Bifurcation in Arbitration: The Fix Everyone Keeps Proposing Already Exists

What bifurcation is and why it appeals

Bifurcation splits an arbitration into separate phases — typically jurisdiction or another preliminary objection first, merits second — on the logic that resolving a threshold issue early can moot the rest of the case entirely. Tribunals also bifurcate liability from quantum, or separate any issue whose early resolution would narrow what remains to be argued. The appeal is obvious: if a tribunal lacks jurisdiction, nothing is gained by hearing the merits regardless.

The case against bifurcation, and why it persists

The empirical case against bifurcation is genuinely strong, and has only hardened with more data. Lucy Greenwood’s original 2011 study of ICSID cases found bifurcated proceedings took an average of 3.62 years to conclude against 3.04 years for non-bifurcated ones. Her 2018 follow-up, reviewing a larger set of concluded cases, found the gap had widened: bifurcated cases averaged four years three months against three years two months for non-bifurcated cases — and, tellingly, cases where the jurisdictional objection ultimately failed took an average of five years two months to conclude, meaning an unsuccessful bifurcation attempt is close to the worst possible outcome for time and cost.

Tribunals nonetheless keep bifurcating, because the underlying tension is real. A respondent should not have to defend a case on the merits where the tribunal may well lack jurisdiction to hear it at all — but a jurisdictional objection can equally be a dilatory tactic dressed up as a threshold issue, and a tribunal has to weigh procedural fairness against the risk of delay and prejudice without perfect information about which one it is looking at.

The fix is already here — just not as a new institution

Rather than a new supranational body to screen preliminary objections before any tribunal is even constituted, the actual reform underway addresses the delay problem inside existing arbitral rules. ICSID’s 2022 Arbitration Rules introduced Rule 44, imposing strict timelines on the bifurcation process itself: a tribunal must decide whether to bifurcate promptly after a request is filed, and where bifurcation is granted, it must decide the preliminary objection within 180 days of the final submission on it. Early data from cases decided under the new rules is encouraging — tribunals are ruling on bifurcation requests in an average of roughly two months, are granting bifurcation far more sparingly than before, and are meeting the 180-day deadline once bifurcation is granted.

The reform that actually works targets the clock, not the forum — and Nigeria adopted a version of it in 2023. Nigeria’s own Arbitration and Mediation Act 2023 moved in the same direction on a different point in the process. Under the old Arbitration and Conciliation Act, a tribunal’s ruling on its own jurisdiction — the competence-competence principle — was effectively final, with no structured route to court review before the award stage. The 2023 Act now allows a party to apply to the court to determine a jurisdictional question within 30 days of the tribunal’s preliminary ruling on it, giving jurisdictional objections a fast, bounded review path without derailing the arbitration into open-ended litigation over its own competence.

Both reforms share a structural insight the case for a new global institution overlooks: party autonomy and the competence-competence doctrine — a tribunal’s inherent authority to rule on its own jurisdiction — are not obstacles to be routed around by a supranational body, but the foundation the whole system runs on. What arbitral users actually needed was not a different decision-maker, but a bounded clock on the decision-maker they already have. ICSID and Nigeria have each supplied a version of that clock inside their existing frameworks, without touching who decides.

What this means in practice

For parties arbitrating under ICSID rules or seated in Nigeria, the practical response to the bifurcation debate is no longer theoretical: a jurisdictional objection can be raised with real confidence that it will be resolved on a defined timeline rather than open-ended delay, which changes the calculus on whether raising it is worth the cost. For arbitrations under institutional rules that have not adopted equivalent timelines — many ICC, LCIA and ad hoc UNCITRAL proceedings still leave bifurcation entirely to tribunal discretion without a deadline — parties should consider negotiating procedural timetables that build the same discipline in by agreement, since the rules themselves may not supply it.

Key takeaways and conclusion

Greenwood’s updated data shows the case against bifurcation has strengthened since 2011: bifurcated ICSID cases now average over a year longer than non-bifurcated ones. ICSID’s 2022 Rule 44 imposes strict timelines on bifurcation decisions — a structural fix inside the existing framework, not a new institution. Nigeria’s Arbitration and Mediation Act 2023 gives parties a bounded 30-day route to court review of a tribunal’s jurisdictional ruling, departing from the old Act’s finality rule. The better answer to the bifurcation debate targets the clock on the existing decision-maker, not a new decision-maker altogether.

The call for a wholly new institution to pre-screen jurisdictional objections misreads where the real inefficiency in bifurcation sits. It was never really about who decides a preliminary objection — tribunals, exercising competence-competence, are well placed to do that. It was about how long the decision takes once bifurcation is granted. ICSID and Nigeria have each answered that question inside their own frameworks in the last few years, and any arbitration user structuring a clause today should be asking which set of rules gives them that discipline before a dispute ever arises.