one way. Müller’s “discriminatory legalism,” the full force of the law for opponents and protection for allies, is the pattern to watch. It shows up the day a senior Tisza official faces a credible allegation and the case stalls, or the new asset-recovery office only ever finds Fidesz money.

The third test is alternation, and it is where the risk sits. Orbán cannot return under the new rule, and the 12-year cap will clear many of Fidesz’s most experienced politicians off the 2030 ballot. The Helsinki Committee’s lawyers call that cap the amendment’s most problematic provision, because it falls disproportionately on the opposition and arrived without debate. The term limit also binds Magyar, who leaves by 2034; Orbán’s 2011 constitution tied future majorities’ hands without tying his own. The questions for 2030 are concrete: can Fidesz, or a successor party, field viable candidates, reach voters through plural media and replace Tisza? If the rules now stabilize, the reforms will read as rebuilding a competitive system. If amendments keep arriving whenever a serious challenger appears, they will look like the strategic manipulation Bermeo places at the center of modern erosion.

Hungary’s problem is harder than deciding whether Magyar is “democratic” or “authoritarian.” Institutions built to entrench one government may need dismantling, and dismantling them creates precedents for partisan control. My claim is falsifiable on a calendar: on current evidence the courts are functioning and the law is running in both directions, so the case turns on whether competition survives to 2030. The question is not whether Magyar uses powerful tools. It is whether the system he leaves behind can constrain Magyar too.