Article 73 allocates who decides territorial change but does not itself authorize cession, and read with Articles 2, 8, 9 and 157 it reaches only changes consistent with integrity

The Gist

Article 73 says changes to Ukraine's territory must be decided by a nationwide referendum, and a 2021 law sets out how such a referendum would work. But the same law bars referendum questions aimed at violating territorial integrity, so the best reading is that a referendum can approve small adjustments or swaps, not giving away a region. This is contested: the Constitutional Court has never ruled on where the line falls, some scholars think a referendum route is legally conceivable, and the Court in 2014 named the national referendum as the proper channel for territorial questions. This packet steelmans the argument in its best light, for analysis; it is not an endorsement of its conclusions, and it is not legal advice.

Conclusion

Article 73, which provides that issues of altering Ukraine's territory are resolved exclusively by an all-Ukrainian referendum, allocates who must decide a territorial change but does not itself authorize cession; read with Articles 2, 8, 9 and 157 and the 2021 referendum law, it is best understood to reach only changes consistent with territorial integrity, such as agreed adjustments and exchanges, not the transfer of a region to another state.

Premises

1. Article 73 provides that issues of altering the territory of Ukraine are resolved exclusively by an all-Ukrainian referendum; it states who decides, not what may be decided.
2. Article 8 makes the Constitution the highest law, so a referendum held under Article 73 is subject to Articles 2, 9 and 157 like any other exercise of power.
3. The 2021 referendum law provides for a referendum on altering territory but also bars from any all-Ukrainian referendum a question aimed at violating territorial integrity, so the statute itself draws a line between permitted change and forbidden violation.
4. The most coherent way to give both provisions effect is to read 'altering territory' as covering changes consistent with integrity, such as delimitation adjustments and reciprocal exchanges, and 'violating territorial integrity' as covering the loss of a region or comparable part of the state.
5. On that reading, a referendum could approve an exchange or adjustment, but could not lawfully approve the transfer of a region to another state.

Assumptions

• Contrary to a common assumption, Ukraine does have a referendum statute: the 2012 law was struck down in 2018, but Law No. 1135-IX of January 26, 2021 is in force and expressly provides for a referendum approving the ratification of a treaty that changes Ukraine's territory.
• Ukrainian legal commentators describe the line between a permitted 'change of territory' under Article 73 and a forbidden 'violation of territorial integrity' under Article 157 as unresolved, and the Constitutional Court has not given an official interpretation of it as of October 2026.
• Some international lawyers read Article 73 and the 2021 referendum law together as making a referendum on territorial change 'legally conceivable', and Ukrainian and U.S. negotiators have discussed submitting any final deal to a referendum.
• In its March 14, 2014 decision the Constitutional Court said changes to Ukraine's territory must be resolved by an all-Ukrainian referendum called by the Rada and cited peaceful change of frontiers by agreement under international law, without deciding whether a cession approved that way would violate Article 157.
• A March 2026 survey found 50 percent support for holding a referendum on a peace deal and showed that question wording sharply changes outcomes, so a referendum route is not politically foreclosed; the claim here is legal, not a forecast of how voters would decide.

Analysis

Argument type: Inductive.

Potential Fallacies

• Begging the question (partial circularity) (P4 to P5 and the conclusion) — P4 classifies 'loss of a region' as a 'violation of territorial integrity' and then uses that classification to exclude cession from Article 73. That classification is the contested point, so the interpretive premise carries much of the conclusion. The argument would be stronger if it showed why cession falls on the prohibited side, independent of the harmonizing goal.
• Unsupported inference from coexistence to boundary (P3 to P4) — That the 2021 law contains both a permission for territorial referendums and a bar on integrity-violating questions shows that the two categories are distinct. It does not show where the boundary lies, nor that cession sits on the prohibited side. Other harmonizing readings exist, for example that Article 73 is a specific rule (lex specialis), or that Article 157 binds only constitutional amendments.
• Extension of authority beyond its scope (P4 (Venice Commission and 'recognized reading')) — The Venice Commission's finding concerned removal of a constituent part through a prohibited constitutional amendment under Article 157. Using it to settle whether a referendum-approved treaty cession under Article 73 is barred is an analogy, not a direct holding. The Commission's opinions are also advisory.
• Argument from silence (P1 and the title) — Treating the absence of express cession authority in Article 73 as evidence that cession is excluded is not compelled. Silence on substantive limits is equally consistent with Article 73 being the very mechanism through which the people may decide such questions.
• Overstated certainty / sharp category in a continuous space (P4 and P5) — P5 states categorically that a referendum 'could not lawfully' approve a transfer, although the assumptions concede that the line is unresolved. The distinction between 'adjustment/exchange' and 'loss of a region or comparable part' is also drawn as a binary with no criterion (size, population, reciprocity, consent) for intermediate cases.

Counterarguments

These challenges are not ranked.

• Premise 4 (High impact) — Article 73 is a specific, exclusive competence rule (lex specialis) that exists because the drafters foresaw that some territorial changes, including losses, would require popular decision. Article 157 limits amendments to the Constitution's text, not referendum-approved border treaties. Reading Article 73 to cover only trivial adjustments leaves it nearly empty and violates the presumption against surplusage. The drafters could have written 'exchange' or 'adjustment' if they had meant it.
• Premise 2 (High impact) — Under Article 5 the people are the bearer of sovereignty. A direct all-Ukrainian referendum is the exercise of constituent power and so is not an ordinary act bound by Article 157. On this view the people may authorize what the ordinary amendment process cannot.
• Premise 3 (Medium impact) — The 2021 law expressly provides for a referendum approving ratification of a treaty that changes Ukraine's territory. The paradigm of such a treaty is a cession, so the statute suggests the legislature contemplated real losses. The bar on questions violating integrity may target unilateral secession or dismemberment, as with the Crimean plebiscite. Moreover, if the Constitution authorizes cession, the statutory bar would be unconstitutional to that extent.
• Premise 4 (Medium impact) — The Constitutional Court's 2014 decision cited peaceful change of frontiers by agreement, and some international lawyers see a territorial referendum as 'legally conceivable'. These support the view that consensual territorial change is an available category, and that a referendum-approved cession is not a 'violation'.
• Premise 5 (Medium impact) — The permitted category can swallow the prohibition. A cession can be recast as an 'exchange' or 'adjustment' (a nominal swap, or territory for security guarantees), so the narrow reading offers little real protection and the line between permitted and forbidden collapses on hard cases.
• Conclusion (Low impact) — Martial law currently bars all-Ukrainian referendums and constitutional amendments, so even the permitted referendum route is practically unavailable. The analysis also leaves aside alternative paths, such as a constitutional amendment once martial law ends.

Suggested Improvements

• Comparing rival harmonizations — Set out the main alternative readings explicitly (Article 73 as lex specialis; Article 157 binding only amendments; a popular-sovereignty reading) and show why each fails or is less coherent than the narrowing reading. The claim that the reading is 'most coherent' is currently asserted. A comparative analysis would convert it into a defended inference to the best explanation.
• Bridging Article 157 to referendums — Argue directly why a referendum-approved treaty cession is functionally the same as a prohibited amendment, or why Article 2's indivisibility binds the people acting by referendum. Do not rely only on the Venice Commission's amendment-focused reasoning. This is the key gap in P2 and P4. The Commission's finding concerns amendments and must be linked to the Article 73 setting.
• Defining the line — Offer criteria for distinguishing an 'adjustment or exchange' from 'loss of a region or comparable part', such as reciprocity, proportion of area and population, administrative-unit status, and net effect on the state. Without a criterion the boundary is indeterminate and open to gaming. P5 depends on it.
• Engaging the 2021 law's treaty provision — Quote the statute's treaty-ratification referendum clause and explain what class of treaty it covers if cessions are excluded. This is the strongest textual evidence for the opposing reading. Addressing it would answer the surplusage objection.
• Calibrating P5 — Reframe P5 to match the conclusion's hedge, for example 'on the better reading, a referendum under Article 73 would not by itself validate a transfer of a region'. The categorical wording outruns the acknowledged absence of a Constitutional Court ruling.
• Context and sourcing — Cite the Venice Commission opinion precisely, note its non-binding status, add the martial-law bar on referendums and amendments, mention the alternative amendment route, and state an 'as of' date for the 'no Court interpretation' claim. This would make the argument more accurate and more useful to practitioners and negotiators.

Scenario Tests

• The Constitutional Court issues an official interpretation that Article 73 is a specific rule permitting a referendum approving cession, or that a sovereign-people referendum is an exception to Article 157. (Challenges) — This would directly refute the conclusion. The argument's standing depends on the absence of such a ruling, and the Court's independence under wartime pressure is a relevant uncertainty.
• The Court adopts a harmonizing reading and holds that a referendum cannot validate the transfer of a region, so such a transfer would require a constitutional amendment. (Supports) — This would confirm the reading and show that Article 73 and Article 157 operate as complementary rather than conflicting norms. It would also confirm that a real cession requires the amendment route, which is itself barred during martial law.
• A peace proposal pairs a small reciprocal swap of border districts with a modest net loss of populated land, and is presented as a 'delimitation adjustment' for a referendum. (Challenges) — This tests the adjustment/violation boundary. The argument offers no criterion, so it cannot say whether the proposal falls on the permitted side. It shows that the framing can be exploited and that a workable test is needed.
• Negotiators propose a referendum on a deal that formally recognizes transfer of an administrative region to Russia. (Supports) — On the argument's reading this question would be one 'aimed at violating territorial integrity' and would be barred under the 2021 law and Articles 2 and 157. Opponents could challenge it in the Constitutional Court. This is the paradigm case the argument addresses, though the reading remains unresolved.
• Martial law remains in force through any peace-deal ratification window. (Neutral) — No all-Ukrainian referendum or constitutional amendment could be held, so the Article 73 question is academic in the near term. The legal analysis stays relevant for later periods and for how a deal is structured.
• Polling shows majority support for a referendum on a peace deal, but results shift with question wording. (Neutral) — As A5 states, the claim is legal and not a forecast. Political support does not alter the constitutional limits and does not show that the reading is wrong.

Coherence & Relevance

The argument is internally coherent and honest about its limits. From P4 it moves cleanly to P5 and the conclusion, and the hedged wording 'best understood' matches the abductive character of the central inference. Its weaknesses are concentrated in a single step, the interpretive classification of cession as a 'violation' of integrity, which is partly circular and not tested against rival harmonizations. The supporting authorities (the 2021 statute, the Venice Commission, the 2014 decision) point in several directions or address neighboring questions, and the assumptions candidly concede that the Constitutional Court has not resolved the line. The result is a plausible, well-hedged interpretive case rather than a demonstrated one.

• Article 73 provides that issues of altering the territory of Ukraine are resolved exclusively by an all-Ukrainian referendum; it states who decides, not what may be decided. (Strong) — It sets the textual base, but the 'who, not what' gloss is interpretive. It does not itself show that substantive limits apply.
• Article 8 makes the Constitution the highest law, so a referendum held under Article 73 is subject to Articles 2, 9 and 157 like any other exercise of power. (Strong) — It establishes that other constitutional norms bind a referendum, but Article 157 addresses amendments. The link to referendums requires a further argument about its reach, and a lex specialis reading of Article 73 is not excluded.
• The 2021 referendum law provides for a referendum on altering territory but also bars from any all-Ukrainian referendum a question aimed at violating territorial integrity, so the statute itself draws a line between permitted change and forbidden violation. (Moderate) — The statute shows that the two categories are distinct but does not locate the boundary, and as a subordinate law it cannot settle constitutional meaning. Its treaty-ratification provision is not reconciled with the narrow reading.
• The most coherent way to give both provisions effect is to read 'altering territory' as covering changes consistent with integrity, such as delimitation adjustments and reciprocal exchanges, and 'violating territorial integrity' as covering the loss of a region or comparable part of the state. (Strong) — It is the pivot of the argument but is an abductive claim. It does not rank rival harmonizations, supplies no threshold between adjustment and loss, and risks surplusage.
• On that reading, a referendum could approve an exchange or adjustment, but could not lawfully approve the transfer of a region to another state. (Strong) — It follows from P4 and is only as strong as P4. Its categorical wording is not matched by the unresolved status acknowledged in the assumptions.

View this argument on LogicFirst.ai