Development finance is having a difficult conversation about where its money should go. Budgets are tightening, shareholders expect more impact from every unit of capital, and the reform agenda for multilateral development banks (MDBs) requires every institution to justify its footprint with stronger evidence. In this climate, one question stands out: why should a development bank remain active in countries that already have deep capital markets, EU membership and investment-grade sovereigns?

Our recently published evaluation examined the EBRD’s engagement in twelve economies more advanced in transition over fifteen years, from 2010 to 2024. The time frame is long by design: it captures the euro area crisis, the pandemic, the energy shock and the war on Ukraine. It allows effects to be observed over the time horizons on which markets actually change, rather than according to project life cycles.

The short answer is that relevance and additionality did not decline as these economies matured. They changed shape, but the systems we use to measure them did not change with them. Without adequate monitoring and reporting, MDBs risk underestimating their contribution, missing opportunities for institutional learning, and struggling to articulate their role to internal and external stakeholders. The full report, “Flying together”, is open access.

 

Additionality shifted towards innovative financing structures, standard-setting and policy engagement

The classic case for a development bank’s additionality is the financing gap: providing money that would not otherwise be available, on terms the market will not offer. In mature markets that gap narrows as commercial finance becomes easier to access, and the temptation is to read this as diminishing additionality overall. The evidence points elsewhere. Across the evaluation period, the Bank’s contribution adapted rather than diminished, shifting from filling broad financing gaps towards risk-taking, longer tenors, innovative financing structures, standard-setting and sustained policy engagement. 

Additionality also proved non-linear. Financial additionality became especially pronounced during systemic stress, when the Bank acted as a counter-cyclical investor and confidence anchor. In more stable periods, non-financial additionality gained prominence as client demand evolved towards support for corporate governance, sustainability and other institutional objectives, delivered through advisory work, capacity building and policy dialogue. Notably, this shift did not come at the cost of financial performance. 

The evaluative implication is more interesting than the operational one. If additionality is dynamic, context-dependent and often delivered through implementation rather than at approval, then a binary tick-box applied once, at entry, cannot fully capture it. It systematically under-reports repeat engagements, programmatic work and sector-level effects, and it says nothing about whether the claimed additionality actually materialised.

Our recommendation was therefore to introduce a more systemic framing of additionality, complementing transaction-level assessment with mechanisms that capture portfolio- and market-level dynamics, and closing the loop through regular ex-post verification so that what proved additional in practice informs what is claimed next time.

This challenge is not unique to EBRD. It was also central to a panel on rethinking the additionality of MDBs at Asian Evaluation Week, held in Manila on 1-4 September 2026, where a recurring theme was that the difficulty lies less in defining additionality than in evidencing, monitoring and carrying it through from ex-ante assessment to ex-post learning. The same point emerged at EBRD Evaluation Week in March, as discussed in this blog. For anyone working with the MDB Harmonised Framework for Additionality, that is the live methodological frontier.

 

Demonstration effects are the pathway – and the blind spot

If additionality is the mechanism, demonstration and replication effects are how impact travels beyond a single transaction. This is where the evaluation found both the strongest evidence and the weakest measurement.

Three conditions consistently amplified diffusion: local presence, visibility and credible signalling, and engagement during crisis. But perhaps the most powerful vehicle was the framework – sequenced, repeated engagement over years. Frameworks standardise documentation and due diligence until arrangers and regulators treat them as reference points; they sustain policy dialogue long enough for innovations to enter national practice; and they keep institutional memory alive when market conditions are not yet ready. Frameworks are therefore more effective at embedding a new process or product in a market than a standalone transaction, however innovative that transaction may be.

However, this finding cuts against how most results systems are built. Frameworks can successfully introduce and socialise new ideas without immediately producing visible market change. Lithuania’s covered-bond framework, for example, did not immediately lead to domestic issuance, partly because of market size. Stakeholders nevertheless valued the prepared structures as creating optionality for when market conditions matured. Issuance in Estonia, including transactions using pan-Baltic assets, provided proof of concept for the wider regional framework. The absence of rapid replication, then, is not the absence of impact.

This is precisely what project-bounded, ex-ante results frameworks cannot see. Demonstration effects in advanced markets are indirect, time-lagged and system-level; they materialise after the project closes, sometimes in institutions the project never financed. Where they are not explicitly designed for, and never revisited once diffusion becomes observable, they simply vanish from the institutional record. The Bank, in practice, used more advanced markets as a testing ground for new instruments but did not manage them as one, so the learning was not captured systematically.

 

What this means for the evaluation community

• Match the time horizon to the mechanism. Evaluating market-shaping work only within a project cycle risks producing a false negative. Our fifteen-year window was the methodological choice that made the findings possible.
• Evaluate portfolios, not only projects. The strongest effects came from clusters of related operations and policy dialogue. Unit of analysis is not a technicality; it determines what you are able to find.
• Triangulate with new methods, carefully. We combined AI-assisted sentiment analysis of media coverage across the twelve economies with deeper natural-language analysis in Croatia and Greece. These tools did not replace evaluative judgement; they helped us test hypotheses at scale and surface patterns that project documents do not contain.
• Close the feedback loop. Ex-ante claims that are never verified ex-post are assertions with a budget line attached.

 

Advanced does not mean complete

The wider lesson travels well beyond one bank and one region. Market gaps do not disappear as economies develop – they evolve, becoming narrower, more technical and more institutional. Development banks that continue measuring value as they did when financing gaps were more visible will struggle to evidence their wider contribution, even where their additionality and relevance remain strong.

For evaluators, the task is not to defend continued engagement in richer markets. It is to build the evidence that shows, credibly and ex-post, whether that engagement changed how a market works. With current systems, many institutions cannot yet answer that consistently.

 

Authors’ Bios:

Dr Natalia Kryg, Principal Evaluation Manager II, EBRD Independent Evaluation Department

Dr Natalia Kryg is a Principal Evaluation Manager in the Independent Evaluation Department of the European Bank for Reconstruction and Development (EBRD). She leads and contributes to strategic evaluations covering additionality, concessionality, financial sector development, crisis response, and the development effectiveness of multilateral development banks. Natalia co-led the evaluation of the EBRD’s COVID-19 response and, during a secondment to the International Monetary Fund, contributed to the evaluation of the IMF’s response to the pandemic. Her recent work has focused on additionality, demonstration and replication effects, mobilisation of private capital, and the pathways through which development finance contributes to systemic change. She has authored knowledge papers and presented widely on development effectiveness, financial resilience, international financial institution cooperation, and evaluation methods. Natalia is an active contributor to multilateral evaluation initiatives and communities of practice. She holds a PhD in Economics from University College London.

 

Claire D’Alessio, Consultant Analyst, EBRD Independent Evaluation Department

Claire D’Alessio is a Consultant Analyst for the European Bank for Reconstruction and Development’s (EBRD) Independent Evaluation Department. She supports evaluations covering EBRD corporate processes and the additionality and impact of EBRD engagements in its region. Claire contributed to the recent evaluation of EBRD’s engagement in economies more advanced in transition and has focused on the additionality and demonstration effects of multilateral development bank investment and policy work. Prior to her work with the Independent Evaluation Department, Claire supported the EBRD office of the US Executive Director as an adviser, where she covered project review and institutional accountability processes. Claire holds an MA in International Affairs from the Johns Hopkins School of Advanced International Studies.