Introduction
In April 2023, the government launched an experimental project to restore a number of localities damaged by the war. The project was to be distinguished by its comprehensive approach — not merely rebuilding individual structures or infrastructure facilities, but engaging in holistic planning and transformation of the affected localities or their parts.
The initiative was expected to generate new, effective solutions for restoring localities whose infrastructure had suffered severe destruction due to Russian aggression and required a comprehensive, rather than fragmented, recovery approach.
Six localities were included in the experiment:
- The urban-type settlement of Borodianka and the village of Moshchun in Kyiv region
- The town of Trostianets in Sumy region
- The village of Posad-Pokrovske, located on the border of Kherson and Mykolaiv regions
- The village of Tsyrkuny in Kharkiv region
- The village of Yahidne in Chernihiv region.
Problems arose, however, in the very first year of implementation. Because there was no single approach to selecting restoration projects, the rebuilding of Moshchun dropped out of the experiment, and of the more than three hundred restoration facilities planned, work was completed on just one. In mid-2024, TI Ukraine found that the successful launch of the first phase had been undermined by several factors at once:
- gaps in the terms of the experiment;
- delays in financing;
- instability in the leadership of the experiment’s key stakeholders — the Ministry of Infrastructure and the Agency for Restoration;
- the failure to transfer construction contracting-authority rights.
No substantial progress was made during the second year of implementation either. Meanwhile, the list of restoration facilities doubled compared with the original — from 295 to 593. The problems with the actual disbursement of financing, which had appeared at the very start of the project, made it impossible to implement on time and efficiently. As a result, restoration was completed on a further 57 facilities during the second year, and on slightly more than 9% of the total planned number over the two years combined.
2025 was to be the final year of the experimental project. We therefore set out to establish what had been accomplished, whether the stated goals had been met, what problems had prevented the plan from being carried out, and what would happen to the project next.
Summary
The government’s experimental project on comprehensive restoration, which ran from 2023 to 2025, has formally concluded but did not achieve its main goal: none of the selected localities was fully restored within the set time. Within the experiment, 80 facilities were completed — slightly more than a tenth of the current restoration need. The experiment thus laid the groundwork for continuing the work, but it did not deliver comprehensive restoration.
The decisive factor behind the delays was unpredictable and untimely financing. Over two and a half years, UAH 3.99 billion was directed to the project — roughly a third of its estimated value — and 73% of that money fell in the final year of the experiment.
The situation was compounded by planning shortcomings, repeated changes to the lists of facilities, revisions to project documentation, problems transferring contracting-authority functions from private owners, the regional Restoration Services’ limited experience in building housing and social facilities, and security risks. Against this backdrop, contracts were terminated and extended more often, and average competition in open bidding fell over the course of the experiment from 3.6 to 1.5 bidders.
After the experiment ended, the unfinished work was moved into the format of a public investment project running for four years, covering five localities and 591 facilities. This mechanism has its advantages: a longer implementation horizon, the possibility of medium-term budget planning, and the reflection of projects in the DREAM system.
The experiment showed that comprehensive restoration cannot be reduced to a set of separate construction projects with unpredictable annual financing. Delivering it effectively requires integrated territorial planning, community participation, properly prepared projects, uninterrupted financing, and clear accountability for results. Learning these lessons should become the basis for a more consistent implementation of the comprehensive restoration concept.
From experiment to public investment project
At the start of the experimental project, the list included 295 restoration facilities. More than 90% of them were housing. The rest consisted of social infrastructure, administrative buildings, road infrastructure, and utilities.
By March 2025, the list already numbered 593 facilities — twice as many as at the beginning. This sharp increase was driven by a rise in the number of residential infrastructure facilities to be restored in two localities, Posad-Pokrovske and Borodianka.
Yet as of mid-2025, after two years of the comprehensive restoration experiment, construction and repair works had been fully completed on only 58 facilities — roughly 10% of the total. For the vast majority of the others (64%), work had only begun or picked up pace in late 2024.
At the same time, under the terms of the experiment, 2025 was to be its final year. By then it was already clear that restoration would not be completed on time. According to the Restoration Services, work in Borodianka and Trostianets, at least, was not expected to be finished before 2026. The actual state of the other localities gave no grounds to expect full completion by the end of 2025 either: in Posad-Pokrovske the work involved hundreds of private houses and infrastructure facilities; in Yahidne, unfinished residential and social facilities remained; and in Tsyrkuny the work had not entered an active phase.
In the end, that is exactly what happened: none of the localities selected for the experiment was fully restored by the end of 2025.
In 2026, in response to our inquiry, the Agency for Restoration reported that, following the experiment, work had been fully completed on 80 facilities, while for the rest, project documentation had been developed and approved and construction was under way. Thus, over the two and a half years of the government experiment, the regional Restoration Services and the Kyiv RMA managed to restore roughly 13% of all the planned facilities.
According to the contracting authorities, as of the end of May 2026 the highest level of completion was recorded in Borodianka, where restoration was finished on 33 of 77 facilities (42.9%). In Yahidne, 36 of 119 facilities (30.3%) were restored. In Posad-Pokrovske, which concentrated the largest number of facilities in the experimental project, only 10 of 392 facilities (2.6%) were completed. In Tsyrkuny, as of the reporting date, neither of the two planned facilities had been restored. These figures point to a considerable unevenness in the project’s implementation.
Ultimately, the experiment ran out of time, yet most of the facilities it was meant to restore remained unfinished. As far back as April of last year, the government commissioned an assessment of the experimental project which, according to the head of the Agency for Restoration, was to involve international experts and produce its findings in the autumn. However, the Agency did not respond to our inquiry about whether such an assessment had been carried out and what it found. Nor is there any information about it in open sources — which gives reason to doubt that the assessment ever took place.
The first official audit of the experiment was recently carried out by the Accounting Chamber. It found that the project, which was meant to become a model of “build back better,” had not achieved its goals — it was implemented without an adequate strategic, organizational, and financial foundation, amid inconsistent regulatory rules and a mismatch between the powers and the responsibilities of the coordinators.
The experimental project’s conclusion, without meeting its main goal, cast doubt on the situation of the facilities whose restoration had begun but was not finished. This looked especially critical given that a significant share of those facilities involve the construction or repair of housing, whose restoration directly affects residents’ ability to return to normal life.
Despite this, the state did not abandon the unfinished facilities. To avoid a complete halt to the work after the experiment ended, a decision was taken last autumn to continue the comprehensive restoration project — now within a corresponding public investment project. It covers five localities that took part in the experiment and 591 restoration facilities in them.
This means the experiment was not concluded in any substantive sense but reformatted. The facilities that could not be finished within the experiment partly moved into a new financing and management mechanism. The difference is that they are now to be implemented not as a temporary experiment with rather ambitious deadlines, but as a public investment project with a longer horizon. The program’s implementation period is set at 48 months, which looks considerably more realistic.
The transition to a public investment project has several potential advantages:
It corresponds better to the real scale of the work. The comprehensive restoration of a locality is a multi-year process involving design, expert reviews, procurement, construction, technical supervision, changes to project documentation, and the commissioning of facilities.
This format should be more compatible with medium-term budget planning.
All public investment programs and their components, as well as public investment projects, are reflected in the DREAM system. This makes it possible to trace every stage of their implementation: from the technical solution and contracting processes to the option of viewing photographs of a facility during restoration or after its completion. For a project that, in the experiment format, often suffered from a lack of timely and complete public information, this is undoubtedly a step forward.
That said, the shift to a public investment project format does not in itself guarantee that the work will be finished. It merely creates a new management framework for the facilities that could not be completed within the experiment. How effective that framework proves will depend, above all, on the volume and timeliness of financing, which is discussed in the next section.
Funding
Over the course of the experimental project, more than 86% of all actual expenditures (UAH 3.44 billion) went to the restoration of Borodianka and Posad-Pokrovske. The highest ratio of needs met by funds actually spent was recorded in Yahidne, where more than half (50.2%) of the estimated restoration cost was disbursed. By contrast, in Tsyrkuny only UAH 6 million was spent — roughly 1.6% of the projected financing need. In Borodianka and Posad-Pokrovske, despite the largest volumes of actual expenditure, only 37.5% and 24.2%, respectively, of the estimated restoration cost was financed.
In all, over 2023–2025 the state agreed to allocate UAH 9.06 billion to the comprehensive restoration of the selected localities — 72% of the estimated cost. But for two years running, the problems with the actual disbursement of this financing made it almost impossible to implement the project on time and efficiently.
As of the end of 2025, UAH 3.99 billion had been directed to the project — roughly 32% of the estimated cost of the entire comprehensive restoration effort. The lion’s share of these funds — UAH 2.93 billion, or 73% — was allocated in the experiment’s final year, which helped to achieve at least minimal progress on comprehensive restoration.
Unfortunately, reformatting the experimental project into a public investment program did not substantially change the financing situation — neither in the volume of funds nor in the timeliness of their release. For continued work on the restoration facilities in 2026, the government earmarked UAH 1.94 billion, but as of the end of May the funds had not been disbursed.
This once again led to delays in paying for works and services. In particular, the Restoration Service in Kyiv region lacks the financing to pay for 88 contracts, including for construction, installation, and design works, as well as technical and designer supervision. A similar situation has developed in the Chernihiv region. And in the Sumy region, the lack of funds became grounds for terminating a number of contracts.
According to the Agency, as of August the funds had in fact been allocated and delivered to the spending units. The estimated additional need to complete all the facilities in the coming budget periods amounts to UAH 4.92 billion.
Procurement
Between June 2025 and June 2026, the authorities contracting restoration works and services announced 301 tenders with a total expected value of UAH 2.08 billion. Of these, 295 were completed and 3 failed. The tenders resulted in contracts worth UAH 2.01 billion.
The most notable change is in the mix of procurement methods. In the previous period, almost the entire estimated value fell on special open bidding: roughly 97% of the estimated value. Direct contracts then accounted for only around 3% of the estimated value, even though there were more of them by count.
In the new period, the picture changed: the value of direct contracts came close to the value of competitive procurement. Of the 301 lots, 269 — 89% by count — were conducted without using the electronic system. They accounted for UAH 949.5 million, or 45.5% of the total estimated value. There were only 32 instances of special open bidding, or 10.6% of the lots, but they made up UAH 1.136 billion of the estimated value — 54.4% of the total.
It is important to note that 93% of direct procurement by value consists of additional works from contractors with whom a contract had already been signed.
Contracting authorities are permitted to buy additional works directly from the same contractor for amounts not exceeding 50% of the main contract. Such works are often procured, for instance, because of revisions to project documentation. And such revisions are fairly common for construction that drags on — a problem for the facilities in the experimental project, in particular because of interruptions in financing.
By number of tenders, the most active was the Regional Development Department of the Kyiv RMA — 143 lots with an estimated value of UAH 1.46 billion. Together with the Restoration Service in Kyiv region, the department signed more than 200 contracts with contractors to restore the infrastructure of Borodianka, worth UAH 1.62 billion.
Among the other localities, the largest estimated value fell to the Restoration Service in Chernihiv region — UAH 257.6 million across 38 lots. These tenders resulted in contracts worth UAH 250.2 million. It was the Chernihiv service that was responsible for restoring Yahidne, where in 2026 the main focus shifted from housing repairs to social infrastructure, individual residential buildings, and the street and road network. This also explains why, in terms of expenditure structure, the largest share of contracts in the new period fell on social infrastructure.
The Restoration Service in Sumy region had the smallest volume of procurement: 14 lots with an estimated value of UAH 31.7 million, resulting in contracts worth UAH 26.0 million. These relate to Trostianets, where after June 2025 the situation became mixed: some facilities were completed or commissioned, some required revisions to project documentation, and for certain contracts problems arose because of a lack of funds.
By type of work, the structure of the contracts did not change compared with the previous period. The bulk of the contracted funds still goes to construction works — UAH 1.95 billion, or 96.7% of the total contract value.
The largest share of funds went to apartment buildings — UAH 1.14 billion, or 56% of the total contract value. The second-largest category was social infrastructure: UAH 697 million, or 34%. Restoration of streets, roads, and utilities accounted for UAH 160.8 million, and private houses for UAH 14.4 million.
Competition in open bidding with features weakened in the final phase of the experiment. On average, 1.5 bidders submitted proposals for competitive lots. That is fewer than in the previous period, when the average was 2.25 bidders, and considerably fewer than in the experiment’s first year, when competition was higher still (3.6 bidders). The best figure in the new period belonged to the Sumy service — an average of 3 bidders per lot. The Chernihiv service averaged 2 bidders. The lowest competition was at the Kyiv service — 0.71 bidders per lot, partly explained by three open tenders canceled for lack of proposals.
The low competition was also reflected in nominal savings — the gap between the estimated value and the contract price. On average, this gap on completed competitive lots came to 2.37%. In the previous period the figure was 7.5%, and in the experiment’s first year it was 21%. The highest nominal savings again belonged to the Sumy service — 31.32% — but this result was produced by a single competitive lot with three bidders and so cannot be considered representative of the procurement statistics as a whole. At the Kyiv service, nominal savings were 3.88%, and at the Chernihiv service, 2.65%.
Unlike the previous period, when potential overpayments on materials were identified in the Borodianka and Posad-Pokrovske tenders, the new period showed no signs of inflated prices or breaches of the law in procurement.
In sum, procurement within the experiment between June 2025 and May–June 2026 reflects not so much a major new stage of reconstruction as an attempt to finish, adjust, and follow through on what could not be completed earlier.
Earlier procurement
Besides the recent tenders, we also re-examined the procurement of the project’s first two years. As of July 2026, 9.47% of the contracts resulting from those tenders had been terminated — 55 of 581 successfully completed tenders. Across the system as a whole, for the CPV codes under which these tenders were conducted, terminated contracts accounted for 3.66% over the same period. In other words, within the project, contracts are terminated more than two and a half times as often as in comparable ordinary procurement. Moreover, the agreements terminated were not only those for construction itself, but also those for developing and revising design and estimate documentation and for designer supervision.
Contracts within the project were also amended fairly often: amendments were signed for 62% of the tenders in the project’s first two years. For 42.9% of the lots, the term of the contract was extended. That is roughly twice as often as the average for comparable procurement in Prozorro over the same period — 19.8%.
Borodianka
Borodianka remains the most visible example of restoration within the government experiment. It is here that the most public results have appeared over the past year: restored apartment buildings, housing repairs on the central streets, landscaping of the areas around buildings, and an effort to present the locality as a model of comprehensive reconstruction. Even so, by the end of the experiment Borodianka had not been fully restored.
As of the end of May 2026, 33 of the 77 facilities requiring restoration had been completed here — almost 43%. The best progress was shown by the part of the project for which the Kyiv Regional Military Administration acted as the contracting authority: by its account, as of the end of 2025, 33 of the locality’s infrastructure facilities had been completed, 31 of them apartment buildings. Work was continuing on 17 facilities.
The part of Borodianka for which the Restoration Service in Kyiv region is responsible progressed considerably more slowly. The service reported that construction works were under way on 23 facilities, but that as of the date of its reply no construction had been completed. In addition, the question of payment for works and services remained unresolved for dozens of contracts.
Trostianets
In Trostianets, the list of restoration facilities remained short but consisted of large infrastructure projects: the reconstruction of the railway station and the station square, the restoration of 40th Army Square, and the repair of sections of the H-12 Sumy–Poltava road. For that reason, progress in the town should be judged not only by the number of facilities but also by their complexity.
As of the end of May 2026, 1 of the 4 facilities on the list could be considered fully completed. There are, however, individual results in other areas too: one section of the H-12 road has been commissioned, the first stage of work on 40th Army Square has been finished, and on another section of the H-12 work is continuing and requires revisions to the design and estimate documentation.
The most ambiguous situation is that of the railway station. The Restoration Service in Sumy region reported that procurement and preparatory procedures had been carried out for this facility, but that in December 2025 the delegated contracting-authority functions were returned to JSC Ukrzaliznytsia. That is, the facility left the track along which the service was responsible for finishing it.
Trostianets therefore cannot be called a fully stalled case, but neither did it become a completed one. Some infrastructure results are in place, yet the overall project split into several different trajectories: some things were completed, some were handed over to another contracting authority, and some were carried over to the next stage of restoration.
Yahidne
Yahidne remains one of the few localities where the experiment produced tangible physical progress. Whereas earlier the main attention was on repairing housing and the street and road network, over the past year the focus also shifted to social infrastructure.
As of the end of May 2026, 36 of 119 facilities had been restored in Yahidne — 30.3%. According to the Restoration Service in Chernihiv region, procurement of construction works had been announced for 17 facilities, work was under way on 12, and construction had been completed on 5. Separately, the Agency for Restoration reported on work on social infrastructure, in particular buildings associated with public services and with the memorialization of events of the occupation.
However, Yahidne did not escape the experiment’s typical problems. Some private residential facilities still depend on the transfer of construction contracting-authority functions. According to the service, such functions had not been transferred to it for 19 houses, and two more contracts were terminated at the initiative of private house owners.
So Yahidne looks like one of the more advanced cases of the experiment, but not a completed one. There is already a visible result here — some housing and infrastructure has been restored — but a number of facilities have moved on to the next stage, and completing the work still depends on resolving organizational and resource questions.
Tsyrkuny
Tsyrkuny remains the most problematic locality in the pilot project. It is a frontline village in Kharkiv region that continues to be affected by security risks. It was the security situation, above all, that came to explain why restoration here never moved into a full construction phase.
By the end of the experiment, neither of the two facilities still on the current list had been completed in Tsyrkuny. The Restoration Service in Kharkiv region reported that procurement of services for developing project documentation had been announced, and that the documentation had been developed and passed expert review. However, procurement of construction works was not announced, and no construction, reconstruction, or major repairs were started.
The Tsyrkuny military administration, meanwhile, reported a broader list of four facilities for which contracting-authority functions were being transferred to the service: a lyceum, a hospital, the community administrative building housing an Administrative Services Center, and a preschool. Project documentation was prepared for them and positive expert opinions obtained, but beyond the preparatory stage the project did not advance.
Tsyrkuny is thus an example of how an experiment can remain almost entirely on paper. Formally, the village is within the framework of continued restoration, but no financing is provided for it in the first year of the new stage. As a result, Tsyrkuny starts the next stage from the worst position of all: with no completed facilities and no construction works begun.
Posad-Pokrovske
Posad-Pokrovske remains the largest case of experimental restoration by scale. It is here that the greatest number of facilities is concentrated — above all private housing, engineering networks, street and road infrastructure, and social facilities. Because of this, the village simultaneously shows considerable construction activity and a very low completion rate.
As of the end of May 2026, only 10 of the 391 facilities requiring restoration had been completed in Posad-Pokrovske — that is, 2.6%. This does not mean, however, that no work is being done. According to the Restoration Service in Kherson region, construction was under way on 328 facilities, among them new construction of private houses, a Safety Center and an outpatient clinic, major repairs of private housing, street and road facilities, and the reconstruction of water supply and wastewater systems.
In the autumn of 2025, the Ministry for Development of Communities and Territories also reported that roughly 300 facilities in the village were in progress, with hundreds of workers and dozens of pieces of equipment engaged in the construction. Posad-Pokrovske has thus genuinely moved from the preparatory stage to a mass construction phase. But because of the scale of the restoration, the number of completed facilities so far remains disproportionately small.
Conclusions
The pilot project has formally concluded, yet its main goal — the comprehensive restoration of the selected localities over 2023–2025 — was not achieved. None of them was fully restored within the set time. Moshchun dropped out of the project at the very start of implementation; in Tsyrkuny construction works never began; and in the other localities a significant share of facilities remained at various stages of readiness. The experiment completed only 80 facilities — slightly more than a tenth of the current restoration need.
However, the experiment cannot be called wholly fruitless. Partial results were achieved in Borodianka and Yahidne, where some housing and social and communal infrastructure was restored. In Posad-Pokrovske, the power supply system was reconstructed, road repairs were completed, and work was launched on hundreds of residential and infrastructure facilities. In Trostianets, project documentation was developed, procurement was carried out, and construction works began. In this way the experiment laid the groundwork for continued restoration, even though within the project itself this work was never seen through to completion.
Yet in terms of delivering a coherent transformation of these territories, the goals set remained unmet. Comprehensive restoration was reduced, in large part, to the parallel implementation of a great many separate projects. The terms of the experiment did not require comprehensive restoration plans to be drawn up, did not establish common criteria for selecting localities and facilities, and did not define how the “build back better” principle was to be put into practice. As a result, the composition and scale of the work varied widely: from hundreds of residential and infrastructure facilities in Posad-Pokrovske to two facilities in Tsyrkuny and a complete absence of projects in Moshchun.
The decisive factor behind the delays was unpredictable financing. The government took decisions to allocate funds in 2023 and 2024 only in August, and the budget appropriations were opened later still. Because of this, contracting authorities could not carry out procurement on time, pay for completed work, or plan construction for the following periods. Some work was halted, contracts were terminated, and the active construction phase for most facilities began only in late 2024 or in 2025. In all, over 2023–2025 the project actually spent UAH 3.99 billion — roughly a third of its estimated value — with almost three quarters of that money falling in the experiment’s final year.
The financing delays compounded weaknesses that had been built in during the project’s planning. Restoration of some private housing was blocked because owners did not transfer contracting-authority functions or did not admit contractors to the sites. Repeated changes to the lists, the poor quality of the initial project documentation, the discovery of additional works, and price fluctuations all led to numerous project revisions. Within the project, contracts were terminated more than two and a half times as often as in comparable ordinary procurement, and their terms were extended roughly twice as often. The regional Restoration Services, which until then had worked mainly on road infrastructure, had limited experience building housing and social facilities and were at the same time managing hundreds of projects with different owners and asset holders.
The systemic problems in implementing the experimental project — the financing delays in particular — made taking part in it less and less attractive to contractors with each passing year. This can be traced in the dynamics of competition in open bidding with features: in the experiment’s final phase, an average of 1.5 bidders submitted proposals per competitive lot, whereas in the previous period the figure was 2.25, and in the project’s first year it was 3.6. In other words, over the course of the experiment the average level of competition more than halved.
Implementation was also complicated by shelling, repeated damage to already-restored facilities, air-raid alerts, power outages, labor shortages, and logistics disruptions. In Tsyrkuny, security risks made it impossible to move from design to construction at all. However, these objective circumstances do not account for all of the experiment’s failures: a large part of the delays could have been avoided through better planning, timely funding, and a clearer allocation of responsibility.
After the experiment ended, the state did not abandon the unfinished facilities. Their restoration continued in the format of a public investment project designed to run for the next four years. It included all five localities and their unfinished facilities. A longer implementation horizon, the possibility of medium-term budget planning, and the reflection of projects in the DREAM system potentially create a more realistic and transparent framework than a temporary experiment.
For continued work in 2026, UAH 1.94 billion was earmarked, but the funds were again released with a delay. The estimated additional need to complete all the facilities amounts to a further UAH 4.92 billion. Without predictable long-term financing, then, the new stage risks repeating the experiment’s scenario — with overdue payments, work stoppages, and slipping deadlines.
To avoid this, the government needs not only to allocate funds but also to set a schedule for completing the facilities, prioritizing them together with the communities. Separately, it should publish the promised independent assessment of the experiment — or carry it out, if no such assessment has yet taken place.
The main result of the experiment, then, was not restored localities but the lessons it revealed about the limits of the current reconstruction system. It proved that comprehensive restoration cannot be delivered as a set of separate construction projects with annual, unpredictable funding. What this requires is integrated territorial planning, community participation, projects that are ready to start, uninterrupted financing, and accountability for the final result.
It is now important to take this experience on board and not to repeat the experiment’s mistakes. At the same time, the very idea of comprehensively restoring localities has lost none of its relevance. Rather the opposite: this experiment showed precisely which preconditions were missing for it to be fully completed. The concept of comprehensive restoration, then, is still waiting for a more consistent and effective implementation.
This material is funded by Norway. Its content is the sole responsibility of Transparency International Ukraine and does not necessarily reflect the views of the Norwegian Government.
This research was developed by
Team lead:
Ivan Lakhtionov, Deputy Executive Director of TI Ukraine for Innovative Projects
Author:
Andrii Shvadchak, Legal Advisor at Transparency International Ukraine
