New comparative work across forty European contexts finds five governance configurations for just transitions and no universally superior one. That is not a disappointing result. It is the finding, and it changes what a successful pilot is evidence of.
Comparative research based on the ESPON NoStaGeo project has examined forty European institutional contexts and identified five broad governance configurations for just green transitions, ranging from centralised policy-driven approaches to community-led models. The researchers report no universally superior arrangement. Effectiveness depends on matching multilevel coordination, territorial flexibility and stakeholder inclusion to local institutional conditions.
Results of that shape are often read as inconclusive, which is a misreading. A finding that no configuration dominates is a substantive claim about how transition policy works, and it has a direct consequence for how European programmes treat a successful demonstration.
The underlying idea has a long history in comparative political economy. Peter Hall and David Soskice built Varieties of Capitalism around institutional complementarities: the observation that institutions in different spheres of a political economy, labour markets, corporate finance, skill formation and inter-firm collaboration, reinforce one another, so that a configuration performs well as a package and poorly when a single element is transplanted.
Their specific finding concerned innovation, with liberal market economies better placed to sponsor radical innovation and coordinated market economies better placed for incremental innovation. The general mechanism matters more than the taxonomy. An institution's performance is conditional on the institutions around it, which means that an instrument which works in one configuration is not thereby demonstrated to work anywhere.
Apply that to the transition instruments that circulate through European policy networks. A renewable auction, a low emission zone, a coal region fund, a citizens' assembly, an electric vehicle incentive and a one-stop permitting shop are all presented as portable designs with documented results.
Each depends on complements that are rarely named alongside them. An auction requires a credible counterparty, a bankable contract and a developer market deep enough to produce competitive tension. A low emission zone requires enforcement capacity, an alternative mode for displaced drivers and a political settlement about who is exempted. A coal region fund requires a regional administration able to originate projects and absorb money on a disbursement schedule, which is the clearance point problem that defeats small administrations regardless of the funding available.
So a successful pilot is evidence about an instrument operating inside a particular configuration, and the configuration is usually unrecorded. That is why replication results disappoint so regularly, and the disappointment is typically attributed to implementation quality rather than to the fact that the instrument was never characterised together with its complements.
If configuration determines outcome, then an instrument should travel with a description of what it requires, and that description is assemblable.
Administrative capacity at the implementing tier. Staff numbers, technical competence and whether the implementing body can procure, contract and supervise without external help.
Fiscal autonomy. Whether the implementing tier controls a budget, can carry money across years, and can commit beyond a political term.
Enforcement reach. Whether a rule can be monitored and a breach sanctioned. Several European instruments assume this and most of their documented successes come from jurisdictions where it holds.
The incumbent position. Who loses, how organised they are and whether they have a veto point. This is frequently the variable that decides, and it is almost never in the instrument description.
Trust in the implementing institution. Participation instruments in particular work very differently where people expect to be heard and where they do not, and prior experience of a scheme that did not deliver is a durable and well documented obstacle.
And the existing practice being displaced. Every instrument competes with something that already works well enough, a point we have made about cookstoves and that applies identically to a policy.
Six variables, each observable, each recordable at the point an instrument is documented as successful. A European instrument library that carried them alongside each design would be a more useful product than another set of case studies, and the ESPON work has effectively produced the beginning of one.
European programmes increasingly ask projects to demonstrate replicability, and the usual response is to pilot in several countries and report that the instrument worked in each.
If configuration determines outcome, that design tests the wrong thing. Piloting in three similar jurisdictions demonstrates robustness within a configuration and says nothing about transfer across configurations. The informative design is the opposite, deliberately selecting contexts that differ on the variables above, and treating a failure in one of them as the most valuable observation in the set rather than as a reporting problem.
That is difficult to fund, because a consortium is rewarded for success in every site. It is also the only way to learn which complements are load bearing, since an instrument that works everywhere it was tried has taught nobody anything about its requirements.
The same argument extends beyond Europe, and this is where the comparison has analytical rather than decorative value. The configuration variables above take values in European contexts that cluster within a fairly narrow range by global standards. Testing the same instrument where administrative capacity, fiscal autonomy or enforcement reach sit well outside that range is what reveals which complements were doing the work, and that is a reason for European research to design in non-European sites rather than to add them for dissemination.
The configuration, recorded at pilot stage. The six variables above, scored for every site before results are reported, so that outcomes can be read against context rather than averaged across it.
Failures, retained in the dataset. The sites where an instrument did not work hold the information about what it needs, and they are the ones most likely to be dropped from a final report.
Which complements were present without being noticed. Established by asking implementers what they relied on that was not in the design. This is ordinary qualitative work and it is the only route to the variables nobody thought to record.
And whether a transferred instrument was modified. An instrument that arrives intact and fails, against one that is adapted and succeeds, are different findings, and reporting usually collapses them into whether the project met its targets.
The ESPON result is being described in some quarters as showing that governance of just transitions is context-specific, which is true and sounds like a counsel of despair. It is better read as an instruction. Characterise the context with the same care as the instrument, and a good deal of what currently looks like inconsistent policy performance becomes predictable.
The Lab works on this through measuring change and reporting to funders, and on what an instrument meets when it arrives somewhere through entering a new context.
If you are designing a replication work package and want the context recorded as data rather than as background, tell us what you need to know.
If this reads close to a decision you are weighing, we would like to hear about it. A short note is enough to start.
The Lab is an independent research team, not an advocacy organisation. We do not campaign, endorse products, or accept commissions where a finding is expected in advance. Where a partner's own theory of change is what needs testing, that is a legitimate engagement, and the Lab tests it honestly.