Ratification Roulette: The Long Road from Signature to Reality

There is something almost theatrical about the signing of a trade agreement. The leather folder, the heavy pen, the handshake calibrated for the cameras. Two leaders smile. The flashbulbs confirm: something has been accomplished. The press releases go out. "Historic." "Landmark." And then the document enters a labyrinth from which it may not emerge for a decade, if it emerges at all.

The European Union and Australia signed their free trade agreement in 2026 after negotiations that stretched back to 2018. For most of the public, that signature looked like an ending. A deal, done. But anyone who has watched EU trade policy knows that the signature is where the uncertainty begins.

The Ceremony and What Comes After

The EU does not ratify trade agreements the way most countries do. There is no single chamber that votes yes or no. The process is layered, sequential, and profoundly dependent on a classification that most people outside Brussels have never heard of: whether the agreement is "EU-only" or "mixed."

An EU-only agreement requires approval from the Council of the European Union and the European Parliament. That is complex enough. A mixed agreement, which touches on areas where member states retain their own competence, must also pass through every single national parliament. In Belgium, that means not just the federal parliament but the regional ones too, including Wallonia, Flanders, Brussels-Capital, and the German-speaking community. Twenty-seven member states, each with their own constitutional requirements, their own political calendars, their own domestic pressures.

In May 2017, the Court of Justice of the European Union delivered Opinion 2/15 on the EU-Singapore free trade agreement. The ruling clarified what belongs to the EU's exclusive trade competence and what does not. Portfolio investment and investor-state dispute settlement, the court said, require national ratification. Everything else, the bulk of any modern trade deal, falls under Brussels alone. That opinion reshaped how the Commission designs trade agreements. But it did not eliminate the problem. It only sharpened the question: can the EU-Australia agreement be structured to avoid national parliaments entirely?

The Walloon Moment

To understand why that question matters, consider what happened in October 2016. The Comprehensive Economic and Trade Agreement between the EU and Canada, years in the making, was ready for signature. The European Parliament was expected to consent. The Council had authorized it. And then the parliament of Wallonia, a Belgian region of 3.6 million people, said no.

Paul Magnette, the Minister-President of Wallonia, refused to grant the Belgian federal government the authority to sign. His concerns centered on investor-state arbitration and the potential impact on Walloon agriculture and public services. For several days, the entire EU-Canada trade relationship hung on the decision of a regional parliament that most Europeans could not locate on a map.

Was this an absurdity? Or was it democracy functioning exactly as designed?

The answer depends on what you believe a democratic process owes its citizens. Belgium's constitutional structure requires all regional parliaments to consent before the federal government can bind the country to an international treaty. Magnette was not obstructing; he was exercising a right that Belgian law explicitly grants. The CETA signature eventually went ahead on 30 October 2016, after additional interpretive declarations were attached to address Walloon concerns. But the episode revealed something that the architects of EU trade policy already knew and preferred not to discuss: in a union of 27 democracies, any single democratic body can become a veto point.

A Treaty in Limbo

CETA was signed in October 2016. Provisional application began on 21 September 2017, covering those parts of the agreement that fall under exclusive EU competence. Trade in goods, services, public procurement, sustainable development provisions - these went into effect. But the investment court system, the very mechanism that had triggered the Walloon revolt, remained suspended, waiting for full ratification by all member states.

As of early 2026, that ratification is still incomplete. Ten of the EU's 27 member states have not ratified, among them Belgium, France, Italy, Ireland, and Poland. France's National Assembly approved CETA in July 2019, but the Senate voted against ratification in March 2024, rejecting the agreement 211 to 44 amid concerns over agricultural competition and food sovereignty. The National Assembly would need to override that rejection, and the political appetite to do so has not materialized. Other member states have stalled for their own reasons, some political, some procedural, some constitutional, as in Ireland where the Supreme Court ruled in 2022 that ratification required changes to existing law.

The result is a treaty that exists in a kind of half-life. It functions, mostly. Businesses trade under its provisions. Tariffs have been reduced. But the full architecture of the agreement, particularly the investment protections, remains in suspended animation. Nearly a decade after signing, CETA is neither fully alive nor fully dead.

Is this sustainable? The European Commission insists that provisional application works well enough. Critics argue that a permanently provisional treaty undermines legal certainty and signals to trading partners that the EU cannot deliver on its own commitments. Both are right, which is precisely why the problem persists.

Mixed or Exclusive: The Legal Fault Line

After Opinion 2/15, the Commission adjusted its approach. The EU-Japan Economic Partnership Agreement, which entered into force in February 2019, was deliberately structured as an EU-only deal. Investment protection was carved out into a separate agreement that would go through national ratification on its own track. The trade deal itself sailed through the Council and the European Parliament without needing approval from 27 national legislatures.

That model worked for Japan. Whether it works for Australia depends on the political choices made during the final negotiation phase. If the EU-Australia FTA includes investment protection provisions within the main agreement, it becomes mixed, and every national parliament gets a vote. If those provisions are separated, the core trade deal can be ratified more quickly, but the investment chapter faces its own uncertain journey.

The Commission, having watched CETA spend a decade in ratification limbo, has every incentive to pursue the EU-only route. But member states, particularly those whose parliaments view trade ratification as one of their few remaining levers of influence over EU external policy, may resist being cut out of the process. The tension between efficiency and democratic legitimacy is not a design flaw. It is the central architectural question of the European Union itself.

What the Pen Cannot Promise

The EU-Australia free trade agreement now enters the same procedural landscape that CETA has been navigating since 2016. The timeline, if the CETA precedent holds any predictive value, stretches well beyond any single electoral cycle. Five years from signature to full ratification would be optimistic for a mixed agreement. Ten years is not unusual. And during that time, governments change, political alignments shift, new crises emerge that push trade ratification further down the legislative agenda.

Provisional application may cushion the wait. If the agreement is structured carefully, the bulk of its trade provisions could take effect within a year or two of signature, while the full ratification grinds forward in the background. But provisional application has its own fragility. It depends on continued political will, and it cannot deliver the parts of the agreement that require the deepest democratic buy-in.

There is something worth sitting with in all of this. The ceremony of signing suggests that political leaders control the process. They do not. The process controls them, or more precisely, it belongs to dozens of parliaments, thousands of legislators, and millions of voters who may never read a single clause of the agreement that will shape their economic lives for decades.

Perhaps that is exactly as it should be. The gap between the speed at which geopolitics moves and the pace at which democracies deliberate is not a bug to be patched. It is the price of requiring consent from the governed. The pen on the leather folder was always just the beginning of a conversation. Whether that conversation reaches a conclusion, and how many years it takes to get there, is not something any signature can guarantee.

See also: When Courts Draw the Line: Budapest, Rome, and the Limits of EU Governance for another perspective on institutional tensions within the European Union.