When does confidentiality make mediation less effective?

We tend to treat confidentiality in mediation as an unquestioned strength. But it also comes at a cost that may be holding mediation back from becoming mainstream. If mediation is meant to reduce uncertainty, what happens when confidentiality removes all consequences of failure?

In much of continental Europe, mediation failure leaves no trace. There is no record of where parties stood, no procedural consequence, and no real effect on what comes next. For large companies, a failed mediation is usually absorbed as another cost item. Without legal insurance or the financial capacity to continue, a failed mediation can effectively be the end of the road. If the next step is too slow, too expensive, or simply unrealistic, failure is not neutral. It closes options.

Mediation is frequently defended by pointing to high settlement rates. Across jurisdictions, a large majority of mediations do end in agreement. But that statistic says as much about what enters the system as it does about what happens inside it. If in one jurisdiction 90% settles, but only 300 cases were mediated, there is not too much to celebrate.

Where failure carries no consequence, parties tend to bring cases that are already likely to settle. Harder disputes, where there is real uncertainty, power imbalance, or a genuine risk of non-agreement, are less likely to be tested in mediation at all. The result is a system that appears highly effective, but only within a relatively narrow band of cases.

In England and Wales, courts may examine how parties have engaged in mediation. An unreasonable refusal, or a purely formal participation, can later affect cost decisions. That alone changes behaviour during the mediation.

In Canada and Australia, mediation is often embedded in procedural timelines. Even where no agreement is reached, the way parties engaged, the offers made, and the level of preparedness can shape what follows. In some contexts, elements of that engagement may even be taken into account later.

Back home in the Netherlands, employment mediation did not gain traction because parties suddenly became more willing to engage. It became more widely used once institutions began attaching consequences to refusal. When engagement started to matter procedurally, mediation stopped being a purely optional step.

Business mediation largely lacks that kind of institutional backing. If failure has no effect, it is easy to discount the process altogether. Optional processes tend to attract optional effort, and high success rates can coexist with a relatively small and cautious market.

From a business perspective, that raises a more fundamental question. The issue is not whether parties can be persuaded to like mediation. It is whether the system is designed in a way that makes engagement matter, even when agreement is not reached.