Coordination

Evidence for missing indicators

Each record describes a case the indicators above cannot see. It does not affect the score or confidence because one case is not comparable across countries. A gap can become an indicator when a comparable series covers at least two countries.

• 🇩🇪Germanylinked to University-industry collaborationstarted 1973still operating

  The Fraunhofer contract-research model

  Germany funds the Fraunhofer-Gesellschaft's institutes in proportion to what they earn from industry, and in 2025 the society performed 3.2 billion euros of contract research, 966 million euros of it paid by industry.

  966,000,000 euros, industrial revenue, 2025. 3,200,000,000 euros, contract research volume, 2025. Fraunhofer-Gesellschaft, facts and figures, retrieved 2026-08-27.

  How it worked: The state pays each institute a base grant that grows with the industrial income it wins, so every institute director is structurally forced to sell applied research to firms while the public base keeps the science honest.

  Limits: The figures are the society's own. Revenue measures demand for applied research, not whether knowledge moves between universities and firms, which is what the indicator asks about, and Fraunhofer sits beside the university system rather than inside it, so a reader can argue the model routes around university-industry collaboration as much as it evidences it. Roughly two-thirds of the budget is still public money.

• 🇰🇪Kenyalinked to Public-private collaborationstarted 2007still operating

  M-Pesa, and the regulator that let it run

  Kenya's central bank let a phone company build a national payments system in 2007 and supervised it into ubiquity, and by June 2026 Kenya had 94.2 million registered mobile money accounts and 572,104 agents, in a population of about 55 million.

  94,200,000 accounts, registered mobile money accounts, 2026-06. Central Bank of Kenya, mobile payments statistics, retrieved 2026-08-27.

  How it worked: The central bank issued a letter of no objection instead of demanding a banking licence, set conditions on trust accounts holding customer float, and wrote the e-money regulations after the system worked, legalising what it had learned rather than licensing what it could imagine.

  Limits: The delivery was Safaricom's, a majority-private company, so the record bears on public-private collaboration only through what the state did: permit, then supervise, then regulate interoperability. Registered accounts overstate people, since many Kenyans hold several, and account counts say nothing about fees, which are materially higher than card rails elsewhere. A single dominant operator is also a concentration the indicator's construct does not reward.

• 🇨🇭Switzerlandlinked to Public-private collaborationstarted 2004still operating

  Dual-track vocational training

  Switzerland runs most upper-secondary education through firms: of 218,259 students enrolled in vocational education and training in 2020, 90.6 percent, 197,782, were apprentices trained inside companies under federal law.

  197,782 apprentices, apprentices in dual-track training, 2020. 90.6 % of VET students, share of vet students in the dual track, 2020. Cedefop, European Centre for the Development of Vocational Training, dual-track VET scheme fiche, retrieved 2026-08-27.

  How it worked: Federal law makes firms, cantons and professional associations joint owners of training: associations write the curricula for their own occupations, firms hire and pay apprentices to do real work, and the state examines and certifies, so the labour market signals what to teach.

  Limits: Enrolment counts participation in the system, not what the training is worth; the wage and employment evidence sits in separate studies. The system long predates the number: 2004 is the current federal act, not the start of apprenticeship in Switzerland, and the record cannot show how much of the outcome is the law versus a century of employer habit. Firms train because it pays them net of apprentice output, and that arithmetic, not the statute, is the load-bearing part.

• 🇧🇷Brazillinked to Public-private collaborationstarted 1974still operating

  Itaipu Binacional, the treaty-built power system

  Brazil and Paraguay built and operate Itaipu through a treaty-based binational entity, which reached 14,000 megawatts of installed capacity and has supplied both countries since the first unit entered operation in 1984.

  14,000 MW, installed generating capacity, 2007. Itaipu Binacional, institutional information, retrieved 2026-08-29.

  How it worked: Brazil and Paraguay converted a contested shared river into a jointly governed operating institution with its own legal personality, fixed allocation rules and a long-lived technical system.

  Limits: Installed capacity shows the scale of the physical system, not the quality of the bilateral relationship, the project's cost or the distribution of its benefits. A treaty structure can hold a joint asset together while leaving wider coordination between the countries unresolved.

• 🇧🇷Brazillinked to University-industry collaborationstarted 2013still operating

  EMBRAPII, industry-led applied research

  Created in 2013, EMBRAPII funds industrial research that companies contract with accredited public and private research units, and it contracted a record 811 new projects in 2025, 25.35% more than in 2024.

  811 projects, new projects contracted in the year, 2025. EMBRAPII, Embrapii em 2025: veja como foi nossa atuação (published 3 February 2026), archived copy, retrieved 2026-10-02.

  How it worked: Public funding follows an industry's defined technical problem into a research unit, sharing early risk while leaving the company and researchers responsible for solving a concrete challenge.

  Limits: Project counts show the throughput of a funding network, not the commercial value, technical success or additionality of the research. EMBRAPII is a social organisation under a management contract with four federal ministries, and the figure is its own report of its activity, drawn from its Power BI panel on 3 February 2026. Its 90 accredited units are research institutions of several kinds, university units among them, and the page does not say how many of the 811 projects ran in a university, so the record evidences firm-research-unit collaboration of which university collaboration is a part, not all of it. Projects vary widely in scale and maturity. The live page answered 'Conteúdo temporariamente indisponível' (HTTP 503) on 2 October 2026, so the source is the Internet Archive copy of 22 April 2026, which carries the 811 figure.

• 🇧🇷Brazillinked to University-industry collaborationstarted 2007delivered and closed

  SIBRATEC, the national technology networks

  To foster university-firm interaction, the Federal Government set up SIBRATEC, coordinated by the Ministry of Science and Technology with FINEP, BNDES and INMETRO taking part, as a national system of 56 research and development networks by 2010: 14 innovation-centre networks, 20 technological-service networks and 22 state-level extension networks.

  56 networks, sibratec national r&d networks reported, 2010. Ministry of Science and Technology, Plano de Ação em Ciência, Tecnologia e Inovação 2007-2010: principais resultados e avanços (Programa 5.1, SIBRATEC), archived copy, retrieved 2026-10-02.

  How it worked: The ministry linked research groups, technology centres and state-level extension networks so firms could reach technical support outside universities and turn a production problem into a shared R&D project.

  Limits: The network count records an institutional architecture, not how many firms worked with a university, how much research was done jointly or what it produced. The report states the purpose (university-firm interaction) and the network count, but the networks also include technology institutes and metrology laboratories, and the extension networks support small firms with technical services rather than joint research. It is the ministry's own account of its plan and does not evaluate outcomes. The started year is the first year of the 2007-2010 action plan under which SIBRATEC was implemented; the report gives no creation date. The report says nothing about the networks after 2010, so the concluded status records the end of this delivery period, not the programme's fate. The ministry repository did not answer on 2 October 2026, so the source is the Internet Archive copy of 29 April 2025; the passage is on page 74.

• 🇰🇷South Korealinked to Public-private collaborationstarted 1961delivered and closed

  Economic Planning Board, long-run industrial coordination

  South Korea's Economic Planning Board operated from 1961 to 1994 as the government body responsible for economic development planning and budget planning before its functions moved through later finance ministries.

  1961 year, year the economic planning board was established, 1961. Ministry of Strategy and Finance, ministry history, retrieved 2026-08-30.

  How it worked: Planning, budgeting and economic coordination were placed in one central institution, giving government a durable route to align public investment with industrial priorities and private-sector expansion.

  Limits: The board's mandate and lifespan show a coordination architecture, not the independent contribution of planning to industrial outcomes or the distribution of its gains. A ministry history is an official account and is not a comparative evaluation.

• 🇺🇾Uruguaylinked to Public-private collaborationstarted 2006still operating

  SNIG, universal individual cattle traceability

  Uruguay required producers to register every calf in a national traceability system from 2006, and by October 2011 the registry covered 100 percent of the herd, 11.75 million live cattle, while identification tags reached more than 45,000 producers each year.

  11,750,000 cattle, live cattle registered and individually identified, 2011-10. 45,000 producers, producers receiving identification tags each year, 2011. Presidency of Uruguay, SNIG traceability report, retrieved 2026-08-30.

  How it worked: The agriculture ministry made lifetime identification compulsory, issued paired visual and electronic tags, reused the existing producer register and enlisted the postal service to deliver tags to farms, while producers and processors recorded every movement through one national system.

  Limits: A complete administrative register does not prove that every movement is reported correctly, that inspectors can reconcile cattle with the database, or that the commercial benefit exceeds the public and producer cost. The source is a government report from 2011, and the collaboration rests on a legal obligation rather than voluntary agreement, so it evidences coordinated national implementation more strongly than mutual public-private agenda-setting.

• 🇳🇱Netherlandslinked to Public-private collaborationstarted 1950still operating

  Social and Economic Council, the institutionalised polder bargain

  The Netherlands established the Social and Economic Council (SER) by law in 1950, giving employers, trade unions and independent Crown-appointed experts a standing forum for social and economic policy; the current council has 36 members and collective agreements cover more than 84 percent of employees.

  36 members, ser council members, 2026. 84 % of employees, employees covered by collective agreements, 2026. Social and Economic Council of the Netherlands, frequently asked questions, retrieved 2026-08-31.

  How it worked: The state gave organised employers and workers a formal route to negotiate national trade-offs, added independent experts as bridge-builders and required the Cabinet to explain when it rejects a broadly supported recommendation, turning social conflict into a repeated bargaining process.

  Limits: A tripartite membership structure and collective-agreement coverage show the reach of a consultation system, not whether its compromises are equitable, whether advice changes policy or whether workers outside agreements benefit. The SER describes its own role and the coverage figure; it is not an independent evaluation of the polder model's effects.

• 🇳🇱Netherlandslinked to Public-private collaborationstarted 2023still operating

  Top Sectors and mission-driven innovation policy

  In November 2023 the Netherlands renewed its mission-driven innovation compact: businesses, knowledge institutions, governments and other organisations committed an estimated €5.7 billion a year, coordinated through five missions and eight Knowledge and Innovation Agendas.

  5,700,000,000 euros per year, annual joint mission-driven innovation effort, 2023. 30 organisations, partners signing the innovation covenant, 2023. Ministry of Economic Affairs, signing of the Knowledge and Innovation Covenant 2024-2027, retrieved 2026-08-31.

  How it worked: Government gave industry, research organisations and public agencies a shared mission vocabulary, placed them in sector teams and tied public instruments to jointly written agendas so collaboration could shape research priorities rather than appear only at the procurement stage.

  Limits: The €5.7 billion is an estimated annual commitment, not audited additional spending, and the partner count measures signatures rather than durable collaboration or successful products. The ministry's announcement does not provide a counterfactual against which the policy's innovation effects can be judged. Funding agendas can coordinate priorities without changing how projects are selected in practice.

• 🇳🇱Netherlandslinked to Public-private collaborationstarted 2005still operating

  iDEAL, interoperable online payments through Dutch banks

  The Netherlands built iDEAL as a shared bank-connected online payment rail; in 2024 it processed 1.47 billion payments worth €141 billion, almost matching the €147 billion spent through debit cards at the point of sale.

  1,470,000,000 payments, ideal payments, 2024. 141,000,000,000 euros, ideal turnover, 2024. Dutch Payments Association, payment traffic factsheet 2024, retrieved 2026-08-31.

  How it worked: Banks and the scheme operator gave merchants one recognisable checkout method while allowing each bank to keep its own customer interface; the shared protocol made account-to-account payment immediate and gave merchants a common confirmation path.

  Limits: Transaction volume and turnover show adoption of one payment rail, not fraud rates, consumer welfare, merchant costs or whether iDEAL caused the growth of Dutch e-commerce. The Dutch Payments Association aggregates industry data, and the figures do not separate domestic from cross-border use. A successful network also benefits from incumbency and high bank-account ownership.

• 🇳🇱Netherlandslinked to Public-private collaborationstarted 2009still operating

  eHerkenning, a public-private business identity framework

  The Netherlands developed eHerkenning with commercial identity providers so businesses could use one trusted login across public services; in 2024 the framework recorded 1,140,751 authentication means and 639 connected organisations.

  1,140,751 means, eherkenning authentication means, 2024. 639 organisations, connected organisations, 2024. Logius, 2024 service figures, retrieved 2026-08-31.

  How it worked: Government set the assurance rules and service-provider obligations while private identity providers competed to issue credentials, giving businesses a choice of provider without asking every agency to build its own business identity system.

  Limits: Credential and organisation counts show the reach of a trust framework, not whether small firms can afford a credential, whether logins reduce administrative burden or how identity providers distribute risk. The service statistics are operator-reported and do not reveal failed attempts, fraud or the cost of compliance at each assurance level.

• 🇳🇱Netherlandslinked to Public-private collaborationstarted 2004still operating

  Digipoort, an electronic post office for government–business data

  The Netherlands runs Digipoort as a central gateway that receives, validates and routes secure messages between businesses and public agencies; in 2024 it handled 75,433,353 messages, including a channel with 3,075 connected organisations for financial and social reporting.

  75,433,353 messages, messages handled by digipoort, 2024. 3,075 organisations, connected organisations in sbr financial & social, 2024. Logius, 2024 service figures, retrieved 2026-08-31.

  How it worked: The state provided one secure gateway and common message standards so software suppliers, accountants and firms could submit reports to several agencies without negotiating a new transport and validation layer for each one.

  Limits: Message volume and connected-organisation counts show use of a shared gateway, not whether submissions are correct, cheaper or less burdensome for firms. The 2024 total combines different message streams with different users and service levels, and operator-reported availability does not measure the quality of the underlying tax, payroll or regulatory decisions.

• 🇨🇱Chilelinked to Public-private collaborationstarted 2003still operating

  ChileCompra, a national public-procurement marketplace

  Chile runs Mercado Público as a shared e-procurement platform; in 2024 it transacted US$17.643 billion through 2,031,670 purchase orders involving 110,255 participating suppliers.

  17,643,000,000 US dollars, mercado público transaction value, 2024. 2,031,670 orders, mercado público purchase orders, 2024. ChileCompra, 2024 Mercado Público transactions report, retrieved 2026-08-31.

  How it worked: ChileCompra set common tender rules, a digital marketplace and open transaction records so public agencies could buy from a broad supplier base through one visible channel rather than each agency maintaining a separate procurement system.

  Limits: The report's totals describe purchase-order commitments rather than payments, and they are operator-reported. They show the reach of a common market, not whether procurement is cheaper, more competitive, more innovative or less corrupt, nor whether small suppliers receive an equitable share. The supplier classification is based on administrative data and does not establish the quality or delivery of the contracts awarded.

• 🇨🇱Chilelinked to Public-private collaborationstarted 2010still operating

  Start-Up Chile, a public accelerator for global startups

  Since 2010, Start-Up Chile has supported more than 2,500 startups from Chile and abroad through an equity-free public accelerator that connects founders with investors, mentors and corporates.

  2,500 startups supported, startups supported by start-up chile, 2026. Start-Up Chile (CORFO), portfolio, retrieved 2026-08-31.

  How it worked: CORFO supplied equity-free early funding, a landing platform and a founder network while private investors, mentors and corporates supplied market knowledge, follow-on capital and commercial links. The state used a public accelerator to widen the pipeline rather than trying to pick every winning company itself.

  Limits: The programme's portfolio count is self-reported and records support, not startup survival, additionality or the share of value retained in Chile. It does not show whether public funding displaced private investment, whether benefits reached domestic firms or how outcomes compare with other accelerators. The equity-free model and network describe an intervention, not proof of a causal economic return.

• 🇨🇴Colombialinked to Public-private collaborationstarted 2015still operating

  SECOP II, a transaction platform for public procurement

  Since March 2015, Colombia Compra Eficiente has used SECOP II to move public contracting from publication to online transactions; in 2024 the platform managed COP117.642 trillion, 84.4% of the value managed across the three SECOP platforms.

  117,642,419,358,639 Colombian pesos, value managed through secop ii, 2024. 84.4 % of value managed, secop ii share of value managed across secop platforms, 2024. Colombia Compra Eficiente, 2024 institutional management report, retrieved 2026-08-31.

  How it worked: The procurement agency supplied common transaction rules and a shared digital workflow so public buyers and private suppliers could publish, bid, award and manage contracts in one auditable space rather than exchanging documents through disconnected systems.

  Limits: The managed-value totals show the reach of a transaction platform, not whether procurement is cheaper, more competitive, more innovative or less corrupt. Colombia Compra Eficiente reports the figures and combines different contract types and platforms; value managed is not the same as payments or delivered outputs. Open contracting data still require supplier capacity, audit and citizen scrutiny to change outcomes.

• 🇵🇪Perulinked to Public-private collaborationstarted 2009still operating

  Obras por Impuestos, firms financing public infrastructure

  Peru's Obras por Impuestos scheme lets private firms finance public projects against future tax liabilities; the finance ministry recorded S/5.504 billion in certificates issued from 2009 through January 2024 and six projects identified as physically and financially complete in 2024.

  5,504,000,000 Peruvian soles, certificates issued through obras por impuestos, 2024-01. 6 projects, projects identified as complete in 2024, 2024. Ministry of Economy and Finance, Obras por Impuestos detailed report, retrieved 2026-08-31.

  How it worked: Peru let companies propose and finance public works and recover the cost through tax certificates, creating a bridge between private capital and local public priorities while the state retained project approval and service ownership.

  Limits: Certificates and completion counts show the scale of a tax-linked financing mechanism, not whether projects were additional, delivered on time, well maintained or distributed fairly. The ministry compiles project information from public systems and companies; the scheme shifts financing and execution roles but does not remove public responsibility for selecting, regulating and operating the asset.

• 🇨🇷Costa Ricalinked to Public-private collaborationstarted 1997still operating

  FONAFIFO Payment for Environmental Services

  Costa Rica's FONAFIFO Payment for Environmental Services programme paid contracts covering 208,089 hectares in 2024, disbursing ₡9.234 billion while using public, donor and market-linked finance to keep private land in forest uses.

  208,089 hectares, hectares under payment for environmental services contracts, 2024. 9,234,000,000 Costa Rican colones, payments disbursed through the programme, 2024. FONAFIFO, Board meeting record and 2024 programme accounts, retrieved 2026-08-31.

  How it worked: Costa Rica created a dedicated forest-finance institution that collects earmarked public revenue and combines it with donor and market funds, then pays landowners under monitored contracts for water, carbon and biodiversity services.

  Limits: Contract hectares and payments show the reach of a conservation-finance mechanism, not whether forest cover is additional, permanent or evenly distributed. The board record is an administrative account from the programme operator; it does not independently verify carbon permanence, biodiversity outcomes or the counterfactual land use.

• 🇪🇨Ecuadorlinked to Public-private collaborationstarted 2012still operating

  SRI electronic invoicing at economy-wide scale

  Ecuador's tax authority received approximately 3.299 billion electronic receipts from January through December 2024, using a legally mandated, digitally signed rail that connects private taxpayers to the public tax system.

  3,299,680,838 receipts, electronic receipts received by sri, 2024. Servicio de Rentas Internas, 2024 accountability report, retrieved 2026-08-31.

  How it worked: Ecuador made electronic invoicing a common legal format, supplied a public authorization and validation service, and required private firms to send signed transaction data to the SRI so tax administration and business records shared one machine-readable event.

  Limits: Receipt volume measures use of the tax rail, not the accuracy of declarations, additional revenue, compliance costs or service quality. The SRI report notes that totals can change through later cancellations and transmissions, and adoption is partly enforced by regulation rather than voluntary collaboration.

• 🇪🇨Ecuadorlinked to Public-private collaborationstarted 2007still operating

  Crédito de Desarrollo Humano for beneficiary-led businesses

  Ecuador's Ministry of Economic and Social Inclusion reported delivering 44,622 Crédito de Desarrollo Humano loans in 2024, including 6,622 to Indigenous people, turning a social-transfer registry into small-enterprise finance and market-access support.

  44,622 credits, development credits delivered, 2024. 6,622 credits, development credits delivered to indigenous people, 2024. Ministerio de Inclusión Económica y Social, 2024 accountability report, retrieved 2026-08-31.

  How it worked: Ecuador let eligible transfer recipients convert part of a social benefit into a productive credit, then paired public finance with group or individual ventures and a government market-access service rather than leaving a cash transfer disconnected from enterprise support.

  Limits: Credit counts show administrative disbursement, not repayment, business survival, income gains or whether the projects are productive. The ministry's report combines several beneficiary groups and credit modalities, so the total is not a comparable measure of private investment or a causal poverty effect.

• 🇵🇾Paraguaylinked to Public-private collaborationstarted 2017still operating

  Ruta PY02 public-private road concession

  Paraguay's public-works balance records the Ruta PY02 duplication and expansion as a 149.5-kilometre public-private partnership, awarded under the 2013 APP law with a USD 507 million construction contract and a concession running to 2046.

  149.5 kilometres, road length under the ruta py02 app contract, 2024. 507,000,000 US dollars, construction contract value, 2017 contract. Paraguay Ministry of Public Works and Communications, 2024 public-management balance, retrieved 2026-08-31.

  How it worked: Paraguay used an APP concession to bundle design, financing, construction, maintenance and operation of a national corridor, moving some delivery and lifecycle obligations to a private special-purpose company while the ministry retained contract oversight.

  Limits: Contract length and value show the scale and financing model, not whether the road met its promised schedule, lifecycle cost, safety standard or service-level payments. A government project balance can confirm the contract and progress but is not an independent value-for-money review, and construction and operation continue over the concession term.

• 🇦🇷Argentinalinked to Public-private collaborationstarted 2012still operating

  Vaca Muerta unconventional energy build-out

  Argentina's energy ministry reported that Vaca Muerta supplied 54.9% of national oil output and 50.1% of national gas output in December 2024, within a year in which national oil production reached its highest level since 2003 and gas its highest since 2006.

  54.9 %, vaca muerta share of national oil production, 2024-12. 50.1 %, vaca muerta share of national gas production, 2024-12. Argentina Secretariat of Energy, 2024 hydrocarbon production report, retrieved 2026-08-31.

  How it worked: Argentina combined state control of the resource and export rules with private drilling, midstream investment and a price framework intended to make unconventional wells financeable, turning a large geological resource into a growing national production base.

  Limits: Production shares show geological and industrial output, not the net fiscal return, environmental effects, local infrastructure strain or the durability of investment under changing prices and regulation. The government account attributes the rise to regulation and private investment, but this record does not isolate either contribution or establish the field's full lifecycle economics.

• 🇮🇩Indonesialinked to Public-private collaborationstarted 2015still operating

  Palapa Ring national fibre network

  Indonesia delivered the three-package Palapa Ring fibre network to connect 57 kabupaten/kota; the communications ministry reported average utilisation of 54.35% in 2024, with the western, central and eastern packages carrying different loads.

  54.35 % of network capacity, average palapa ring utilisation, 2024. 57 districts and cities, kabupaten/kota connected by palapa ring, 2024. Ministry of Communication and Digital, Strategic Plan 2025-2029, retrieved 2026-08-31.

  How it worked: The Indonesian state split a national backbone into regional packages and used operating partners to extend public connectivity into places where commercial demand alone was too thin, then tracked capacity and utilisation by package.

  Limits: Network utilisation and the number of connected districts show a delivered platform and the demand placed on it, not the procurement cost, construction schedule, service reliability or whether local users can afford the resulting bandwidth. The utilisation figure is a ministry planning statistic, and its average hides large differences between the western, central and eastern packages.

• 🇲🇾Malaysialinked to Public-private collaborationstarted 2020still operating

  JENDELA national digital-network plan

  Malaysia's JENDELA Phase 1 connected 7.74 million premises with fibre by 2022, above its 7.5 million target, and raised 4G coverage in populated areas to 96.92% through a regulator-led programme delivered with telecommunications providers.

  7,740,000 premises, premises passed by jendela fibre, 2022. 96.92 % of populated areas, 4g coverage in populated areas after jendela phase 1, 2022. JENDELA, Phase 1 concluding report, retrieved 2026-08-31.

  How it worked: Malaysia's communications regulator set common coverage and fibre targets, coordinated multiple providers and retired 3G spectrum to concentrate investment on a shared 4G foundation before expanding 5G.

  Limits: Premises passed and populated-area coverage measure network availability, not affordability, uptime, actual household subscriptions or rural service quality. The programme report combines regulator and industry-reported delivery and its coverage denominator excludes unpopulated areas, so it is not a national land-area measure.

• 🇹🇭Thailandlinked to Public-private collaborationstarted 2016still operating

  PromptPay national instant-payment rail

  Thailand's central bank and commercial-bank network have run PromptPay as a shared instant-payment rail since 2016; by December 2024, 87.5 million proxy numbers were registered and average daily volume reached 66.4 million transactions, with a one-day peak of 89.2 million in September.

  87,500,000 registered numbers, promptpay registered proxy numbers, 2024-12. 66,400,000 transactions per day, average daily promptpay transactions, 2024. Bank of Thailand, Annual Report 2024, retrieved 2026-08-31.

  How it worked: The Bank of Thailand set a common proxy-ID and QR standard, while banks connected their existing accounts and customer channels to a central registration and switching layer. That let a small merchant receive an interbank payment with a phone number or citizen ID instead of negotiating a bespoke integration with every bank.

  Limits: Registration and transaction volume show a heavily used payment rail, not the distribution of access, reliability during outages, fraud losses or the share of activity that is genuinely additional to cash and cards. The Bank of Thailand reports the aggregate totals, and proxy registrations can include more than one number for a person or business. The public-private mechanism is inferred from the central-bank standard and participating-bank network rather than isolated by the report as a causal estimate.

• 🇪🇸Spainlinked to Public-private collaborationstarted 2021still operating

  Kit Digital SME digitalisation programme

  Spain's Red.es Kit Digital programme used public grants, volunteer representatives and private digitalising agents to help small firms adopt software and online services; by 3 August 2026 the programme reported 1,566,846 transformation projects delivered by 10,727 agents, 96 percent of them ICT SMEs.

  1,566,846 projects, digital transformation projects launched through kit digital, 2026-08-03. 10,727 agents, adhered digitalising agents, 2026-08-03. Red.es, Kit Digital public-private collaboration results, retrieved 2026-08-31.

  How it worked: Red.es standardised eligible solutions and payment rules, while private agents supplied implementation and volunteer representatives handled applications for firms that lacked administrative capacity. The public side made demand legible and portable; the private side supplied local skills and delivery capacity.

  Limits: Projects and agents measure programme throughput, not whether the adopted tools raise productivity, remain in use or narrow the digital divide. The results page is a programme dashboard updated in 2026 and does not provide a counterfactual for firms that did not receive a grant. Projects can be delivered to the same beneficiary in multiple agreements, and agent counts include providers rather than unique firms served.

• 🇵🇱Polandlinked to Public-private collaborationstarted 2018still operating

  Baltic Pipe gas interconnection

  Poland's Gaz-System, Denmark's Energinet and Norway's Gassco coordinated the Baltic Pipe cross-border gas corridor; the interconnection launched on 1 October 2022 and reached its target technical parameters by November, allowing up to 10 billion cubic metres per year toward Poland.

  10,000,000,000 cubic metres per year, baltic pipe capacity toward poland, 2022-11-30. 3 operators, transmission-system operators coordinating the interconnection, 2022. Polish Energy Regulatory Office, National Report of the President of URE 2023, retrieved 2026-08-31.

  How it worked: Three transmission-system operators aligned standards, permits, offshore construction and compressor upgrades around one operating date. Poland's public strategy created the demand and political priority, while the network companies supplied the engineering and cross-border interfaces needed to make the corridor usable.

  Limits: Capacity and commissioning dates show infrastructure delivered, not actual annual flows, gas-price effects, environmental costs or the security value of a single corridor under changing demand. The regulator reports technical parameters and the three-operator arrangement, while the project also depended on EU and national permitting outside this metric. The record describes a cross-border build rather than a purely Polish project.

• 🇸🇪Swedenlinked to Public-private collaborationstarted 2012still operating

  Swish mobile instant payments

  Sweden's six-bank-owned Getswish consortium built Swish into a national mobile payment rail: in 2023 it served just over 8.7 million private users and about 335,000 merchant or business users, processing approximately 977 million payments worth SEK 480 billion.

  977,000,000 payments, swish payments, 2023. 8,700,000 users (over), swish private users, 2023. Sveriges Riksbank, The Swedish Financial Market 2024, retrieved 2026-08-31.

  How it worked: Competing banks jointly owned a simple brand and API, used BankID for authentication and connected the service to the national settlement infrastructure. The consortium solved the coordination problem of getting rival banks to make one person-to-person rail interoperable before merchants and public services built on it.

  Limits: The Riksbank's figures are service totals, not a causal evaluation of the consortium or a measure of consumer welfare. Swish depends on mobile BankID, smartphones and participating banks, so people without those prerequisites can be excluded; high payment volume also says nothing about fraud, outage resilience or market competition.

• 🇮🇪Irelandlinked to Public-private collaborationstarted 2019still operating

  National Broadband Plan rural fibre rollout

  Ireland's National Broadband Plan combined commercial investment with a state intervention contract for rural areas covering more than 560,000 premises; by October 2024 the government reported over 102,600 premises connected to the National Broadband Ireland network.

  102,600 premises (over), state-intervention premises connected, 2024-10. 560,000 premises (over), premises in the state-intervention area, 2020. Government of Ireland, National Broadband Plan benefits and rollout update, retrieved 2026-08-31.

  How it worked: The state defined a universal-service gap, subsidised a network builder in that gap and left commercially viable areas to private operators. A single contract, a published intervention map and connection points for community facilities let public policy steer private construction without requiring the state to own every fibre asset.

  Limits: The connection count is a rollout snapshot, not completion of the national plan or a measure of connection quality, take-up, affordability or economic impact. The state-intervention area is a target geography and the programme's future milestones can move with construction, demand and commercial coverage; premises passed and premises connected are different stages.

• 🇦🇪United Arab Emirateslinked to Public-private collaborationstarted 2018still operating

  UAE Pass, the shared digital identity layer

  The UAE Telecommunications and Digital Government Regulatory Authority reported that UAE Pass had 6.84 million registered users accessing more than 15,000 digital services from government and private-sector providers in 2023.

  6,840,000 users, registered uae pass users, 2023. 15,000 services, digital services available through uae pass, 2023. UAE Telecommunications and Digital Government Regulatory Authority, Digital Enablers Report 2023, retrieved 2026-08-31.

  How it worked: TDRA coordinated federal and emirate-level governments around one identity, signature and document layer, then allowed banks, telecoms and other private providers to connect to it. A shared authentication rail lets separate organisations reuse verified identity instead of each rebuilding onboarding.

  Limits: Registered accounts and listed services do not show active use, completion rates, accessibility, privacy experience or whether a transaction became faster for residents. The figures are reported by the authority coordinating the platform, and provider totals combine government, semi-government and private services with different quality and accountability arrangements.

• 🇬🇹Guatemalalinked to Public-private collaborationstarted 2003still operating

  Guatecompras, the national procurement marketplace

  Guatemala's Ministry of Finance reported 152,676 enabled Guatecompras users in April 2023, including 137,950 registered suppliers alongside public buyers, auditors and oversight bodies using the national procurement system.

  152,676 users, enabled guatecompras users, 2023-04. 137,950 users, enabled supplier users, 2023-04. Government of Guatemala, Ministry of Public Finance, Guatecompras user report, retrieved 2026-08-31.

  How it worked: Guatemala put public buyers, suppliers, auditors and oversight bodies on a shared procurement record, giving firms a common route into state demand and giving monitors a common transaction identifier. The platform turns a dispersed paper process into a searchable public-private interface.

  Limits: Enabled accounts are not the same as completed tenders, competitive prices, delivered works or lower corruption. The ministry's count is an administrative snapshot and does not show how often suppliers bid, whether data are corrected, or whether every public entity follows the system in practice.

• 🇬🇧United Kingdomlinked to University-industry collaborationstarted 1975still operating

  Knowledge Transfer Partnerships

  UKRI reported more than 1,000 Knowledge Transfer Partnerships running at one time and more than 1,100 Associate projects in May 2026, joining businesses with universities or other research organisations through jointly supervised innovation work.

  1,000 partnerships, knowledge transfer partnerships running, 2026-05-08. 1,100 projects, ktp associate projects, 2026-05-08. UK Research and Innovation, Knowledge Transfer Partnerships overview, retrieved 2026-08-31.

  How it worked: Innovate UK funds a three-way arrangement in which a business names a strategic problem, a university or research organisation supplies expertise and a graduate associate works inside the firm under joint supervision. The structure gives firms absorptive capacity while giving academic knowledge a route into operations.

  Limits: The partnership and project counts show programme scale, not whether collaborations produce durable productivity, patents, spinouts or jobs. UKRI reports the programme's administrative totals and benefit estimates; participation is not a representative sample of all UK firms or universities, and the figures do not provide a counterfactual against which the programme's effects can be judged.

• 🇦🇺Australialinked to University-industry collaborationstarted 1991still operating

  Cooperative Research Centres programme

  Australia's Department of Industry reported that its Cooperative Research Centres portfolio facilitated 304 businesses establishing research-focused collaborations in 2023–24, across 114 CRC and 190 CRC Projects partners, above the year's target of 248.

  304 businesses, businesses establishing research-focused collaborations, 2024. 248 businesses, portfolio target, 2024. Australian Department of Industry, Science and Resources, Annual Report 2023–24, retrieved 2026-08-31.

  How it worked: Australia funds industry-led research centres for three to ten years and shorter CRC Projects for up to three years, giving firms and researchers a shared programme, matching resources and a governance structure for applied work. The long and short formats let a mature collaboration and a focused translation project coexist.

  Limits: The partner count measures collaborations facilitated through grant agreements, not completed research, commercialisation, additionality or university-only links. It is an administrative portfolio measure whose reporting and retrospective calculations can change, and partners include research organisations and community bodies alongside universities and businesses.

• 🇵🇹Portugallinked to University-industry collaborationstarted 2017still operating

  Portugal's CoLAB applied-research network

  Portugal's Agência Nacional de Inovação reported 41 recognized Collaborative Laboratories (CoLABs), and its 2024 innovation report recorded 639 highly qualified staff contracted by CoLABs, 32 percent of them PhDs.

  41 CoLABs, recognized collaborative laboratories, 2024. 639 staff, highly qualified staff contracted by colabs, 2024. Agência Nacional de Inovação, Relatório Nacional de Inovação 2024, retrieved 2026-08-31.

  How it worked: Portugal gives mission-led applied-research organisations a formal recognition and monitoring route, then combines public base or competitive funding with industry-facing agendas so researchers have an institution whose job is to carry work across the university-business boundary.

  Limits: The network and staffing totals show institutional reach, not research quality, business adoption, additionality or durable jobs. CoLABs are separate private non-profit entities and report through the programme; the figures do not cover every university-industry link or establish that public recognition caused the reported staffing.

• 🇵🇹Portugallinked to Public-private collaborationstarted 2021still operating

  Agendas Mobilizadoras and green innovation alliances

  Portugal's PRR transparency portal reports 52 Agendas Mobilizadoras and Verdes representing €5,278,338,665.70 of approved investment and €3,051,632,468.10 of approved support as of 17 August 2026.

  52 consortia/agendas, agendas mobilizadoras and verdes represented, 2026-08-17. 5,278,338,665.7 EUR, approved investment across the agendas, 2026-08-17. Portal Mais Transparência / Estrutura de Missão Recuperar Portugal, PRR project C05-i01.01, retrieved 2026-09-02.

  How it worked: Portugal used an open call to assemble firms, research organisations, clusters and public bodies into contract-based consortia, then tied public support to innovation, production, skills and technology-transfer work inside each agenda.

  Limits: Approved investment and support are programme commitments, not money spent, products delivered, export growth or additional private R&D. The portal aggregates consortia with different sectors and execution rates, and the figures are reported by the programme itself; they show a coordination mechanism at national scale, not that every alliance will complete its milestones by the 2026 deadline.

• 🇮🇳Indialinked to Public-private collaborationstarted 2016still operating

  Unified Payments Interface, the bank-and-app network

  NPCI's UPI rail connected 731 live banks and settled 22,716.07 million transactions worth ₹28,92,138.67 crore in June 2026 through participating banks and payment apps.

  22,716,070,000 transactions, upi transactions settled in the month, 2026-06. 731 banks, banks live on upi, 2026-06. National Payments Corporation of India, UPI Product Statistics, retrieved 2026-09-02.

  How it worked: The Reserve Bank-regulated NPCI set a common protocol and settlement layer while banks and private payment apps competed at the customer interface, allowing one public standard to scale through many firms.

  Limits: UPI volume and bank participation show a shared transaction rail, not the quality of public-private coordination across other national objectives. NPCI publishes the figures for the system it operates, the table excludes on-us transactions from August 2018 onward, and volume says nothing about access, fraud, outages or distributional reach.

• 🇿🇦South Africalinked to Public-private collaborationstarted 2011still operating

  Renewable Energy Independent Power Producer Procurement Programme

  South Africa's IPP Office reported 95 independent power producers operational with 7,335 MW online by 31 March 2025, alongside 9,618 MW procured from 116 projects at financial close and R292.2 billion attracted for energy infrastructure.

  7,335 MW, operational ipp capacity, 2025-03-31. 292,200,000,000 South African rand, investment attracted by projects at financial close, 2025-03-31. Independent Power Producers Office, South Africa, IPPPP Quarterly Report Q4 2024/25, retrieved 2026-09-02.

  How it worked: The state set capacity and development rules, ran repeated competitive tenders and gave private producers bankable power-purchase agreements, while Eskom and government agencies carried the grid and policy interfaces.

  Limits: The IPP Office reports the programme it manages, and its figures combine several procurement windows and technologies. Operational MW and financial close show a functioning public–private delivery channel, not affordability, reliability, local ownership or emissions outcomes; terminated projects are excluded from current totals and the investment figure is attracted capital rather than completed public value.

• 🇲🇽Mexicolinked to Public-private collaborationstarted 2019still operating

  Jóvenes Construyendo el Futuro, government-workplace training

  Mexico's labour ministry linked young people to more than half a million registered workplaces for on-the-job training, reporting 2.95 million beneficiaries nationwide after five years of Jóvenes Construyendo el Futuro in February 2024.

  2,950,000 people, young people benefited, 2024-02. 500,000 workplaces, workplaces registered to host training, 2024-02. Secretaría del Trabajo y Previsión Social, Government of Mexico, retrieved 2026-09-02.

  How it worked: The labour ministry paid a stipend and social insurance while firms, shops and civil-society organisations supplied real work placements, with a shared digital platform matching young people to host workplaces.

  Limits: The ministry's total is a cumulative programme count and the workplace figure is rounded; neither shows unique completions, the quality of mentoring or whether trainees stayed employed. The government funds the stipend and insurance, so the record demonstrates a large public-private delivery interface rather than private financing of the programme.

• 🇹🇷Turkeylinked to Public-private collaborationstarted 2008still operating

  e-Devlet Gateway, a shared public-service front door

  Türkiye's e-Devlet Gateway ended 2024 with 66,753,526 users, 8,309 online services and 1,080 participating institutions, including 127 private providers alongside central government, universities and municipalities.

  66,753,526 users, registered e-devlet gateway users, 2024. 1,080 institutions, institutions offering services through the gateway, 2024. Digital Transformation Office / e-Devlet Gateway, Government of Türkiye, retrieved 2026-09-03.

  How it worked: A central gateway gave public agencies, universities, municipalities and selected private providers one identity and service front door, while the state kept the common authentication and catalogue layer.

  Limits: User registrations and service listings measure reach and integration, not whether transactions complete, how long they take or whether private providers add value. The gateway's own statistics aggregate institutions with very different roles, and the 127 private providers are included as participants rather than evidence of joint financing or shared accountability.

• 🇹🇷Turkeylinked to University-industry collaborationstarted 2024still operating

  TÜBİTAK 1505 university–industry support programme

  TÜBİTAK's 1505 University–Industry Collaboration Support Programme received 116 project applications and began support for 59 projects in 2024, with 52 fellows working across the collaboration pipeline.

  59 projects, projects entering the support process, 2024. 116 applications, project applications received, 2024. Scientific and Technological Research Council of Türkiye (TÜBİTAK), 2024 activity report, retrieved 2026-09-03.

  How it worked: TÜBİTAK made firms' needs the entry point for university research, with a public grant and named academic and customer organisations sharing the project rather than leaving technology transfer to informal contacts.

  Limits: Applications and support starts measure the interface, not the number of products commercialised or the durability of university–firm relationships. The activity report is an administrative source and does not identify the firms, project quality or completion rate; 52 fellows are a staffing input, not an outcome.

• 🇮🇱Israellinked to Public-private collaborationstarted 2007still operating

  Seawater desalination under private concessions

  Israel's Water Authority contracted private operators to build and run large seawater desalination plants under long-term concessions, and by 2022 the five largest held concessions for about 596 million cubic metres a year and supplied 33 percent of the country's fresh water, with the state buying 540 million cubic metres from them that year for about NIS 1.5 billion.

  33 % of fresh water, share of fresh water supplied from desalination, 2022. 596.4 million cubic metres a year, production capacity in the concessions of the five largest plants, 2022. State Comptroller of Israel, Regulation and Oversight of Seawater Desalination Facilities, November 2024, retrieved 2026-10-01.

  How it worked: The state kept the planning, the water purchase and the regulation, and let private consortia finance, build and operate each plant on a long-term concession with a guaranteed buyer, so capacity arrived on a tender schedule without the capital sitting on the budget.

  Limits: The source is an audit, and most of it is criticism: it finds that the Water Authority never completed the national desalination master plan a 2010 government resolution required, that the agreed penalties for poor water quality are too low to enforce anything, that 14 seawater contamination incidents halted production between 2007 and 2023, that three of the largest plants sit near one river mouth, and that no serious preparation had begun for the Palmachim concession expiring in 2029. A third of fresh water is the 2022 national share; desalination's share of drinking water is higher, but the audit does not state it. The started year is the first year the audit reports plants in operation; the first plant opened earlier. Capacity is a contractual figure, not output.

• 🇷🇼Rwandalinked to Public-private collaborationstarted 2016still operating

  Government of Rwanda and Zipline drone delivery network

  From 2016 the Government of Rwanda contracted Zipline to fly blood and later other medical products to health facilities, and by 2024 the Rwanda Biomedical Centre reported deliveries to 624 sites across 27 districts and over 12,000 packages in a single month.

  624 sites across 27 districts, delivery sites served by drone, 2024. 12,000 packages per month (lower bound), packages delivered in a single month, 2024. Rwanda Biomedical Centre, Head of the Blood Transfusion Service, 'Revolutionizing Rwanda's Supply Chain: The Impact of UAVs (2016-2023)', presented at the ICAO Advanced Air Mobility Symposium 2024 and hosted by ICAO, retrieved 2026-10-01.

  How it worked: The Ministry of Health and the Rwanda Biomedical Centre bought a delivery service rather than drones: Zipline built and flew the network, the state paid per delivery, centralised the blood supply, and the civil aviation authority opened the airspace.

  Limits: The source is a slide deck by the public partner, presented at an aviation conference. It is not a statistical release, and its operating figures almost certainly come from Zipline, the private contractor, so the number is the two partners reporting on themselves. 'Today' is undated beyond the 2024 symposium, and 'over 12,000 packages in a single month' is a lower bound for an unnamed month, not an annual series. The site count mixes health facilities with farms and other drop points, and 27 districts is not all 30. The deck's outcome claims (90 percent less wastage, 99 percent less maternal mortality from blood shortage) are unreferenced on the slide and are not used here. The record shows scale, not value for money: contract prices, per-delivery cost to the health budget, and whether road logistics would have done as well are not shown. The collaboration is with one foreign firm under a service contract, which is a narrow form of the joint public-private delivery the indicator means.

• 🇭🇳Honduraslinked to Public-private collaborationstarted 2016dismantled

  Empresa Energía Honduras loss-reduction partnership, and its reversal

  Honduras's state utility ENEE awarded Empresa Energía Honduras a public-private partnership contract in December 2015, with service starting in 2016, to run national distribution and cut technical and non-technical losses by 17%. Distribution losses fell from 33.38% to 27.11% by February 2018, rose to 35.37% by December 2020 and stood at 30.77% in December 2021. The state intervened the contract and took distribution back when it ended in August 2023.

  30.77 % of energy, electricity distribution losses, december 2021 (technical and non-technical), 2021-12. 27.11 % of energy, lowest distribution losses under the partnership (february 2018), the gain that was lost, 2018-02. Secretaría de Energía de Honduras, Dirección General de Electricidad y Mercados, Informe Estadístico Anual del Subsector Eléctrico Nacional 2021 (Tabla 11, from ENEE Subgerencia de Planificación y Gestión Comercial), retrieved 2026-10-01.

  How it worked: ENEE, through a Congress-approved PPP trust supervised by SAPP and an external supervisor, handed national distribution and loss recovery to a private operator paid from distribution revenue. When losses rose, the state intervened the contract and then let it lapse, moving loss recovery into its own Programa Nacional para la Reducción de Pérdidas.

  Limits: The loss series is ENEE's own commercial accounting, republished by the Energy Secretariat under the government that intervened the contract and campaigned against it, so the source is not neutral. The table stops in December 2021. The 2022 and 2023 values, including the roughly 37 to 38 percent widely reported when the contract ended, appear only in press coverage, and this record does not carry them. The SEN text does not attribute the 2016 to 2018 fall to EEH by name. The 2020 rise coincides with COVID-19 lockdowns and hurricanes Eta and Iota, so not all of the reversal can be laid on the operator. The contract was intervened and then expired at the end of its term in August 2023. It was not cancelled outright, and 'dismantled' here means the partnership model ended and ENEE took distribution back, not that a court annulled it. The SAPP fact sheet has copy errors: a malformed start date, and a repayment source copied from an airport sheet ('COBROS POR SERVICIOS AEROPORTUARIOS'). The 17% target comes from SAPP's project description, not from the contract text. A loss percentage measures the outcome of the partnership and says nothing directly about how deep the collaboration was. It is evidence that a structured government-firm arrangement on a national goal was tried at national scale and failed, not a measure of collaboration.

• 🇸🇻El Salvadorlinked to Public-private collaborationstarted 1993dismantled

  INSAFORP tripartite training institute, and its dissolution

  El Salvador ran its national vocational training system from 1993 through INSAFORP, an institute funded by a 1% employer payroll levy and governed by a tripartite council of four government, four employer and four worker representatives; a 2021 reform gave the President the choice of the employer seats, and Decree 893 of 14 November 2023 dissolved INSAFORP and handed its functions to INCAF, whose council holds only the institute's president and four ministries.

  0 seats, employer-sector seats on the governing council of the national training institute under the 2023 incaf law, 2023-11-14. 4 seats (of 12, four per sector), employer-sector seats on the insaforp consejo directivo under the 1993 ley de formación profesional (peak structure, as read from the recital of the 2021 reform), 2021-06-03. Legislative Assembly of El Salvador, Decreto N.° 893 (Ley del Instituto Nacional de Capacitación y Formación), as published by the Supreme Court's Centro de Documentación Judicial, retrieved 2026-10-01.

  How it worked: El Salvador's Legislative Assembly, on the executive's initiative, first moved the choice of INSAFORP's employer representatives to the President (2021) and then dissolved the tripartite institute, replacing it with a ministry-run council while keeping the employer payroll levy (2023). The state kept the money flow from firms and removed the firms' and workers' seats at the table.

  Limits: The metric is a count of seats read off the text of two laws, not a measured outcome: the value 0 is our reading of Art. 11 of Decree 893, which lists only the INCAF president and four ministries, and the peak of four employer seats comes from a recital in the 2021 reform, not from the 1993 law itself, which we did not open. The 1% levy rate in the claim is read from Art. 19 of Decree 893, which carries the levy over, and was not read from the 1993 text. Neither text shows how much the employer and worker seats shaped training decisions in practice, how many people INSAFORP trained, or whether training volume or quality fell under INCAF; no official participant series was retrieved, so this record documents the end of a governance arrangement, not a fall in delivery. Firms still fund the system: Art. 19 keeps the 1% employer payroll levy, although it reserves only 20% of it for INCAF's training and sends the remaining 80% elsewhere, so the levy-funded collaboration was reshaped rather than abolished in money terms. The collaboration was already eroded before dissolution: the 2021 reform let the President pick employer representatives from ternas built by the Labour Ministry, without requiring membership of employer associations. The Supreme Court documentation host returned intermittent HTTP 522 errors on the day of retrieval before serving both documents.

• 🇺🇸United Stateslinked to Public-private collaborationstarted 2021dismantled

  Affordable Connectivity Program, and its lapse

  The Federal Communications Commission ran the Affordable Connectivity Program from 31 December 2021, paying monthly broadband discounts through voluntarily participating internet providers to more than 23 million enrolled households, and ended it on 1 June 2024 when the funding Congress had provided ran out.

  23 million households, households enrolled in the acp when new enrolments stopped, 2024-02. Federal Communications Commission, press release 'FCC brings Affordable Connectivity Program to a close' (DOC-402930A1), retrieved 2026-10-01.

  How it worked: Congress funded a fixed pot, the FCC turned it into a monthly discount that private internet providers applied on their own bills, and enrolment climbed until the pot was exhausted; when Congress did not refill it, the FCC froze enrolment, cut the benefit to a partial discount and closed the program.

  Limits: The 23 million figure is the regulator's own enrolment count, taken at the February 2024 freeze, and it reports sign-ups, not households that would otherwise have gone without broadband: the FCC's claim that 68 percent of ACP households had inconsistent or no connectivity before rests on a self-reported agency survey. The program was a voucher paid through private providers whose participation was voluntary, so a hostile reviewer would call it a subsidy channel rather than joint delivery; the public-private element is that providers carried enrolment, billing and the discount. It grew out of the pandemic-era Emergency Broadband Benefit, so the 2021 start understates how long the capability was built. It ended through a funding lapse rather than a repeal, and the FCC said it could resume if Congress appropriated money, so 'dismantled' records the end of delivery, not the destruction of the administrative machinery. The release is an agency press statement, not an audited statistical series, and the PDF is the script-readable copy; the fcc.gov program pages block automated requests.

• 🇺🇸United Stateslinked to Public-private collaborationstarted 2020delivered and closed

  Operation Warp Speed

  Operation Warp Speed, an HHS and DOD partnership, financed private vaccine makers' development, at-risk manufacturing and advance purchases, including a $5.97 billion contract for 300 million Pfizer/BioNTech doses. Three of the vaccines it bought held FDA emergency use authorization by 1 March 2021, the first from 11 December 2020.

  3 vaccines, ows-purchased covid-19 vaccines holding fda emergency use authorization, 2021-03-01. 5.97 USD billion, ows contract value for 300 million pfizer/biontech doses, 2021-03-01. Congressional Research Service, Insight IN11560 'Operation Warp Speed Contracts for COVID-19 Vaccines and Ancillary Vaccination Materials', version 7, retrieved 2026-10-01.

  How it worked: HHS and DOD pooled BARDA's countermeasure contracting with defense logistics and paid private manufacturers to build production capacity while trials were still running, carrying the financial risk of failure so that doses existed the day authorization came.

  Limits: The metric counts authorized products, not doses delivered or people vaccinated. It does not show what share of the speed came from federal money rather than private R&D already under way: Pfizer/BioNTech took no federal development funding and joined only through a purchase contract, as the CRS notes. The contract values are awards, not outlays, and CRS cites BARDA and DOD pages without auditing them; the GAO's audit of OWS sits on gao.gov, which blocks automated requests, so it is not the source here. The CRS Insight is a congressional research summary rather than a statistical series, and it does not state the May 2020 launch date, so 'started' is the year vaccine delivery began. Ancillary-supply shortages and distribution problems are flagged in the same document. The Biden administration said it planned to restructure and rename OWS, so 'concluded' marks the end of the vehicle, not of federal vaccine purchasing.

• 🇫🇷Francelinked to Public-private collaborationstarted 2013still operating

  Plan France Très Haut Débit

  Under the Plan France Très Haut Débit, defined by the government in February 2013, private operators and local-authority public initiative networks built fibre so that 43.1 million of France's 45.2 million premises, 95.3 percent, were connectable to fibre at the end of June 2026.

  43,100,000 premises, premises connectable to fibre to the home, 2026-06-30. 970,000 premises, premises still to be made connectable in public initiative network zones, 2026-06-30. Arcep, the French telecoms regulator, Marché du haut et du très haut débit fixe, results to end June 2026, retrieved 2026-10-01.

  How it worked: The state split the territory: private operators made binding rollout commitments in profitable areas, which Arcep can enforce. In the rest, départements and regions tendered public initiative networks, built and run by private firms under concession and co-financed by a state subsidy fund.

  Limits: The national coverage figure mixes the parts of the plan. In the very dense zones and the private-initiative zones (AMII and AMEL), private operators invested on their own account, so not every connectable premise is a product of joint delivery. Only the public initiative network zones are the clearly public-private part. Arcep's release gives what remains to be built there, 5 percent, but not the zone's total size. 'Connectable' means a line passes the premises, not that anyone subscribes or that the service is good. The 20 billion euro envelope and the roughly 3 billion euro state subsidy are 2013 announcements, not audited spending. The envelope was announced for the ten years from 2013, and 2.1 million premises were still not connectable in mid-2026. Arcep says more than a third of those are premises where a third party refused the works or the building is still under construction. The deployment figures are reported by operators to the regulator.

• 🇺🇾Uruguaylinked to University-industry collaborationstarted 2009still operating

  ANII Alianzas para la Innovación

  Uruguay's national research and innovation agency ANII began co-funding joint academia-firm projects through Alianzas para la Innovación in 2009 and had signed 29 such contracts, committing US$3.85 million and drawing US$1.71 million from firms, by 30 June 2016; in 2023 it approved 23 projects joining the productive sector and academia, matching the 2022 record.

  29 projects, alianzas para la innovación contracts signed since 2009, 2016-06-30. 23 projects, projects approved in the year involving the productive sector with academia, 2023. Agencia Nacional de Investigación e Innovación, Unidad de Evaluación y Monitoreo, evaluation of Alianzas para la Innovación and Redes Tecnológicas Sectoriales, retrieved 2026-10-01.

  How it worked: ANII, the state innovation agency, offered non-repayable co-funding only to projects where a firm brought a problem and money and a university or research institute brought the research team, with a firm employee placed inside that team.

  Limits: The scale is small: 29 contracts in seven years and US$3.85 million committed is a national instrument, but a thin one, and it says little about how much Uruguayan universities and firms collaborate outside ANII funding. Both figures are the agency reporting on its own instruments, the first from its internal evaluation unit and the second from a year-end press note. The 2023 figure is not limited to Alianzas: it counts every approved project the agency describes as joining firms with academia, under its technology-transfer line, so the two metrics are not one series and no trend can be drawn between them. Signed contracts measure funding decisions, not completed research, patents, products or lasting relationships; the evaluation shows only 14 of the 29 finished by mid-2016. The evaluation report is undated on its cover and the date comes from its own tables.

• 🇨🇱Chilelinked to Public-private collaborationstarted 1993still operating

  Sistema de Concesiones de Obras Públicas

  Chile's Ministry of Public Works has delivered public infrastructure through private concession contracts since 1993; in April 2026 its Concessions Directorate reported 81 contracts in force with US$29.615 billion of committed investment across motorways, airports, hospitals, prisons, a reservoir, a desalination plant and a cable car.

  29,615 million US dollars, committed investment in concession contracts in force (materialised plus pending), 2026-04. 81 contracts, concession contracts in force, 2026-04. Ministerio de Obras Públicas, Dirección General de Concesiones, Cuenta Pública 2025, retrieved 2026-10-01.

  How it worked: The Ministry of Public Works tendered build-operate-transfer contracts under a dedicated concessions law, letting private consortia finance and run motorways and later airports, hospitals and prisons, recovered through tolls and state payments, while a dedicated directorate supervised the contracts.

  Limits: The figures are the Directorate's own slide figures, read from a presentation PDF, not audited accounts. Committed investment adds money already spent to money still to be spent, and 35 of the 81 contracts are still in construction, so the headline overstates infrastructure delivered. The number covers contracts in force only, not concessions that have ended, so it is not the system's cumulative output. It says nothing about value for money, renegotiations and supplementary agreements (a known weakness of the Chilean model that the slide does not quantify), contingent fiscal liabilities such as guaranteed minimum revenues, or who pays through tolls. Chile also keeps a large conventional public works budget, so concessions are one channel, not the whole of infrastructure delivery.

• 🇲🇾Malaysialinked to Public-private collaborationstarted 2018still operating

  DuitNow on the Real-time Retail Payments Platform

  Bank Negara Malaysia merged its clearing subsidiary with the banks' payment consortium into PayNet, jointly owned by the central bank and the financial industry, which rolled out the Real-time Retail Payments Platform with DuitNow from December 2018; BNM statistics record 5,438.6 million RPP transactions worth RM3,638.1 billion in 2025, up from 109.3 million in 2019.

  5,438.6 million transactions, rpp (duitnow) transaction volume, 2025. 109.3 million transactions, rpp (duitnow) transaction volume in the first full year, 2019. Bank Negara Malaysia, Payment Statistics, Table 3 Payment Systems (transaction volume and value), retrieved 2026-10-01.

  How it worked: Bank Negara Malaysia merged its own clearing house with the banks' payment consortium into one jointly owned operator, then used it to build a shared real-time rail with a national proxy (phone and ID number) addressing database that every bank connects to.

  Limits: Transaction volume measures use of a shared rail, not the depth of collaboration: it does not show how costs, governance or pricing are split between BNM and the banks that owned MEPS (11 domestic banks per the 2017 release), nor whether non-bank participants get fair access in practice. Part of the growth substitutes for other electronic channels (Interbank GIRO, Instant Transfer) rather than adding new payments, and the 2019 base is the platform's ramp-up year, so the multiple overstates organic growth. The joint-ownership fact comes from a 2017 announcement of intent ('will be jointly owned'), not a current shareholder register. The central bank is both shareholder and overseer, so this is a central-bank-led utility as much as a partnership.

• 🇵🇱Polandlinked to University-industry collaborationstarted 2017still operating

  Doktorat wdrożeniowy (implementation doctorate) programme

  Poland's science ministry opened the implementation doctorate in the 2017/2018 academic year, funding doctoral candidates to do their research inside a firm or institution. In its first five editions 2,293 projects were approved, 479 people qualified for the sixth edition in 2022/2023, and 69.6 percent of the doctorates ran in the private sector.

  2,293 approved doctoral projects, implementation doctorate projects approved in the first five editions, 2022. 479 doctoral candidates, candidates qualified for the sixth edition, 2022/2023. Polish Ministry of Education and Science, '5 lat programu Doktorat wdrożeniowy', retrieved 2026-10-01.

  How it worked: The science ministry paid a scholarship to doctoral candidates who were also employed by a firm, with a supervisor on each side. The firm gets a researcher working on its own problem and the university gets a funded doctorate. Each edition was a national competition that universities applied to.

  Limits: The figures are the ministry reporting on its own programme. They restate a report by the state-owned Łukasiewicz Research Network, not a statistical release. 2,293 is approved projects, not completed doctorates. The page gives no completion, dropout or defence rate, and no figure for patents, products or firm R&D that came out of the work. The latest count read here is from 2022. Figures for the editions since then, including later press totals of about 3,000 participants, were not verified at an official source. A co-funded PhD placement is one channel of university-firm collaboration, and it does not measure the intensity of joint research the indicator asks for, such as co-publications, contract research or co-patenting. 30.4 percent of placements are in the public sector, including state-owned companies and public institutions, so not every project links a university with a private firm. Firms may use the scholarship as a wage subsidy for staff they would have employed anyway.

• 🇸🇬Singaporelinked to Public-private collaborationstarted 2013dismantled

  Tuaspring desalination partnership, and its collapse

  PUB contracted Hyflux's Tuaspring Pte Ltd under a water purchase agreement for Singapore's largest desalination plant, opened in 2013, then terminated the agreement on 17 April 2019 after the operator failed to keep the plant reliably operational, took the plant into public ownership in May 2019 and spent around $80 million on repairs due for completion in mid-2021.

  80 million SGD, pub repair cost after taking over the failed partnership plant, 2021-02. Ministry of Sustainability and the Environment, Singapore, written reply to Parliamentary Question on Tuas South Desalination Plant, retrieved 2026-10-01.

  How it worked: PUB bought desalinated water from a private developer under a long-term purchase agreement and, when the developer's finances failed, used the contract's default and termination remedies to take the plant into public ownership and repair it itself.

  Limits: This is one contract failing, not the end of Singapore's public-private water model: PUB's own page lists Keppel Marina East (2020) and Jurong Island (2022) desalination plants built and run by private firms after the collapse, so the partnership approach survived and the record documents the loss of one delivery inside it. The $80 million is a ministerial answer to a parliamentary question, not an audited figure, and it covers repairs only, not PUB's operating costs since takeover or what the private partner's investors lost. The original 25-year term and contracted capacity of the agreement are reported only by press and secondary sources found today, so the record does not carry them. A hostile reviewer can also read the episode as the state's contractual safeguards working, which is how the minister presented it: the water supply was not interrupted.

• 🇨🇦Canadalinked to Public-private collaborationstarted 2009dismantled

  PPP Canada and the P3 Canada Fund, and its dissolution

  Canada's federal Crown corporation PPP Canada, operational from 2009, committed over C$1.3 billion through the P3 Canada Fund to 25 public-private partnership projects across the country before the government dissolved it, with operations ceasing at the end of 2017 and the corporation dissolved effective March 31, 2018.

  1,300,000,000 Canadian dollars (committed, 'over'), p3 canada fund investments committed over the corporation's life, 2018-03. 25 projects, projects in the p3 canada fund portfolio at dissolution, 2018-03. PPP Canada Inc., Annual Report 2017-18 (published by Housing, Infrastructure and Communities Canada), retrieved 2026-10-01.

  How it worked: The federal government created a Crown corporation with a dedicated fund that paid up to a quarter of eligible costs to provinces, municipalities and Indigenous communities that procured infrastructure as public-private partnerships, plus template contracts and screening of federal projects; once it judged the market mature, it folded the corporation into the infrastructure department.

  Limits: This is a dismantling by decision, not a collapse. The government's stated reason was that PPP Canada had 'successfully fulfilled its mandate', and its files were transitioned to Infrastructure Canada. The corporation carried performance-based payment liabilities of approximately C$1.3 billion over five years. A hostile reviewer would say the record shows a federal agency closing, not a national capability lost. The C$1.3 billion is commitments rather than disbursements (the audited statements show far smaller annual payments to recipients) and is a federal top-up of up to 25% of eligible costs, so it does not measure the scale of the joint delivery, its value for money or whether the 25 assets performed. The narrative and the figure are the corporation's own annual report; only the financial statements carry an Auditor General audit opinion.

• 🇨🇳Chinalinked to Public-private collaborationstarted 2015dismantled

  National PPP programme and project library, and its dismantling

  China's Ministry of Finance ran a national government and social capital cooperation (PPP) programme under the 2015 State Council guidance, whose management library held 10,332 projects with 16.6 trillion yuan of investment by the end of October 2022; in November 2023 the State Council ended the 2015 regime for every project not yet procured and replaced it with a concession-only, user-fee mechanism that bars availability payments.

  10,332 projects, projects in the national ppp management library, cumulative, 2022-10. 1,695 projects, projects removed from the library in the 2017-2018 clean-up, 2018-04-23. Ministry of Finance of the People's Republic of China (mof.gov.cn, republishing China Finance and Economics News), retrieved 2026-10-01.

  How it worked: The Ministry of Finance built a national PPP framework, a central project library and disclosure platform from 2014-2015 and pushed local governments to procure infrastructure through it; when the library became a channel for hidden local debt, the Ministry purged it in 2017-2018 and the State Council replaced the regime in 2023 with user-fee concessions run by the planning commission.

  Limits: Both project figures sit in articles by Ministry of Finance affiliated newspapers (China Finance and Economics News, Economic Daily) republished on mof.gov.cn, not in a Ministry statistical release; the Ministry's own monthly library reports on jrs.mof.gov.cn stop at January 2020, so the 2022 figure cannot be checked against a primary table. The source is a ministry press page rather than a statistics release. A count of library projects measures entries in an administrative register, not collaboration: many entries were government-paid availability-payment schemes that the Ministry itself purged in 2017-2018 as disguised local borrowing, so a large share of the 'partnership' was public debt with a private contractor in front. The library count is cumulative and does not show how many projects reached operation or delivered value for money. Dismantling applies to the 2015 regime for new projects; existing contracts keep running and PPP continues as user-fee concessions under the National Development and Reform Commission, so this is the end of one national model, not of public-private delivery in China. The 2023 notice gives the reasons in general terms (hidden local-government debt, private investment) and no figure for what failed.

• 🇮🇪Irelandlinked to University-industry collaborationstarted 2013still operating

  SFI Research Centres programme

  Science Foundation Ireland funded a network of 16 university-based research centres from 2013, whose cumulative SFI investment of €684 million was matched by €638 million of industry commitment by 2021 (Annual Report 2021), and which by the end of 2022 had signed 1,426 collaborative research agreements with 688 companies, 366 of them in Ireland.

  1,426 agreements, cumulative collaborative research agreements signed by sfi research centres with companies, end of 2022. 688 companies, companies party to those agreements, end of 2022. Science Foundation Ireland, Annual Report 2022, retrieved 2026-10-01.

  How it worked: Science Foundation Ireland funded large, multi-university research centres for six-year terms and made part of each award conditional on industry co-funding, so every centre had to sign agreements with firms to keep its public money.

  Limits: These are the funder's own counts of its own programme, in an annual report, inside an 87-page PDF. An agreement count says nothing about the size, duration or research content of each agreement; in 2022 alone 136 agreements carried €32.2 million of industry commitment. The €638 million is industry commitment, not verified cash received, and the programme's design required industry co-funding, so collaboration is partly a condition of funding rather than demand from firms. 317 of the 688 companies are multinationals, so the figure mixes Irish firms' engagement with foreign affiliates attracted by tax and investment policy. Science Foundation Ireland was merged into Research Ireland in 2024, so the series ends with 2023 reporting under the old name and the sfi.ie URLs may move.

• 🇮🇳Indialinked to Public-private collaborationstarted 2017operating below its peak

  UDAN regional air connectivity scheme, and its lapsed routes

  India's Ministry of Civil Aviation paid viability-gap funding to private airlines to fly underserved regional routes from April 2017. Of the 774 routes awarded in the first three rounds, 371 commenced, and only 174 were still flying in March 2023; only 54 of the 112 routes that completed their three-year concession kept operating afterwards.

  174 routes, udan 1 to 3 routes in operation, 2023-03. 371 routes (cumulative), udan 1 to 3 routes that commenced operations, 2023-03. Comptroller and Auditor General of India, Report No. 22 of 2023, Compliance Audit on Regional Connectivity Scheme - UDAN, Ministry of Civil Aviation, retrieved 2026-10-01.

  How it worked: The Ministry of Civil Aviation auctioned regional routes to private airlines, paid the winners a viability-gap subsidy for three years funded by a levy on other flights, and asked states and the Airports Authority to revive unused airstrips; most routes did not survive the end of the subsidy.

  Limits: The 371 is a cumulative count of routes that ever commenced, not a simultaneous peak, so the fall to 174 overstates how many routes were flying at any one time. Routes from the third round had less time to lapse, which flatters their survival rate. The audit covers UDAN 1 to 3 to March 2021 with route figures updated to March 2023, so later rounds and the scheme's current state are not shown. A lapsed route is a commercial failure for the airline as much as a policy failure, and some routes stopped because airlines collapsed. The figures come from the scheme's own RCS cell, as the audit notes. The 120 page report is a 12 MB PDF and the quoted table is read from text extraction.

• 🇹🇷Turkeylinked to Public-private collaborationstarted 1986still operating

  Kamu-Özel İşbirliği, the national public-private partnership programme

  Turkey's central government signed 277 public-private partnership contracts from 1986 to September 2025, worth US$112.09 billion in September 2025 prices across transport, energy and health, with 262 in operation, including 18 PPP city hospitals holding 29,325 registered beds.

  112.09 US$ billion, September 2025 prices, total investment size of signed ppp contracts since 1986, 2025-09. 29,325 beds, registered beds in the 18 city hospitals completed under ppp, 2025. Presidency of the Republic of Türkiye, Strategy and Budget Office (SBB), 2026 Presidential Annual Programme, retrieved 2026-10-01.

  How it worked: Successive governments, and the Strategy and Budget Office as coordinator, used build-operate-transfer, build-lease-transfer and operating-rights contracts to move large transport, energy and hospital builds off the annual budget. The state then pays through guarantees and availability payments over the contract life.

  Limits: The registry's own note on this gap says PPP investment databases measure infrastructure finance, not collaboration capacity, and this record rests on exactly that kind of figure. US$112.09 billion is contracted investment size in current prices, not money spent or value delivered. It includes transfers of operating rights (İHD), which are closer to concessions or privatisation than to joint delivery. The government reports on its own programme. The same document shows what the collaboration costs the public side: in 2026 it programs 57.5 billion TL for goods and services and 78.7 billion TL in capital spending to run the completed PPP city hospitals, and adds 101.8 billion TL in current transfers to finance demand guarantees on completed PPP transport projects. A hostile reviewer would say those guarantees left the risk with the state, so the record shows private construction finance with public risk-bearing more than shared delivery. No cost or schedule performance is reported for individual contracts, and the count spans four decades and very different legal models.

• 🇫🇮Finlandlinked to University-industry collaborationstarted 2008dismantled

  Strategic Centres for Science, Technology and Innovation (SHOKs), terminated in 2015

  Finland ran six Strategic Centres for Science, Technology and Innovation from 2008, company-form consortia of firms, universities and research institutes funded about 60 percent by Tekes, with funding of EUR 88 million in 2014; the Sipilä government's austerity programme ended the earmarked funding at the end of 2015 and the programme was closed, with the last Tekes-funded projects wound down between 2016 and 2018.

  88,000,000 euros, shok funding for industry-academia knowledge transfer, 2014. 6 centres, shok companies operating before termination, 2014. OECD STIP Compass, Knowledge Transfer and Policies project: Case study on the policy mix for science-industry knowledge transfer in Finland (Halme et al., 4Front and Gaia Consulting with Business Finland), Box 1, retrieved 2026-10-02.

  How it worked: The state made long-term, multi-year research programmes conditional on firms, universities and institutes owning a joint company and co-writing its research strategy, with the innovation agency paying the majority share. The structure held collaboration together only while the earmarked public share kept flowing.

  Limits: The figure is inside a PDF (page 12, Box 1). The case study is hosted by the OECD as a country contribution to its knowledge-transfer project, but it was written by Finnish consultants with a Business Finland co-author, so it is the funder's side of the story rather than an OECD assessment. The EUR 88 million is one year's funding and the source does not say whether it is Tekes money alone or the full public and private budget; the same box says Tekes covered on average 60 percent. Funding is an input: the source carries no count of joint projects, firms or outcomes. The end was a budget decision, not a judgement that the centres failed: the same document says two of the six SHOKs (DIMECC and CLIC) continued as open innovation companies owned by firms, universities and institutes, and that ecosystem funding moved to Business Finland's Growth Engine instrument, so collaboration capacity was partly redeployed rather than wholly lost. The 2013 external evaluation the case study cites (Licence to SHOK?) was openly critical of the programme's governance and its limited commercial and scientific impact. The same PDF dates the programme's start to 2006-2007 in Box 4 and to 2008 in Box 1; started follows Box 1.

• 🇨🇭Switzerlandlinked to University-industry collaborationstarted 2018still operating

  Innosuisse innovation projects with implementation partners

  Innosuisse, the Swiss federal innovation agency, recorded CHF 113.95 million of contributions in 2025 to innovation projects in which a Swiss research institution works with a company or other implementation partner that bears 40 to 60 percent of the project cost, and held CHF 144.3 million of further commitments to such projects at the end of the year.

  113,950,000 Swiss francs, contributions to innovation projects with implementation partner, 2025. 144,311,000 Swiss francs, commitments to innovation projects with implementation partner outstanding at year end, 2025-12-31. Innosuisse, Swiss Innovation Agency, Geschäftsbericht 2025 (annual report and financial statements), notes 17 and 22, retrieved 2026-10-02.

  How it worked: The federal agency pays only the research institution's side of a project, and only when a partner in the economy bears a large share of the cost and contributes cash, so public money follows demonstrated private demand for academic expertise.

  Limits: The figures are inside a PDF, in the financial statements (PDF page 32, printed page 29, note 17; and PDF page 36, printed page 33, note 22), in thousands of francs. This is the agency reporting on its own spending. The money is the public share paid to research partners: it does not include the firms' own 40 to 60 percent or their cash contributions, so it understates the joint effort, and the accounts give no count of projects or firms. Research partners include universities, universities of applied sciences and non-university research bodies, and implementation partners can be administrations and non-profits as well as firms, so not every franc is university-firm collaboration. Spending measures input, not whether projects produced products. Started is 2018, the year the current agency took over the instrument from its predecessor; neither cited source states that date, and the instrument is older.

• 🇫🇷Francelinked to University-industry collaborationstarted 1981still operating

  Cifre industrial doctorates

  France's research ministry has funded Cifre industrial doctorates since 1981, paying an employer 42,000 euros over three years to hire a doctoral candidate whose thesis research is run in partnership with an academic laboratory; ANRT, which operates the scheme for the ministry, accepted 1,700 new Cifre in 2025 across 1,029 distinct employers.

  1,700 doctoral contracts, new cifre industrial doctorates accepted, 2025. 1,029 organisations, distinct beneficiary employers, 2025. Association nationale de la recherche et de la technologie (ANRT), operator of the Cifre scheme for the Ministry of Higher Education and Research, Rapport d'activité 2025, retrieved 2026-10-02.

  How it worked: The state subsidises a firm to employ a doctoral candidate whose thesis is supervised by an academic laboratory, and requires a collaboration contract between employer and lab, so each grant creates a person who works in both and a formal research agreement between them.

  Limits: The figures are inside a PDF (PDF pages 17 and 19, printed pages 15 and 17). ANRT is an association that runs the scheme on the ministry's behalf, so this is the programme reporting on itself; the ministry's own page was not reachable to cross-check. Employers include local authorities and associations (13 percent in 2025) as well as firms, so not every contract is university-industry collaboration. The count is of new contracts, not completed theses or research outputs. 2025 was below the scheme's own target: the same report says the ministry expected growth from 1,850 to 1,950 new Cifre and that 1,700 fell 13 percent short, with large firms' demand down 19 percent, so the 2025 figure is a dip against the 2024 level rather than a peak.

• 🇸🇪Swedenlinked to University-industry collaborationstarted 1995delivered and closed

  Competence Centres Programme (Kompetenscentrumprogrammet) 1995 to 2005

  Sweden's NUTEK, and later Vinnova and the Swedish Energy Agency, ran a ten-year programme from 1995 that built 28 competence centres at eight universities, each founded on a three-party contract between a university, a group of firms and the agency; more than 300 companies joined as contracting parties and industry put in 1.7 billion kronor between 1995 and 2005.

  300 companies (more than), companies participating as contracting parties, 1995-2005. 1,700,000,000 Swedish kronor, industry contribution to centre research, 1995-2005. Vinnova (Swedish Governmental Agency for Innovation Systems), VINNOVA Analys VA 2011:07, Effektanalys av starka forsknings- och innovationssystem, retrieved 2026-10-02.

  How it worked: Each centre was a contract among a university, a group of firms and the state agency, with matched funding from all three, boards with a non-academic majority and chair, and staged three-year funding that depended on international evaluation, which kept firms engaged in the whole research agenda instead of single projects.

  Limits: The figures are inside a PDF (PDF page 41, printed page 39). The report is published by Vinnova but written by evaluation consultants (Faugert & Co and Technopolis), and Vinnova notes that its analysis series does not necessarily state the agency's position; it in turn cites Vinnova's own 2006 summary. Roughly 70 percent of the industry contribution was in kind, so the 1.7 billion kronor is mostly valued staff time and equipment, not cash. The company count is of contracting parties over ten years, not firms active at any one time, and about 50 firms were counted in more than one centre. The programme ended as planned and was succeeded by VINN Excellence Centres, so its end is not a reversal; this record says nothing about the current centres.

• 🇪🇸Spainlinked to University-industry collaborationstarted 2014still operating

  Doctorados Industriales (State Research Agency industrial doctorates)

  Spain's State Research Agency funds four-year contracts for doctoral candidates employed in firms and other entities to write a thesis on an industrial research or experimental development project; in the 2024-2 call, resolved in October 2025, it awarded 65 grants worth 7.6 million euros.

  65 grants, industrial doctorate grants awarded, call din 2024-2, 2025-10. 7,600,000 euros, funding awarded, call din 2024-2, 2025-10. Agencia Estatal de Investigación (State Research Agency), news release on the award resolution for Doctorados Industriales DIN 2024-2, 24 October 2025, retrieved 2026-10-02.

  How it worked: The state co-funds a firm's employment of a doctoral candidate for four years on condition that the thesis is built on the firm's own research or development project, tying a university doctoral programme to a company problem through one person.

  Limits: The figure is one call's awards, not the stock of industrial doctorates in progress; The 2024 round was split into two calls and only the second is counted here, so the figure understates a year's awards. It is small next to France's Cifre. The release does not mention universities at all: the university side is implied by the doctoral thesis, not shown. It does not count the firms, and beneficiaries include 'other entities' as well as companies. Awards are funding decisions: the release does not say how many contracts were signed or theses completed. The start year (2014, the first call under the 2013-2016 State Plan) is not stated on the cited page. The release says funding doubled in 2024 and held in 2025, so the current scale is recent.

• 🇵🇭Philippineslinked to University-industry collaborationstarted 2017still operating

  CRADLE, DOST's collaborative R&D grant for academe and companies

  From 2017 the Department of Science and Technology's CRADLE programme funded 82 projects in which a university or research institute did R&D on a problem set by a Filipino company, worth 373.6 million pesos and involving 36 HEIs or RDIs and 83 companies by early 2022.

  82 projects, cradle projects funded since 2017, 2022. 373,600,000 Philippine pesos, value of cradle projects funded since 2017, 2022. Department of Science and Technology (DOST), Science for Change Program, CRADLE call for proposals for CY 2023 funding, retrieved 2026-10-02.

  How it worked: DOST pays for an R&D project only when a Filipino company names the problem, commits counterpart funding and agrees to adopt the result, and a university or research institute does the research.

  Limits: The figures are DOST's own, stated in a call for proposals rather than an audited annual report, and they are cumulative from 2017 to the date of the call (March 2022), so no single year can be read off them. The number sits inside a PDF, on page 1. A funded project is an input: the document does not say how many were completed, how many technologies the partner companies adopted, or what the companies paid as their required counterpart of at least 20 percent. 82 projects over five years is small against the size of the Philippine economy, and the partner pool includes public research institutes as well as universities. No newer official count was found: the operating status rests on DOST's undated S4CP programme page, which still describes CRADLE as being implemented, so the 2022 figures may understate the programme's current size.

• 🇲🇾Malaysialinked to University-industry collaborationstarted 2015still operating

  Public-Private Research Network (PPRN) matching grants

  Malaysia's Ministry of Higher Education approved 680 Public-Private Research Network projects involving 634 companies by the end of 2022, with a total allocation of RM27.7 million, and 469 of them had closed as successfully completed.

  680 projects, pprn projects approved, cumulative, 2022. 634 companies, companies involved in approved pprn projects, cumulative, 2022. Ministry of Higher Education Malaysia (KPT), Deputy Minister's speech at the PPRN 2.0 briefing and exhibition, 22 February 2023, retrieved 2026-10-02.

  How it worked: The ministry matches a company's own contribution, on a sliding scale by company size, to pay a university or public research institute researcher to solve a problem the company brings, and pays the researcher's fee only after the project is certified complete.

  Limits: The figures come from a ministerial speech published on the ministry's site, not from an annual report or audit, and the page is in Malay with no English translation. They are cumulative since the programme began, so they do not show the annual rate. The projects are small: RM27.7 million across 680 projects is roughly RM41,000 of public money each, and the speech describes them as demand-driven projects to raise company productivity rather than joint research of the kind the indicator's survey item imagines. The productivity gain the speech quotes comes from a 2019 study by the Malaysia Productivity Corporation that is not cited here. The page writes the amount as 'RM 27,797,149.25 juta', a typo for RM27.8 million, and the record uses the rounded RM27.7 million the same sentence gives. The start year is not stated in the speech, which places PPRN under the 2015-2025 higher education blueprint.

• 🇿🇦South Africalinked to University-industry collaborationstarted 2001operating below its peak

  TIA Technology Stations Programme, and its erosion

  South Africa's Technology Innovation Agency runs a network of university-based Technology Stations that gave 2,907 small firms science, engineering and technology services in 2024/25, against 3,167 at the five-year high in 2021/22, while the programme's budget fell from R105.9 million in 2020 to R93.98 million in 2025 and only 80 percent of the stations were operational.

  2,907 SMMEs, smmes accessing set services through the technology stations network, 2024/25. 3,167 SMMEs, smmes accessing set services through the technology stations network, five-year peak, 2021/22. Technology Innovation Agency (TIA), Annual Report 2024/25, retrieved 2026-10-02.

  How it worked: The state funds equipped technology stations inside universities, mostly universities of technology, whose staff and students sell subsidised testing, prototyping and process improvement to small firms that could not afford the equipment.

  Limits: The numbers sit inside a 188-page PDF, in the outcome table on PDF page 41 (printed page 39), and the budget figures in the Outcome 3 narrative on PDF page 42. The peak is the highest of the five years TIA reports, not a long-run maximum, and the fall from it is about 8 percent, which is computed here. The yearly count is noisy (1,990, 3,167, 2,671, 3,114, 2,907), and 2024/25 sits above two of the five years, so the erosion rests on what TIA reports about capacity: the budget cut, a fifth of the stations not functioning and no new high-end equipment, rather than on the count alone; the budget cut is in nominal rand and larger in real terms. A count of firms served measures university-to-firm technical service, testing and prototyping, which is a thinner form of research collaboration than joint projects: the same report says 73 percent of station equipment time goes to firms. The agency also reports a separate SMME figure (3,775) across all its programmes; this record uses only the Technology Stations line. A 2022 National Treasury spending review recommended moving the programme out of TIA altogether. The start year is taken from the programme's establishment under the former Department of Science and Technology and is not stated in the cited report.

• 🇨🇱Chilelinked to University-industry collaborationstarted 2004still operating

  CORFO Programas y Consorcios Tecnológicos

  Chile's development agency CORFO had 49 technology programmes and consortia in execution during 2024, 38 technology programmes and 11 technology consortia, the consortia joining firms with universities and technology centres in multi-year R&D portfolios, after awarding six new consortia in 2023 with 1,500 million pesos of co-financing each.

  49 programmes and consortia, technology programmes and consortia in execution, 2024. 6 consortia, new technology consortia funded in the 2023 call, 2023. Corporación de Fomento de la Producción (CORFO), Dirección de Programas y Consorcios Tecnológicos, results presentation at the VI Encuentro de Programas y Consorcios Tecnológicos, January 2025, retrieved 2026-10-02.

  How it worked: CORFO co-finances long-lived consortia and programmes, for up to six years per award, at about 30 percent of cost for new consortia, inside a stated horizon of more than ten years, in which firms in one sector and universities or technology centres share a portfolio of R&D projects and the intellectual property they produce.

  Limits: The count is CORFO's own, from a slide deck hosted on CORFO's asset store, not an audited report, and it is a stock of initiatives in execution, including six new and ten finished in the year, not a flow of research output. The figure sits inside a PDF (page 13, footnote; the 38 and 11 split on page 9). Not every technology programme is led by a university: several of the 38 programmes are executed by a single company with no university partner named in the deck, so only the 11 consortia are collaborative by design and the 49 overstates university-firm collaboration. The second metric comes from a separate Ministry of Economy press release of 12 December 2023 announcing the 2023 awards, which is an award notice rather than evidence of delivery. The deck also gives figures for participating firms and private co-funding whose layout in the PDF does not make clear which label each number belongs to, so they are not used. The instrument began as the 2004 Consorcios Tecnológicos Empresariales and has been redesigned since; that start year is not stated in either cited source.

• 🇺🇸United Stateslinked to University-industry collaborationstarted 2012still operating

  Manufacturing USA innovation institutes

  In fiscal 2023 the 16 reporting Manufacturing USA institutes, public-private consortia sponsored by the Departments of Commerce, Defense and Energy, managed 929 applied research and development projects and held 2,921 memberships, 655 of them universities, community colleges and other academic members alongside 1,315 small and 493 large manufacturers.

  929 projects, ongoing applied research and development projects across the institutes, FY 2023. 655 memberships, academic members (universities, community colleges, etc.) of the institutes, FY 2023. National Institute of Standards and Technology, Advanced Manufacturing National Program Office, Manufacturing USA Report to Congress 2025 (covering FY 2023), retrieved 2026-10-02.

  How it worked: Federal agencies fund a standing consortium per manufacturing technology for a fixed term; firms, universities and colleges pay membership and cost share to sit on a shared applied research and workforce agenda that no single member would fund alone.

  Limits: The figures are the institutes' own reports aggregated by the program office, in Table 2 on PDF page 13 (printed page 11) of the report, and they cover fiscal 2023 although the report was posted in March 2026, so they are about three years old. Sixteen of the 17 institutes report; EPIXC joined mid-year and is excluded. The counts are memberships, not distinct organisations: a university that belongs to three institutes counts three times. The academic count mixes research universities with community colleges, which join mainly for workforce training, and the report does not say how many of the 929 projects actually pair a university with a firm; some are workforce or firm-only projects. Federal base funding was 160.4 million dollars in the year against 379.5 million of cost share and other funding, all as reported by the institutes, with no counterfactual for what members would have done anyway. The start year is the first institute, America Makes, established August 2012 per the report's Appendix B; the program's statute is the Revitalize American Manufacturing and Innovation Act of 2014.

• 🇨🇦Canadalinked to University-industry collaborationstarted 2003still operating

  Mitacs Accelerate research internships

  Mitacs, the federally and provincially funded not-for-profit, placed 4,859 students, recent graduates and postdoctoral fellows in Accelerate research internships with 2,314 host organisations under 2,406 professors in 2024-25, out of 20,885 internships across all its programmes that year.

  4,859 interns, interns in the accelerate program, 2024-25. 2,314 organisations, host organisations in the accelerate program, 2024-25. Mitacs, 2024-2025 Annual Report for Innovation, Science and Economic Development Canada, retrieved 2026-10-02.

  How it worked: A national intermediary splits the cost of a graduate student's or postdoc's research project inside a firm with the firm, using federal and provincial money, and uses the student's supervisor as the university link.

  Limits: This is Mitacs reporting on itself to its federal funder: the Accelerate counts are on PDF page 28 and the 20,885 total on PDF page 2. Mitacs is a not-for-profit rather than a government body; ISED funding was 48 percent of its income in 2024-25 and the rest came from provinces and partners. Host organisations include not-for-profits, hospitals and municipalities as well as firms, so 2,314 overstates the firms involved. An internship is a funded unit of four to six months and one intern can hold several, so internships and interns are different counts; the 20,885 total also includes Globalink internships that bring international students to Canadian labs with no firm involved. The next report (2025-26, July 2026) gives 20,819 internships in total, against 21,432 in 2022-23. The capacity is now shrinking by plan: in 2024-25 Mitacs capped Accelerate units per university for lack of federal money, and its 2026-27 corporate plan (PDF page 9) foresees 12,540 internships on a 120 million dollar ISED grant, against more than 200 million of federal support in 2022-23. If that plan is delivered this record becomes an erosion. The 2003 start year, the first Mitacs research internship program, comes from ISED's evaluation of its funding to Mitacs, not from the cited report.

• 🇯🇵Japanlinked to University-industry collaborationstarted 2020still operating

  COI-NEXT, co-creation centres for industry, academia and government

  Japan Science and Technology Agency ran 42 COI-NEXT co-creation centres in April 2023, 30 full-scale and 12 in a development stage, most led by universities and some by national research institutes, which in fiscal 2022 counted 1,073 participating organisations, 256 universities and research institutes and 817 companies and others, with 2,566 participants from the company side.

  817 organisations, companies and other non-academic organisations participating in coi-next centres, FY 2022. 42 centres, coi-next centres supported (30 full-scale, 12 development-stage), 2023-04-01. Ministry of Education, Culture, Sports, Science and Technology (MEXT), Science and Technology Policy Bureau, briefing for the 2023 autumn administrative project review (JST operating grant for co-creation programmes), retrieved 2026-10-02.

  How it worked: JST funds a university-led centre for up to ten years around a stated vision for society, requires the university's leadership to commit and firms to contribute resources, and steps down its funding in the final years so the centre must find its own.

  Limits: The source is the funding ministry's own briefing, written to defend the programme's budget before a government spending review, and the participant counts are its self-reported totals. The counts are on PDF page 13 (printed page 12) and the centre counts on PDF page 10 (printed page 9). The organisation counts are for fiscal 2022 alone and are now more than three years old. The category is 'companies and others' (企業等), which can include municipalities and other non-academic bodies, so 817 is an upper bound on firms. The ministry's claim that external income of about 15.4 billion yen already exceeds about 14 billion yen of support counts resources the firms provided at their own valuation. The 2020 start year is inferred from the heading 'activities FY2020 to FY2022' on the same page. MEXT's fiscal 2026 budget sheet plans support for about 37 centres in the main fields, fewer than 42, which partly reflects centres graduating from support.

• 🇧🇷Brazillinked to University-industry collaborationstarted 2018still operating

  MAI/DAI, innovation scholarships for research with partner companies

  Brazil's research council CNPq, with CAPES co-funding the 2024 call, has run four calls of its Mestrado e Doutorado Acadêmico para Inovação programme (2018, 2020, 2022 and 2024), in which about 80 science and technology institutions and more than 200 companies had taken part and at least 5,000 scholarships had been granted by 2024 for undergraduate and graduate student research on projects defined with partner firms.

  5,000 scholarships (at least), scholarships granted across mai/dai calls, 2018 to 2024, cumulative. 200 companies (more than), companies that have taken part in mai/dai, 2018 to 2024, cumulative. Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), news release on the preliminary result of the first round of the 2024 MAI/DAI call, retrieved 2026-10-02.

  How it worked: The national research council pays scholarships for students whose thesis or initiation project is a research project defined with a partner firm, run through a university that must have a technology transfer office and an innovation policy, and since 2024 asks the firm for a financial counterpart.

  Limits: (append) The 2018 start year is also stated on the cited page, which lists the earlier calls as 2018, 2020 and 2022.

• 🇳🇱Netherlandslinked to University-industry collaborationstarted 2006dismantled

  NWO Perspectief, ended by the 2026 spring budget

  NWO's Perspectief programme, funded mainly by the Ministry of Economic Affairs, awarded 12 public-private technology consortia in 2025 with 46.4 million euros of subsidy, each joining at least three knowledge institutions with at least four companies or other users paying at least 30 percent of costs; the ministry's own page records that the last Perspectief round started in July 2026.

  12 consortium projects, perspectief projects awarded, 2025. 46,400,000 euros, perspectief subsidy awarded, 2025. Ministerie van Economische Zaken, Bedrijvenbeleid in beeld, NWO Perspectief programma (data from NWO), retrieved 2026-10-02.

  How it worked: The research council runs a yearly competition for multi-year consortia in which universities receive the subsidy and companies and civil-society users co-fund in cash and kind, sit on a user committee and get first access to the results.

  Limits: The counts are NWO's awards as published by the ministry, and an award is a project starting, not a result: projects run for up to six years, so awarded consortia will continue until the early 2030s although no new ones will be funded after the 2026-27 round. 2025 was an unusually large year: the same table gives 4 projects in 2024 and 5 to 9 in most years since 2016, so 12 is the peak, not the norm. The page's table of participating companies has two columns whose labels do not make clear which is total and which is SMEs (70 and 79 in 2025), so company counts are not used. The termination is stated on the cited page only as 'the last round started in July 2026'; the reason, the end of ministry funding in the first supplementary budget to the 2026 Spring Memorandum, is in NWO's news release of 2 April 2026, which also states NWO's objection. The final round has 26.5 million euros by NWO's account. The start year is approximate: NWO's release says the programme has run for twenty years, and the ministry's series begins in 2016.

• 🇬🇧United Kingdomlinked to Public-private collaborationstarted 1992dismantled

  Private Finance Initiative and PF2, and their abolition

  HM Treasury's Private Finance Initiative signed on average 55 contracts a year between 1997 and 2010 and left 704 current PFI and PF2 projects with a capital value of £57 billion at 31 March 2018, after signings fell to an average of nine a year after May 2010 and the Chancellor announced at Budget 2018 that the government would no longer use PF2 for new projects.

  704 projects, current pfi and pf2 projects (in operation or construction), 2018-03-31. 55 contracts per year, average pfi contracts signed per year, 1997 to 2010 (peak era; 9 a year after may 2010), 1997-2010. HM Treasury and Infrastructure and Projects Authority, Private Finance Initiative and Private Finance 2 projects: 2018 summary data, retrieved 2026-10-02.

  How it worked: HM Treasury let departments and local bodies buy hospitals, schools, roads and prisons as long design-build-finance-operate contracts paid by performance-linked unitary charges, keeping the capital cost off the up-front budget; after value-for-money criticism it reformed the model as PF2 in 2012 and then stopped using it for new central-government projects in 2018.

  Limits: This is a dismantling by decision of the procurement route, not of the assets: the 704 contracts keep running, with £188.35 billion of forecast unitary charges over the next 30 years, so the record shows a national collaboration model closed to new deals rather than a lost stock of hospitals and schools. The decline started well before abolition (an average of 55 signings a year from 1997 to 2010 against nine a year after May 2010), and PF2, introduced in 2012 'in response to widespread concerns about value for money', was used only six times. A hostile reviewer would say PFI is evidence of joint delivery at scale and of poor value for money at once: the scale figure says nothing about whether the risk transfer paid off, and the Carillion liquidation hit two PF2 projects. Coverage is departments and devolved administrations under standard PFI and PF2 terms only; other PPP forms such as NHS LIFT are excluded, and Scotland and Wales run their own models that the 2018 announcement did not affect. The 1992 start year is not stated in the document, whose earliest year is 1997. The figures are in a PDF: the 55-a-year and nine-a-year averages and the PF2 announcement on PDF page 5 (paragraphs 1.5 and 1.6), the 704 projects and £57 billion on PDF page 8 (paragraphs 2.2 and 2.3). The publication is HM Treasury reporting on its own programme.

• 🇯🇵Japanlinked to Public-private collaborationstarted 1999still operating

  PFI Act projects and concessions

  Under the 1999 PFI Act, Japanese national and local governments had published implementation policies for a cumulative 1,219 PFI projects by 31 March 2026, 80 of them public facility operating concessions, with a cumulative initial contract value of ¥10.47 trillion (104,730 hundred-million yen).

  1,219 projects, cumulative pfi projects with a published implementation policy since fy1999, 2026-03-31. 104,730 hundred million yen (oku-en), cumulative pfi contract value (initial contract amounts), 2026-03-31. Cabinet Office of Japan, PPP/PFI Promotion Office, PFI事業の実施状況（令和7年度）, retrieved 2026-10-02.

  How it worked: The national government legislated a standard route for private finance, design, construction and operation of public facilities, and a Cabinet Office promotion office published guidelines, case collections and annual counts, so that municipalities could procure schools, waste plants and civic buildings through PFI and, from 2011, grant operating concessions for airports and water works.

  Limits: The count is of projects whose implementation policy was published, which the Cabinet Office tracks by its own survey; a published policy is not a delivered asset, although the note says projects cancelled during service or abandoned before service began are excluded. Most projects are small and local: the sectoral table in the same annex shows local governments sponsoring the large majority (for example 397 of 454 culture and education projects), so the record evidences a widely used procurement route rather than large national infrastructure. The contract value covers initial contract amounts where a public payment was fixed and excludes concession fees, and the annex itself says it differs from the project-scale measure in the 2026 PPP/PFI Action Plan. Nothing here shows value for money or service quality. The figures are read from charts and a table in a PDF (project counts on page 1, contract values on page 2, concessions on page 3, sector table on page 4); the summary page at https://www8.cao.go.jp/pfi/pfi_jouhou/pfi_joukyou/pfi_joukyou_r7.html states 1,219 and 80 in text. The Cabinet Office is the programme's own promoter.

• 🇩🇪Germanylinked to Public-private collaborationstarted 1969operating below its peak

  Dual vocational training (duale Ausbildung), and its contraction

  Germany's dual training system, in which firms employ apprentices under contracts registered by the chambers while state vocational schools teach them, recorded 475,950 new training contracts in the year to 30 September 2025, down from 569,379 in 2011, while 54,372 offered places stayed unfilled.

  475,950 contracts, new dual training contracts (bibb survey, 1 october to 30 september), 2025-09-30. 569,379 contracts, new dual training contracts at the series peak (2009 to 2025), 2011-09-30. Bundesinstitut für Berufsbildung (BIBB), Erhebung neu abgeschlossene Ausbildungsverträge zum 30.09., Tabelle 60 (Stand 05.12.2025), retrieved 2026-10-02.

  How it worked: Federal law sets training regulations agreed with employers and unions, firms recruit and pay apprentices and deliver workplace training, chambers register contracts and run examinations, and the Länder run part-time vocational schools, so qualification is a joint product that no single party can deliver alone.

  Limits: The fall of about 16 percent from the 2011 high is computed from two published values (column 1). How much of it is the collaboration weakening is contested. On the applicant side, unfilled training places rose from 17,139 in 2009 to a high of 73,431 in 2023, which points to fewer and differently choosing school leavers, demography and the shift to university. On the employer side, the same table shows the places offered (column 6) falling from 598,935 in 2011 to 530,322 in 2025, about 11 percent, including a drop of 25,338 in 2025 alone, while unfilled places fell to 54,372 and applicants left without a place or an alternative (column 3) rose to 38,784, the highest in the series; the latest year reads as firms offering fewer places, not only as missing applicants. A hostile reviewer would also say the system still enrols nearly half a million new trainees a year, the largest joint public-private training delivery in the corpus. The table does not show the share of firms that train, which is the direct measure of employer participation. The 2020 figure (467,484) is a pandemic low and 2025 sits above it. The series starts in 2009, so earlier and higher levels are not covered, and the 1969 start year (the Vocational Training Act) is not stated in the source. Figures are read from a two-page PDF table, Bundesgebiet block. BIBB is a federal institute that runs the survey with the chambers, the bodies that register the contracts.

• 🇦🇺Australialinked to Public-private collaborationstarted 2018still operating

  New Payments Platform, the shared fast-payment rail

  The New Payments Platform, whose payments the Reserve Bank of Australia settles one by one in central bank money through the Fast Settlement Service of its RITS system, settled a daily average of 3.6 million transactions in 2024/25, almost 19 percent more than in 2023/24, worth A$4.7 billion a day.

  3,600,000 transactions per day, fast settlement service (npp) daily average transactions settled, 2024/25. 4,700,000,000 Australian dollars per day, fast settlement service (npp) daily average value settled, 2024/25. Reserve Bank of Australia, Annual Report 2025, retrieved 2026-10-02.

  How it worked: The Reserve Bank set real-time payment objectives after its 2012 strategic review and built the Fast Settlement Service, while banks jointly financed, built and governed the switching platform and overlay services, so that every participating account could send and receive in seconds around the clock.

  Limits: The figure counts settlements in the Reserve Bank's Fast Settlement Service, which exists to settle NPP payments; it is a volume measure, not evidence of reach, outage resilience, fraud losses or how much activity is additional to existing direct entry. The annual report shows the Reserve Bank's half of the arrangement (settlement in RITS, fifteen institutions settling NPP payments directly) but does not say who built, owns or governs the NPP itself; it names NPP Australia Limited only in its glossary, so the industry side of the partnership rests on sources outside this document. The NPP is still the smaller rail for account-to-account payments: the report's Graph 2.3.4 shows bulk electronic clearing (BECS) still carrying more daily volume, and the industry's proposed decommissioning of BECS, which the report discusses, is not yet done. The 2018 launch year is not stated in the annual report. The number is on page 99 of the PDF. The rba.gov.au host refuses scripted requests (Akamai 'Access Denied'); the PDF was read through WebFetch at the same URL. The Reserve Bank reports on a service it owns and operates.

• 🇳🇬Nigerialinked to Public-private collaborationstarted 2011still operating

  NIBSS Instant Payment (NIP)

  NIBSS Instant Payment, run since 2011 by the interbank settlement company jointly owned by the Central Bank of Nigeria and the banks, grew from 2 billion transactions in 2020 to 11 billion in 2024, with values rising from US$457 billion to US$1.1 trillion, and had 57.7 million unique users.

  11,000,000,000 transactions, nip transactions in the year, 2024. 57,700,000 users, nip unique users (52 percent of adults), 2024. AfricaNenda Foundation, The State of Inclusive Instant Payment Systems in Africa (SIIPS 2025), NIP Nigeria case study, retrieved 2026-10-02.

  How it worked: The Bankers' Committee, the forum of the central bank and the deposit-taking banks, created a jointly owned switch in 1992 and used it to launch a 24/7 account-to-account instant payment service in 2011, with the central bank setting rules and fee caps and every bank connected, so a transfer reached any account in seconds.

  Limits: The publisher is a non-profit foundation, not a statistical agency, and the case study does not name the source of its 2020 to 2024 series ('NIP data for 2020 to 2024 shows'); it is presumably NIBSS operator data, but that is an inference. The Central Bank of Nigeria's e-payment statistics page reports channels (internet, mobile app, USSD transfers) without naming NIP, and NIBSS's own pages carry press reprints that mix NIP and all e-payment totals, so no official NIP series was citable. The figures are rounded (2 and 11 billion), the 2024 value is printed as $1.1 trillion in the text and $1,171,642 million in the chart, and the dollar values are converted at unstated rates in a period of steep naira devaluation, so the value growth says little in real terms. Growth was pushed by the 2023 cash scarcity and the pandemic as much as by the rail; the same page reports 69 percent of transaction volumes from men, a large inclusion gap, and the 57.7 million users are not active-user counts. Ownership and governance (jointly owned by the CBN and banks, a board of a CBN deputy governor and bank chief executives) and the 2011 launch are stated in the same document. Figures are in a PDF, on PDF page 8 (printed page 9) and the chart that follows.

• 🇵🇭Philippineslinked to Public-private collaborationstarted 2018still operating

  InstaPay under the National Retail Payment System

  InstaPay, the real-time automated clearing house organized under the Bangko Sentral ng Pilipinas' National Retail Payment System framework, in which banks and e-money issuers take part, processed 4.7 billion transfers worth ₱11.6 trillion in 2025, up from 232.9 million in 2020, with 95 participating institutions in July 2026.

  4,700,000,000 transactions, instapay transactions in the year, 2025. 11,600,000,000,000 Philippine pesos, instapay transaction value in the year, 2025. Bangko Sentral ng Pilipinas, Payments Bulletin (data as of July 2026), retrieved 2026-10-02.

  How it worked: The central bank set an interoperability framework requiring retail payment providers to join industry-run automated clearing houses, while banks and e-money issuers governed the clearing rules through a self-regulatory payment management body, so any account or wallet could pay any other in real time.

  Limits: Transaction counts show a heavily used rail, not who uses it: a large part of the 2025 jump (230.9 percent in volume against 57.3 percent in value) is consistent with small e-wallet cash-ins and transfers, so value per transfer fell, and the bulletin does not split person-to-person use from wallet top-ups. The participant count (87 sender-receivers and eight receive-only) is computed from the bulletin's table. The bulletin lists InstaPay among the payment systems organized under the National Retail Payment System and lists its participants by type, but does not say who operates or governs the clearing house, so the industry side of the arrangement rests on sources outside this bulletin. The 2018 start year is not stated in the bulletin, whose annual table begins in 2020. The central bank's 2024 digital payments report puts digital at 57.4 percent of retail payment volume, so cash still carries a large share. The bulletin is a slide deck PDF: the annual table is on page 6 and the participant table on page 5. bsp.gov.ph refuses scripted requests (Akamai 'Access Denied'); the PDF was read through WebFetch at the same URL. The central bank reports on a framework it regulates.

• 🇪🇪Estonialinked to Public-private collaborationstarted 2002still operating

  National eID ecosystem with private trust-service partners

  Estonia's state eID ecosystem, in which the state issues ID cards with certificates from a private trust-service partner and accepts the privately issued Smart-ID in its authentication service TARA, handled 27,094,512 Smart-ID authentications through TARA in 2023, 57 percent of the total, beside 22,351,282 by state ID card and Mobile-ID.

  27,094,512 authentications, smart-id authentications through the state authentication service tara, 2023. 22,351,282 authentications, id card and mobile-id authentications through tara, 2023. Riigi Infosüsteemi Amet (Estonian Information System Authority), Identiteedihalduse ja eID valge raamat 2024, retrieved 2026-10-02.

  How it worked: The state made a chip ID card with authentication and signing certificates compulsory and built common authentication and signing services, while a private certification provider issued the certificates, mobile operators issued Mobile-ID SIMs and banks relied on the same credentials, so one identity layer served public and private services and later absorbed a bank-led smartphone credential.

  Limits: The counts are authentications through one state service, TARA, not all eID use (banks authenticate customers directly, outside TARA). The split shows a shift from state-issued means to a private credential rather than growth of joint delivery: the white paper reports that ID card and Mobile-ID signatures fell from 94.8 percent of digital signatures in 2020 to 76 percent in 2022 as Smart-ID was enabled, and state-means signatures through DigiDoc fell from 8.6 million in 2021 to 6.9 million in 2023. A hostile reviewer could read that as the state losing control of its identity layer to a private provider. The document is a strategy white paper, not a statistical release, and draws on the Police and Border Guard Board and SK ID Solutions for issuance data. The 2002 start year (first ID cards) is not stated in the source. Figures are in a PDF on PDF pages 22 and 23 (printed pages 20 and 21), text and Tables 1 to 4.

• 🇫🇮Finlandlinked to Public-private collaborationstarted 2017still operating

  Finnish Trust Network for strong electronic identification

  From May 2017 Finland's Trust Network obliged banks and mobile operators issuing strong electronic identification to let brokers resell their authentications at a regulated maximum price, cut to 3 cents in 2019, and mediated authentication events rose from an estimated 20 million in the second half of 2019 to about 60 million in the first half of 2020.

  60,000,000 authentication events in the half-year, authentication events mediated through trust network brokers (estimate), 2020-H1. 3 euro cents, maximum regulated wholesale price per mediated authentication (from 2019), 2020. Traficom (Finnish Transport and Communications Agency) and Finnish Competition and Consumer Authority, Sähköisen tunnistamisen markkinat, Traficomin tutkimuksia ja selvityksiä 2/2021, retrieved 2026-10-02.

  How it worked: Parliament and the communications regulator turned bank and mobile-operator credentials into shared national infrastructure by law: providers must sell authentications to licensed brokers at a capped wholesale price, so any public or private e-service can accept every strong credential through one contract, and the state's own Suomi.fi service buys them the same way.

  Limits: The volume figures are the regulators' own estimates ('arviolta') built from provider returns that some providers could not separate, and the report says it could not trace a multi-year series; for 2019 it gives only a range of 40 to 80 million. The figures are from 2020 and published in March 2021, so they are dated, and the first half of 2020 overlaps the first months of the pandemic, which also pushed public and private services online. The collaboration was regulatory compulsion as much as partnership: the report says identification providers held at least a weak monopoly over their own credentials before price caps, and the network only worked from 2019 after repeated legislative fixes, with 2020 the first full year of comprehensive brokerage. Bank credentials made up an estimated 85 to 90 percent of mediated events in late 2019, so the system depends on a few banks. The government is building a state digital identity alongside it. Figures are on PDF pages 79 and 80 (printed pages 78 and 79), section 6.5.

• 🇧🇴Bolivialinked to Public-private collaborationstarted 2021still operating

  Interoperable QR immediate payments, built by the banks and the Banco Central de Bolivia

  Bolivia's banks first offered QR payments under their own SIMPLE standard, and the Banco Central de Bolivia then made QR transfers interoperable across banks, cooperatives, housing lenders and development finance institutions with a universal standard compatible with the banks' codes, mandatory electronic channels and free transfers; immediate QR payments rose from 6 million operations in 2021 to 892 million in 2025, 86 percent of electronic fund transfer orders that year.

  892 million operations, immediate payments by qr code processed in the year, 2025. 357,747 million bolivianos, nominal, value of immediate payments by qr code in the year, 2025. Banco Central de Bolivia, Informe de Vigilancia del Sistema de Pagos 2025, retrieved 2026-10-02.

  How it worked: The banking system launched its own QR standard, and the Banco Central de Bolivia then built a universal QR standard compatible with it into its settlement module, required every financial institution to offer electronic channels and free transfers, and settled QR payments through its module and the private clearing houses, so each institution's app could pay any account in the system.

  Limits: Both numbers are data labels on Gráfico 17 of a 72-page PDF (PDF page 35, printed page 29), with the 86 percent share on Gráfico 18 on the next page; there is no machine-readable series. The central bank is the regulator, the standard-setter and the reporter, so this is the system's owner grading its own system. The collaboration mixes a private standard with a public mandate: the banking system built SIMPLE, the BCB built a compatible universal standard and imposed interoperability and free transfers, and private institutions and the clearing houses ACCL and UNILINK carried it out. The 2025 jump coincided with Bolivia's dollar shortage, fuel queues and rising inflation, which push people off cash; the value series is in nominal bolivianos and overstates real growth in a year of high inflation. A payment count does not show inclusion: the report gives no figure for first-time account holders. The series counts every immediate QR payment processed through the clearing houses and the BCB's settlement module, not only codes generated under QR BCB Bolivia: the banks' own SIMPLE QR standard came first, the BCB's 2021 vigilance report describes the BCB's compatible QR standard in its settlement module as due in the last quarter of 2022, and the BCB's QR page (https://www.bcb.gob.bo/?q=pagos_qr_bcb_bolivia) calls QR BCB 'un solo estándar universal compatible con todos los QR implementados por el sistema financiero'. The 2021 figure therefore predates the BCB standard, and the start year is the first year of the series, not the launch of either standard.

• 🇩🇴Dominican Republiclinked to Public-private collaborationstarted 2021operating below its peak

  Plan Nacional de Viviendas Familia Feliz, and its intake freeze

  By May 2024 the Dominican government's Familia Feliz plan, in which private developers build low-cost homes and the state pays down-payment, tax and mortgage-rate subsidies, had built 3,979 homes and assigned RD$1,108 million in subsidies.

  3,979 housing units, homes built under the plan since launch, 2024-05. 1,108 RD$ million, subsidies assigned (bono inicial, bono itbis and bono tasa 7 años), 2024-05. Ministerio de la Presidencia de la República Dominicana, press note of 13 May 2024, retrieved 2026-10-02.

  How it worked: The Presidency, and since Decreto 114-25 the housing ministry, ran a national trust that certifies private housing projects and pays each buyer's down payment, sales tax and seven years of mortgage-rate relief, so private developers and banks build and finance low-cost homes that the state does not build itself.

  Limits: The numbers come from a ministry press note quoting the minister, not from a statistical series, and the government is reporting on its own flagship. 'Built' is not 'handed over': the same note says 704 apartments were delivered in January to May 2024 and 934 were waiting to be inaugurated. It also promised 9,600 deliveries in 2024, and no official page opened today gives a cumulative delivered total to check that against; Diario Libre (12 May 2026) reports the housing ministry putting Familia Feliz deliveries at 5,922 units since 2020, which, if right, means the 2024 promise was badly missed. That figure is press only and is not used. The programme's intake has stopped: FONVIVIENDA's news item of 26 April 2026 (https://fonvivienda.gob.do/2026/04/26/ministro-ito-bisono-reafirma-continuidad-del-programa-familia-feliz-entrega-144-apartamentos-en-laderas-del-este-vf/) says projects approved up to 30 December 2025 continue unchanged, that applications made up to that date are being re-evaluated, and that the focus is on finishing and handing over projects approved up to 2025, and it names no intake after that date; the status is eroded for that reason, although deliveries from the pipeline continue (890 apartments across Familia Feliz and Mi Vivienda in the first half of 2026, per FONVIVIENDA, not split by programme). A few thousand homes is small against the national housing deficit. The private side is paid by subsidy and mortgage, so this is a subsidy-for-supply model; banks still screen buyers through pre-approval, which excludes the informal poor. The start year is taken from Decreto 191-21, which created the FONVIVIENDA trust.

• 🇮🇩Indonesialinked to University-industry collaborationstarted 2021still operating

  Matching Fund Kedaireka university-industry matching grants

  Indonesia's Directorate General of Higher Education, Research and Technology funded 427 university-industry collaboration proposals through the Kedaireka Matching Fund in 2021, its first year, and 1,093 in 2022.

  1,093 proposals, matching fund proposals funded, 2022. 427 proposals, matching fund proposals funded in the first year, 2021. Directorate General of Higher Education, Research and Technology (Ditjen Diktiristek), news release 'Matching Fund-Kedaireka Akselerasi Riset dan Inovasi Perguruan Tinggi', 13 August 2023, as archived by the Internet Archive on 29 April 2025, retrieved 2026-10-02.

  How it worked: The higher-education directorate built an online matchmaking platform (Kedaireka) where university staff and firms register a partnership, then paid public money to match what the partner commits, so the state funds only collaborations a partner has already put resources into.

  Limits: The source is the ministry's own news release, and the cited URL is an Internet Archive copy: the original host (dikti.kemdikbud.go.id) no longer answers, and the successor ministry's copy on kemdiktisaintek.go.id sits behind a Cloudflare human-verification check that this protocol does not get past. The release is self-reporting, not an annual report or audit. The counts are funded proposals, not distinct firms or universities, and the release gives no count of partner firms. Partners can be businesses, industry or government bodies ('pihak bisnis, industri, maupun pemerintahan'), so not every funded proposal is a university-firm pairing. The figures show money granted, not research or commercial results. A separate ministry release of 15 September 2022 (same archive) says 5,407 proposals were submitted in 2022 against 1,231 in 2021, so roughly one in five 2022 proposals was funded; it also reports Rp 11.2 trillion of combined ministry and partner 'collaboration funds' for 2022, which mixes partner and possibly in-kind contributions and is not used here. No funded counts for 2023 to 2025 were found at an open official source. The programme was renamed Dana Padanan for 2024 and passed to the new Ministry of Higher Education, Science and Technology after the October 2024 restructuring; the programme site (backoffice.kedaireka.id) lists a 2025 call and a July 2025 award announcement, which is the basis for 'operating', but not the scale of those rounds, and the 2025 budget cuts to the ministry may have reduced it.

• 🇻🇳Vietnamlinked to Public-private collaborationstarted 2020operating below its peak

  North-South eastern expressway BOT sections, and the retreat to public investment

  Vietnam built three component projects of the 2017-2020 North-South eastern expressway phase, 178 km at about VND 25,343 billion, as BOT public-private partnerships, then financed all ten 2021-2025 component projects on the Hanoi to Ho Chi Minh City stretch as public investment, and in June 2025 the Ministry of Construction reported that it had asked the Prime Minister to keep widening the corridor with public investment money, leaving PPP as a later option once tolls were collected.

  0 of 10 component projects (all public investment), 2021-2025 phase component projects on the hanoi to ho chi minh city stretch procured as ppp, 2025-06. 3 component projects (178 km, about VND 25,343 billion), 2017-2020 phase component projects built as bot, the peak, 2017-2020 phase. Government of Viet Nam portal (baochinhphu.vn), report of the Ministry of Construction's study on widening the North-South eastern expressway, 25 June 2025, retrieved 2026-10-02.

  How it worked: The transport ministry split the corridor into component projects and offered some of them to BOT investors, with state capital covering more than half of each project's cost and tolls repaying the private share. For the next phase the National Assembly financed every section from the budget, with the state to recover its outlay through tolls, which took private investors out of the corridor.

  Limits: The figures are the Ministry of Construction's own account, relayed by the Government's e-newspaper; no audit body has published them. The zero is read off the page, not printed as a number: the ministry lists 11 public-investment component projects, 10 of them from the 2021-2025 phase, against 3 BOT projects from 2017-2020. The full 2021-2025 phase was 12 component projects of about 729 km at VND 146,990 billion, all assigned to public investment by National Assembly Resolution 44/2022/QH15 (scanned PDF at datafiles.chinhphu.vn/cpp/files/vbpq/2022/01/44-qh.signed.pdf: 729 km in 12 component projects as public investment on page 1, Article 2.2; VND 146,990 billion on page 2, Article 2.5). The peak was itself a retreat from a plan. The Ministry of Transport listed eight sections for PPP in December 2017 (Decision 3535/QD-BGTVT, baochinhphu.vn, 26 December 2017), and only three reached contract, out of 11 component projects over 654 km. Most of the money in those three came from the state. The Government portal of 28 April 2024 puts Cam Lam-Vinh Hao at VND 8,925 billion with about 5,139 billion of state capital, and Dien Chau-Bai Vot at 11,157 billion with about 6,067 billion. The same page records the opening of Cam Lam-Vinh Hao and of Dien Chau-Bai Vot as far as National Highway 46B, with Nha Trang-Cam Lam already in service; the Ministry of Transport announced the full Dien Chau-Bai Vot opening for 30 June 2024 (baochinhphu.vn, 28 June 2024). Construction started in 2020-2021, per the Ministry of Transport on the 2024 page. The retreat is a choice as much as a market failure: the same 2025 page says the ministry's own study found BOT widening highly feasible, and it chose public investment, citing toll levels and the push to disburse public funds. This is one corridor. The record does not show whether PPP procurement fell on other expressways. The question is still open: a pre-feasibility report of June 2026 (baochinhphu.vn, 6 June 2026) proposes public investment for 15 component projects and PPP with state support only for the three existing BOT sections. That is a proposal, not a decision.

• 🇺🇦Ukrainelinked to Public-private collaborationstarted 2023still operating

  Brave1 defence technology cluster

  Ukraine's government set up the Brave1 cluster in 2023 under the Ministry of Digital Transformation. By September 2026 it had awarded more than 1,000 non-repayable grants worth UAH 6.2 billion to private defence-technology developers, whose products it tests with the armed forces and moves toward serial production.

  6.2 UAH billion, defence technology development funded by brave1 since 2023, 2026-09. 1,000 grants (published as more than 1,000), grants awarded to ukrainian defence developers since 2023, 2026-09. Brave1 defence innovation cluster (state institution under the Ministry of Digital Transformation of Ukraine): news release, retrieved 2026-10-02.

  How it worked: The Ministry of Digital Transformation, with the Ministry of Defence and the General Staff, set up a state cluster. It registers private developers, funds them through tiered grants, arranges battlefield testing and military feedback, and steers money toward priority areas the military defines.

  Limits: The source is the cluster reporting on itself in a news release, with a minister's quote. Neither the 6.2 billion nor the grant count is broken down by year or audited. The figures measure money and grants going out, not what was delivered: they do not show how many funded products reached serial production or the front. The impact claims on the page, a 1.7-point gain in technology readiness and 79 percent of firms saying they had no alternative, come from a survey of defence-tech manufacturers in a study by the BRDO Office for Effective Regulation and are self-reported. Grants are one-way subsidies, so the record evidences public funding of private developers plus joint testing, which is a thin reading of joint delivery. The standard grant programme accounts for UAH 2.3 billion of the total; most of the rest, UAH 3.7 billion, went to 34 competition projects for missiles, air defence and autonomy, so the money is concentrated in a few large awards. Wartime urgency and secrecy limit what can be verified.

• 🇷🇺Russialinked to University-industry collaborationstarted 2010still operating

  Government Resolution 218 complex projects for high-tech production

  Under Government Resolution 218 of 9 April 2010, Russia's Ministry of Science and Higher Education subsidised 41 joint firm-university complex projects in their R&D stage in 2024, from the 14th and 15th selection rounds. Federal funding was 2.708 billion rubles, and firms put in 4.14 billion rubles of their own (156 percent of the budget funds). The ministry reports 161 technologies developed and handed over to production by January 2025.

  41 projects, complex projects receiving state support in the year (14th and 15th selection rounds), 2024. 4.14 billion rubles, extrabudgetary funds raised from participating firms (156% of budget funds), 2024. Ministry of Science and Higher Education of the Russian Federation, report on the ministry's results for 2024 (June 2025), retrieved 2026-10-02.

  How it worked: The Russian government subsidises the manufacturer rather than the university. A firm wins a competitive subsidy to commission R&D from a Russian university or state research institute, co-funds it at least one for one, and must build the production line and sell the product from its own resources.

  Limits: The ministry is reporting on its own programme in its own annual report, and none of it is audited. The figures are counts of subsidised projects and money, not a measure of how much research firms and universities do together. The 156 percent co-funding largely reflects the rule itself, which requires each firm to match the subsidy at least one for one. The report's cumulative total of 133 projects cannot be reconciled, from this document, with 15 selection rounds since 2010, so the programme's scale over time is unclear. Whether output or new projects per year have grown or shrunk is not shown. The '161 technologies transferred' figure is self-defined and does not show sales. The terms of the scheme and the 2010 to 2020 funding total come from the operator's site, which a private contractor runs. That site says that from 2025 the projects run inside a departmental project of the state programme for scientific and technological development. That reorganisation could change the scheme, and this record does not track it. The figures are read from a 3 MB Russian-language PDF. The figures say nothing about the effect of post-2022 sanctions on the firms involved.

• 🇪🇬Egyptlinked to Public-private collaborationstarted 2015delivered and closed

  Renewable energy feed-in tariff programme and Benban

  Egypt's Ministry of Electricity and Renewable Energy took private sector-owned renewable generation from 0 MW in October 2015 to 1,500 MW at financial closure by the end of FY2017/18 under its feed-in tariff programme, with about US$2 billion of private investment, much of it in the Benban solar park.

  1,500 MW, private sector-owned renewable energy capacity reaching financial closure, FY2017/18 (reported at June 2019). World Bank, Implementation Completion and Results Report, Egypt First, Second and Third Fiscal Consolidation, Sustainable Energy and Competitiveness Programmatic DPF, retrieved 2026-10-02.

  How it worked: The Ministry of Electricity and Renewable Energy legislated a renewable energy law in 2014, set a feed-in tariff by cabinet decree, licensed private developers through the regulator and allocated land in one desert park, Benban, so many small IPPs could share grid connection and permitting.

  Limits: The indicator was softened mid-programme: DPF1 asked for 1,500 MW commissioned, and DPF2 redefined it as 1,500 MW at financial closure, so the number shows contracts financed, not plants generating, and the ICR's statement that Benban was then the world's fourth-largest solar plant sits in a footnote that says it 'promises' capacity. Most developers and lenders were foreign, and the deal flow leaned on IFC debt (US$653 million) and MIGA risk insurance (US$210 million), so how much of the collaboration was the Egyptian state's own capacity and how much was World Bank Group de-risking cannot be separated. The figure does not show the tariff cost carried by the state offtaker, or that the feed-in tariff was later replaced by other procurement. The source is the World Bank's own evaluation of operations whose results framework included this indicator, and the number is read from a PDF results table.

• 🇺🇿Uzbekistanlinked to Public-private collaborationstarted 2019still operating

  Competitive renewable IPP programme, from Navoi Scaling Solar

  Uzbekistan's government signed its first competitively tendered solar PPP in November 2019 at US¢2.679 per kWh, the 100 MW Navoi plant began commercial operation in December 2021, and by January 2024 operational solar PV capacity in the country stood at 1,497 MW, with agreements for a further 5,677 MW of solar and 5,100 MW of wind by 2030.

  1,497 MW, total operational solar pv generation capacity in uzbekistan, 2024-01. 2.679 US cents per kWh, winning fixed offtake tariff in the first competitive solar ppp (navoi, 25 years), 2019. World Bank, Implementation Completion and Results Report, Uzbekistan Navoi Scaling Solar IPP Project (P170598), retrieved 2026-10-02.

  How it worked: The government, through the energy ministry and a new PPP agency, ran a World Bank Group Scaling Solar tender with standard contracts and a sovereign support agreement, and used the record low tariff it produced to raise its renewable target and repeat the tender model.

  Limits: The 1,497 MW is all operational solar, and the ICR's own table of operating projects includes 400 MW from Gezhouba procured non-competitively, so not every megawatt came through the competitive PPP route. One foreign developer, Masdar, holds most of the competitive capacity listed (Navoi, Scaling Solar 2, Sherobod), so the collaboration is concentrated. Most deals carry World Bank Group, ADB or EBRD payment guarantees, so the state and multilateral lenders carry offtake risk and the number cannot separate Uzbek institutional capacity from multilateral de-risking. Capacity is nameplate, not generation, and solar remains a small share of a system of roughly 14 GW of mostly gas-fired plant. The 5,677 MW and 5,100 MW are agreements, not delivery. The source is the World Bank evaluating its own guarantee operation.

• 🇦🇴Angolalinked to Public-private collaborationstarted 2022still operating

  Lobito Corridor rail concession

  Angola's Ministry of Transport signed a 30-year concession with Lobito Atlantic Railway, in which Mota-Engil holds equity, on 4 November 2022 to rehabilitate, operate and maintain about 1,300 km of the Benguela line from Lobito to Luau and the Lobito mineral terminal. By September 2026 the concessionaire was operating the corridor with about 945 workers, 529 of them transferred from state rail and port operators.

  1,300 km, rail infrastructure operated, maintained and modernised under the concession, 2026-09-10. 945 workers, workers employed by the concessionaire, of whom 529 transferred from state-owned operators, 2026-09-10. Multilateral Investment Guarantee Agency (World Bank Group), press release of 10 September 2026, retrieved 2026-10-02.

  How it worked: Angola's Ministry of Transport tendered the freight operation of a state railway to a private consortium for 30 years, keeping passenger service public and moving state rail and port staff into the concession company.

  Limits: Neither figure measures delivery. 1,300 km is the length of an asset handed over, and the press release does not say what was rehabilitated, what freight moved or what the concessionaire invested. The publisher is a guarantor promoting its own transaction, and the April 2024 summary that dates the contract was written by the concessionaire. 2022 is the contract date, and the source does not say when operations passed to the concessionaire. The concession is four years into 30, so it shows that the state handed over operations, not that the partnership delivered. Passenger service stays with the state operator. The corridor is driven by DRC and Zambian copper exports and by US and EU strategic finance, so the national objective it serves is as much foreign as Angolan. It is one contract, and the construct asks for the frequency and scale of public-private delivery across the country.

• 🇲🇿Mozambiquelinked to Public-private collaborationstarted 2004dismantled

  Beira railways concession, and its termination

  Mozambique's government concessioned the Sena and Machipanda railways for 25 years in 2004 to a joint company in which the state rail enterprise CFM held 49 percent; the Sena line reopened to Moatize more than two years late, carried 266,000 tonnes against a target of one million, and the government terminated the concession on 7 December 2011.

  266,000 metric tons, freight carried on the sena line by the end of the project, 2011. 1,000,000 metric tons per annum, target freight on the sena line by the end of the project, 2009. World Bank, Implementation Completion and Results Report, Beira Railway Project (P082618), retrieved 2026-10-02.

  How it worked: The government of Mozambique, through CFM, handed rehabilitation and operation of two state railways to a private concessionaire in a 51 to 49 joint company, financed largely by an IDA credit, and took the lines back when rehabilitation ran late and traffic did not materialise.

  Limits: The source is the World Bank's own completion report on a project it financed, so the lender is grading a failure it shares, and the independent review of that report found its cost tables mutually inconsistent: the text claims a cost overrun of more than US$50 million while the annex shows actual cost below appraisal, which is why no cost figure is used here. The ICR also dates the termination both to September and to December 2011. The 266,000 tonnes is coal moved in the months after the line opened in August 2011, against a target set for 2009, so it measures lateness as much as capacity. The concession covered two lines in the centre of the country, not the national network. The termination ended the partnership but not the asset: CFM kept operating the rehabilitated Sena line, and later coal traffic on it is not in this record. Blame is contested: the report faults the concession design, the state's 49 percent stake and ineffectual Bank supervision as well as the private operator.

• 🇬🇭Ghanalinked to Public-private collaborationstarted 2018still operating

  Sankofa Gas Project (Offshore Cape Three Points)

  Ghana structured the Sankofa offshore gas development as a public-private deal in which Eni and Vitol invested and the state oil company GNPC bought the gas under take-or-pay terms backed by an IDA payment guarantee, mobilising US$4,653 million of private capital and bringing first gas to power plants in 2018.

  4,653 US$ million, private sector capital mobilized for the sankofa gas field development, 2018. 90 million standard cubic feet per day, average daily gas off-take in 2019, against take-or-pay volume of 170 mmscfd, 2019. World Bank, Implementation Completion and Results Report ICR 147538-GH, Ghana Sankofa Gas Project (P152670), retrieved 2026-10-02.

  How it worked: The Government of Ghana and GNPC contracted to buy domestic gas from a privately financed offshore field on take-or-pay terms, and the World Bank guaranteed those payments so that private sponsors and banks would commit capital to a sub-Saharan gas project.

  Limits: The capital was private and the build was done by the Eni-Vitol consortium; Ghana's institutional move was the deal structure and the offtake, and that side performed worst. GNPC could not pay: arrears reached US$192 million by March 2020 and the sponsors drew US$191.6 million on the letter of credit backed by the IDA guarantee, so the public half of the partnership needed a World Bank backstop within two years of first gas. Off-take in 2018 and 2019 was a fraction of the 170 mmscfd the state contracted to take or pay for, because the pipeline interconnector to the eastern market was late and demand in the west was too small, which means Ghana paid for gas it did not use. The source is the World Bank's own completion report on its own guarantee, written in 2020, and its figures are not published as a machine-readable series; the source stops in mid-2020. One deal, however large, says little about the frequency of public-private delivery the indicator asks for.

• 🇨🇮Côte d'Ivoirelinked to Public-private collaborationstarted 2011still operating

  Henri Konan Bédié toll bridge

  The Ivorian government let a third Abidjan lagoon crossing as a toll-bridge concession to SOCOPRIM, the project company formed in 1997, whose investors include Bouygues Travaux Publics, after a 1996 tender. It funded initial works from October 2011 after the conflict. Private equity and loans, insured by MIGA guarantees of $145 million, built the 1.5 km bridge and 6.7 km road at a cost of about $282 million, and the bridge opened on 16 December 2014.

  282 US$ million, approximate construction cost of the bridge, 2014. 145 US$ million, miga guarantees covering the private investment, 2015. World Bank Group (MIGA / IFC), Public-Private Partnerships Briefs: Côte d'Ivoire Toll Bridge, April 2015, retrieved 2026-10-02.

  How it worked: The government kept a 1990s concession alive through a decade of conflict and then restarted it in 2011, funding the first works itself. A political-risk guarantee from MIGA let a French contractor and development lenders put equity and debt into a toll bridge in a post-conflict country.

  Limits: The source is a two-page promotional brief from the World Bank Group, whose agency MIGA guaranteed the deal. It is an interested party describing its own transaction. The cost is given as approximate, with no baseline budget, so the record shows no cost or schedule performance against plan. The 100,000 vehicles a day is a forecast ('will use'), not observed traffic. No toll revenue, traffic or state guarantee and payment figures are published here, so the fiscal risk the state kept (minimum revenue guarantees, its own up-front works) is not visible. The brief is from 2015. This is one PPP. It does not show how often or at what scale Côte d'Ivoire delivers national objectives jointly with the private sector, which is the construct, and the 18 years from tender to opening were spent mostly in conflict, not in delivery.

• 🇿🇼Zimbabwelinked to Public-private collaborationstarted 2019still operating

  Beitbridge Border Post modernisation (Zimborders build-operate-transfer concession)

  Zimbabwe's government contracted the Zimborders consortium under a build-operate-transfer concession to rebuild the Beitbridge Border Post with a US$300 million investment; works began in March 2019 and the President commissioned the modernised post in September 2022.

  300 US$ million, investment in the beitbridge border post upgrade, 2021-02. Zimbabwe Revenue Authority (ZIMRA), news release 'Beitbridge Border upgrade takes shape', retrieved 2026-10-02.

  How it worked: The Ministry of Finance concessioned the rebuild of the country's busiest port of entry to a private consortium under build-operate-transfer, so the consortium financed construction and recovers it from user fees before handing the facility back.

  Limits: The US$300 million is a headline project value stated in a revenue authority news release during construction, not an audited outturn, and the release itself credits 'the government' with the investment, which blurs how much the concessionaire financed. The concession's term, tariff and risk allocation are not given in the official sources behind this record. One concession, however large, says little about the frequency of public-private delivery, which is half of what the indicator asks; the Treasury's own text shows the model only 'being considered' for other border posts. The record has no official series on clearance times or traffic before and after, so it shows that the facility was built and commissioned, not that it performs better.

• 🇧🇮Burundilinked to Public-private collaborationstarted 2005dismantled

  Private-led coffee sector governance through InterCafe, and its dismantling

  Burundi had run its coffee value chain through a public regulator, ARFIC, beside private-led bodies: InterCafe, whose members were producers, millers, roasters and exporters, and the growers' consortium CNAC with about 109,000 members. The government named all three as joint implementing agencies of the Coffee Sector Competitiveness Project. In January 2020 it replaced them by decree with ODECA, an office under state control. The project closed early in 2021, and 44 percent of the grant was cancelled.

  44 % of grant funding, share of the ida coffee sector competitiveness grant cancelled after the government returned the coffee sector to state control, 2021. 109,000 members, individual members of cnac, the national coffee growers' consortium, one of the three private-led implementing bodies, project appraisal (c. 2016). World Bank, Implementation Completion and Results Report ICR00004849, Burundi Coffee Sector Competitiveness Project (P151869), March 2022, retrieved 2026-10-02.

  How it worked: Over two decades the Government of Burundi handed coffee sector governance to an independent public regulator alongside a private interprofessional association and a growers' consortium, and wrote them into a donor project as co-implementers. In January 2020 it took the value chain back by presidential decree under a new state office, without consulting the partners.

  Limits: The only source is the lender's own completion report, and it is written from the side that lost the argument: the World Bank calls the 2020 decision unilateral and made without advance warning. The 44 percent is the share of one donor project cancelled. It measures what the dismantling cost that project, not how much public-private delivery Burundi lost. The ICR publishes no series on the scale of the private-led arrangement, such as the private share of washing stations, InterCafe's budget or the volumes it governed. The 109,000 CNAC members is a membership count at appraisal, not a delivery figure. The liberalisation began under structural adjustment in the 1980s and was pushed by donors (IFC advised on the third phase), so how far the arrangement was Burundi's own institutional choice is contestable. The 2005 start year rests on a decree the ICR names as the one that reformed the sector. The ICR does not say in which year InterCafe itself was created. Nothing here shows whether coffee output, quality or grower prices fell after ODECA took over. The ICR reports the share of AA/A/FW+ grade coffee rising from 59 to 66 percent, 70 percent of the incremental target. No data exists for the second quality indicator, the specialty share of exports, because ARFIC was removed and ODECA provided none.

• 🇲🇺Mauritiuslinked to University-industry collaborationstarted 2014still operating

  Collaborative Research and Innovation Grant Scheme (CRIGS)

  The Mauritius Research and Innovation Council's Collaborative Research and Innovation Grant Scheme, which co-funds projects led by a local company with a local academic or research institution, had completed 17 projects and had 14 more under way as at 30 June 2020, with 61.5 million rupees of private contribution to the completed ones.

  17 projects, crigs projects completed (project value rs 115.9 m, private sector contribution rs 61.5 m, mric amount rs 53.4 m), cumulative to 30 June 2020. 14 projects, crigs projects ongoing (project value rs 102.18 m, private sector contribution rs 57.22 m), 30 June 2020. Mauritius Research and Innovation Council, Annual Report September 2019 to June 2020, Portfolio section, retrieved 2026-10-02.

  How it worked: The national research council pays roughly half the cost of an R&D project only when a registered company applies with a local university or research institute as partner, so public money follows a firm's own commitment.

  Limits: The council reports on its own scheme. The figures sit in a PDF on page 23 of 66 (printed page 22); the portfolio table on the same page gives a different split for the reporting year alone (10 ongoing and 3 completed CRIGS initiatives), so the cumulative and annual counts must not be mixed. 2019-20 is the latest annual report the council has published: its publications page lists the 2020-21 to 2023-24 reports as awaiting audit or in process, so nothing after June 2020 can be checked against an annual report. The scheme page (mric.mu/collaborative-research-and-innovation-grant-scheme) still describes the scheme and lists 18 funded projects, all marked completed, with no ongoing project and no call dates, so the scheme may have funded little or nothing since 2020; 'operating' rests on the page still offering it and is the weakest part of the record. Thirty-one projects over six years is small even for a country of 1.3 million people, the counts are projects rather than distinct firms, and project value includes the firm's in-kind contribution. The start year is the first call for proposals, June 2014, stated in the council's Annual Report 2014, not at the cited URL. Nothing here shows the research results or commercial outcomes.

• 🇲🇺Mauritiuslinked to Public-private collaborationstarted 2019still operating

  MauCAS Instant Payment System, run by the Bank of Mauritius with banks and non-banks

  The Bank of Mauritius launched the Mauritius Central Automated Switch in August 2019 as a central hub through which banks and non-bank payment providers route instant payments, and its Instant Payment System grew from 126,000 transactions in FY2019-20 to 19.4 million transactions worth Rs63.0 billion in FY2024-25, with 11 banks and 4 non-banks participating.

  19.4 million transactions, maucas instant payment system transactions in the financial year, FY2024-25. 126,000 transactions, maucas instant payment system transactions in its first financial year, FY2019-20. Bank of Mauritius, Annual Report for the year ended 30 June 2025, retrieved 2026-10-02.

  How it worked: The central bank built and ran a national switch that every bank and licensed non-bank could connect to, making transfers between any two accounts settle in seconds, then cut fees and raised limits to push use.

  Limits: The central bank owns and operates the switch, sets the fees and reports the volumes, so this is the system's owner grading its own system. The FY2024-25 figures are in the text on PDF page 88 of the 2025 annual report; the FY2019-20 baseline and the August 2019 launch are in the Annual Report for the year ended 30 June 2020 (https://www.bom.mu/AnnualReport/2020/BOM_June_2020.pdf, PDF pages 129 to 130), where only seven banks and one non-bank were connected. The 95.4 percent growth in FY2024-25 followed the Bank's own 2024 rule changes (interchange revisions, removal of person-to-person charges, a higher limit), so it measures a policy push as much as adoption. Nineteen million transactions a year in a country of about 1.3 million people is still modest beside card and cash use; the report gives no per-capita or merchant figure. The collaboration is a public switch with private participants, not a contractual PPP.

• 🇺🇬Ugandalinked to Public-private collaborationstarted 2005dismantled

  Private electricity distribution concessions, and the 2022 decision to end them

  Uganda ran most of its electricity distribution through private concessions, with Umeme selling about 96 percent of grid energy to end users in 2024, but a Cabinet directive of October 2022 approved the non-renewal of all expiring concessions, and in 2024 the state utility UEDCL took over the assets of two private distributors whose licences were not renewed.

  2 licences, private distribution licences not renewed and taken over by uedcl in 2024 (kil and krecs), 2024. 96 %, umeme's share of grid energy sold to end users, the private concession at its peak (about), 2024. Electricity Regulatory Authority (Uganda), Electricity Supply Industry Performance Report for the Year 2024, retrieved 2026-10-02.

  How it worked: The government concessioned the national distribution network to a private operator for 20 years under ERA regulation, then the Cabinet decided not to renew any expiring concession and folded each private distributor back into the state company as its licence ran out.

  Limits: The figures are in a PDF (the 96 percent and Table 6 on PDF page 21; the Cabinet directive and the KIL and KRECS takeover on PDF page 24). The report covers 2024, when Umeme still operated; it does not record the end of Umeme's own concession, which ran to 2025. That the state utility took over Umeme's area is documented by UEDCL itself: its '100 Days in the Driving Seat' report (https://www.uedcl.co.ug/wp-content/uploads/2025/07/100DaysInTheDriving-Seat-.pdf, page 1) says that on 1 April 2025 UEDCL began its national mandate of electricity distribution; ERA's tariff schedules for Umeme on its site also stop after January to March 2025. The started year is the start of the Umeme concession as commonly dated, not stated in this report. The concession's record is contested: ERA reports Umeme's losses at 16.0 percent in 2024 against a 13.24 percent target, and the decision to end private distribution was a policy choice by the Cabinet, so 'dismantled' records a decision, not a collapse. The report gives no buyout amount; an ERA annual report on the same site (wpdmdl=7485) lists 'no clear asset registers to determine buyout amount for UMEME' as a risk of the concession's expiry.

• 🇲🇱Malilinked to Public-private collaborationstarted 2003dismantled

  Transrail concession of the Dakar to Bamako railway, and its collapse

  Mali and Senegal concessioned the 1,288 km Dakar to Bamako railway to Transrail in 2003 for 25 years, terminated the concession on 5 December 2015 after it failed, and rail traffic, which at its peak had carried about 1.5 million tons of freight a year, stopped completely in 2018.

  0 tons (traffic stopped completely), freight and passenger rail traffic on the dakar to bamako line after the concession's termination, 2018. 1.5 million tons, freight carried by the dakar to bamako railway at its peak (about), peak (year not stated). World Bank, Concept Project Information Document, Dakar-Bamako Intermodal Corridor Project (P171122), September 2019, retrieved 2026-10-02.

  How it worked: Two governments handed a run-down binational railway to a private consortium in a single full concession covering track and operations, keeping minority stakes; the operator could not fund maintenance from traffic revenue, service degraded, the states terminated the contract and the line stopped.

  Limits: The facts are in paragraphs 7, 9 and 16 of a concept-stage document (pages 4 to 6 of 14 in the plain-text version); the zero is the document's statement that traffic 'stopped completely in 2018', not a published tonnage, and the 1.5 million ton peak has no year, so it may predate the concession. The World Bank supported the wave of African rail concessions this one belonged to and is preparing the follow-on corridor project, so the publisher is not neutral about why it failed; the PID also says both governments ignored the Bank's advice to restructure the concession as a lease. The concession was binational: Senegal is an equal party and the record is filed under Mali, where Transrail was headquartered, but it describes both states' collaboration. The PID's diagnosis is shared blame: a 'full concession' design that made the operator pay for track and trains from revenue that could not cover both, decades of underinvestment by both governments before 2003, and road competition after the 2002 Cote d'Ivoire crisis sent Mali's trade through Dakar. After termination both governments paid staff salaries through an interim company (Dakar Bamako Ferroviaire) with a heavy fiscal cost.

• 🇦🇹Austrialinked to University-industry collaborationstarted 2007still operating

  COMET, Competence Centers for Excellent Technologies

  Austria's COMET programme, run by the FFG for the federal ministries and co-funded by the Länder, had approved 81 competence centres and K-projects with eligible costs of 1,669.80 million euros by 30 March 2017, in which about 450 scientific partners and about 1,000 company partners cooperated, according to the Court of Audit.

  81 centres and projects, comet centres and projects approved since programme start (21 k1 centres, 5 k2 centres, 55 k-projects), table 2, 2017-03-30. 1,000 company partners, company partners cooperating within comet (rounded), 2017. Rechnungshof Österreich (Austrian Court of Audit), Bericht 'Forschungsförderungsprogramm COMET: Competence Centers for Excellent Technologies', Reihe BUND 2018/38, retrieved 2026-10-02.

  How it worked: Two federal ministries fund, through their research promotion agency, multi-year competence centres in which universities and firms run a joint research programme, with the Länder adding about half the federal share and companies paying roughly 40 to 60 percent of costs.

  Limits: The figures sit in an 86-page PDF: Table 2 on printed page 33 (PDF page 35) gives the 81 approved centres and projects and 1,669.80 million euros of eligible costs (federal funding 522.06 million, Länder 262.79 million), and the partner counts are on printed page 24 (PDF page 26), both rounded ('rd. 450' and 'rd. 1.000'). The counts are approvals since 2007, not centres running in a given year, and the partner figures count organisations once without saying over what period; the court notes elsewhere about 970 company participations including multiple participations, so 1,000 is not a count of distinct firms. Scientific partners include non-university research bodies, not only universities. The audit period ended in 2016 and the numbers are nine years old; the FFG programme page said on 2026-10-02 that the network currently comprises 23 COMET centres, 17 COMET modules and 15 COMET projects, with no reference date, which supports 'operating' but is not the metric. The court audits spending and compliance, not research output, and it criticised conflicts of interest in centre governance.

• 🇨🇿Czechialinked to University-industry collaborationstarted 2012delivered and closed

  Centra kompetence, the Technology Agency's competence centres programme 2012 to 2019

  The Technology Agency of the Czech Republic funded 34 competence centres under its Centra kompetence programme from 2012, joining 334 participants, 115 research organisations, 118 SMEs and 101 large firms, against a planned 31 centres and 248 participants.

  34 centres, projects (centres) supported, realised against the programme's planned 31 (table 2), 2015-06-30. 334 participants, supported participants in total (115 research organisations, 118 smes, 101 large enterprises), table 2, 2015-06-30. Technologická agentura České republiky (TA ČR), Průběžné hodnocení programu Centra kompetence (interim evaluation of the Competence Centres programme), retrieved 2026-10-02.

  How it worked: The national technology agency funded long-term consortia of research organisations and firms, each at least one research organisation and several companies, to run shared applied research programmes with a minimum private share of costs.

  Limits: The agency evaluated its own programme, and the figures sit in a 101-page PDF: Tables 1 and 2 on printed pages 23 to 24 (PDF page 25), with parameters updated to 30 June and 15 September 2015 from the agency's IS Patriot and the national R&D information system. The counts are of supported centres and participants at contract signature, not of active collaboration in a given year, and a participant is an organisation, so one university can count several times across centres. The evaluation itself says the main objective of developing cooperation between research organisations and firms proved 'nepříliš dostatečné' (not sufficiently met), with results showing it only in part, and that the reported 25.9 billion koruna of product sales is dominated by one centre and was not verified. Funding received by the centres in 2012 to 2014 was 1.53 billion koruna. The programme ran to 2019 and was followed by Národní centra kompetence; this record does not cover the successor.

• 🇩🇰Denmarklinked to University-industry collaborationstarted 1970still operating

  Erhvervsforsker, the Industrial PhD and Industrial Postdoc programme

  Denmark's innovation fund (Danmarks Innovationsfond, formerly Innovationsfonden) awarded 138 Industrial Researcher grants in 2025, from 252 applications and for 147.4 million kroner, each funding a researcher employed by a company and enrolled at or supervised by a university.

  138 grants, grants awarded for industrial researcher (industrial phd and industrial postdoc), 2025. 147.4 million Danish kroner, amount invested in the programme, 2025. Danmarks Innovationsfond (Innovation Fund Denmark), programme page 'Industrial Researcher', retrieved 2026-10-02.

  How it worked: The state pays part of a company's salary cost for a researcher it employs and the university's supervision cost, so that a doctorate or postdoc is carried out on a company problem with joint supervision.

  Limits: The fund reports on its own programme, and the page states only 'All figures are from 2025' with no earlier years, so no trend can be read from it. The page was read through a fetch service; the fund's web firewall refused direct script requests and the annual report PDFs, so the multi-year series could not be checked. A grant is an award at the start of a three-year project, not a completed doctorate, and the count mixes Industrial PhDs with Industrial Postdocs and does not say how many host organisations are public bodies rather than firms. The start year is the commonly cited 1970 founding of the Danish industrial PhD scheme and is not on the cited page. Applications fell to 252 in 2025, and a search snippet gave 140 grants and 149.6 million kroner for 2024, which we could not open at source.

• 🇮🇹Italylinked to University-industry collaborationstarted 2022delivered and closed

  Dottorati innovativi con le imprese (PNRR M4C2 investment 3.3), XXXVIII cycle

  Italy's Ministry of University and Research funded 1,708 company co-financed innovative doctoral scholarships in the 38th doctoral cycle (academic year 2022/2023), against the 5,000 that decree 352/2022 had allocated for that year, and cut the measure's target from 15,000 scholarships and 450 million euros to 6,000 scholarships and 360 million euros.

  1,708 doctoral scholarships, scholarships assigned ('borse assegnate'), xxxviii cycle, summed over tabella a of d.d. 1424/2024, 2022/2023 academic year. 5,000 doctoral scholarships, scholarships planned for 2022/2023 by d.m. 352 of 9 april 2022 (allocation, not delivery), 2022. Ministero dell'Università e della Ricerca, Decreto Direttoriale n. 1424 del 26 settembre 2024 (rideterminazione, XXXVIII ciclo, ex D.D. 2153/2022), as archived by the Internet Archive, retrieved 2026-10-02.

  How it worked: The ministry offered universities EU recovery money for doctoral scholarships on condition that a company co-financed each one and hosted the candidate, then cut the target and raised the public share when firms took up far fewer places than planned.

  Limits: The 1,708 is our sum of the 'Borse assegnate' column over 533 rows of Tabella A (PDF pages 3 to 21), which prints no total; it agrees with the decree's Article 2 total of 85,400,000 euros at 50,000 euros a scholarship. mur.gov.it refused script requests, so the cited copy is the Internet Archive's. The 15,000 target, the 450 million euro original allocation and its redetermination to 360 million euros are in the decree's preamble; the 6,000 figure is on the ministry's measure page (archived 13 June 2026, https://web.archive.org/web/20260613095831/https://www.mur.gov.it/it/pnrr/misure-e-componenti/m4c2/investimento-33-introduzione-di-dott-innovativi-che), not in this decree. The 5,000 allocated by decree 352/2022 covered the 2022/2023 year across courses accredited under both the XXXVII and XXXVIII cycle rules; we read all of them as XXXVIII cycle scholarships, which the decree does not state in so many words. A scholarship counts as assigned when funded, not when the doctorate is completed, and firm co-financing is a condition of the measure rather than a separately counted payment. The shortfall against the 5,000 and the cut in the target are a plan that was not met, not a loss from a delivered peak, so the record is not classed as a reversal: it shows that far fewer firms co-funded than the ministry budgeted for, and the decree's response was to raise the ministry's unit contribution from 30,000 to up to 60,000 euros per scholarship (capped at 20,000 euros extra for this cycle), which shifts cost from firms to the state. Later cycles (decree 117/2023) are not counted here, and the measure ended with the recovery plan in 2026.

• 🇱🇹Lithuanialinked to University-industry collaborationstarted 2012delivered and closed

  Inočekiai LT, innovation vouchers for SMEs to buy research services from universities and institutes

  Lithuania's Agency for Science, Innovation and Technology (MITA) funded 815 innovation-voucher projects in three calls from 2012 to 2014, in which small firms bought applied research, experimental development or feasibility services from universities and research institutes, and 776 of them were completed successfully.

  815 projects, inočekiai lt projects funded across three calls, 2012 to 2014, 2012 to 2014. 776 projects, projects completed successfully, 2015. Lietuvos Respublikos ūkio ministerija (Ministry of Economy of Lithuania), final evaluation report 'Ūkio ministerijos priemonės Inočekiai LT poveikio verslo išlaidoms MTEP vertinimo paslaugos', prepared by UAB Visionary Analytics, 6 February 2017, as held in the SIPER evaluation repository, retrieved 2026-10-02.

  How it worked: The innovation agency gave small firms a fast, small grant that could only be spent on research services from accredited universities and research institutes, to make a first contact between the two cheap and quick.

  Limits: The report was commissioned and paid for by the ministry and written by a consultancy, and the copy that opens is held by SIPER, an academic repository of innovation policy evaluations, not by the ministry; the counts themselves come from MITA's activity reports for 2012 to 2015, which the evaluation cites. The figures are on printed page 19 (PDF page 19) of a 112-page PDF, with the call-by-call table (Table 6: 185, 174 and 456 funded applications). Each voucher was small, up to 5,792.4 euros at up to 80 percent public funding, so a project is a short service purchase rather than joint research, and 104 firms were funded in two calls, so 815 projects are fewer distinct firms. The third call relaxed the private contribution and favoured firms with no prior research contact, which raised volume. The vouchers were EU structural fund money and the measure ended with the 2007 to 2013 period; a successor, 'Inovaciniai čekiai', was being prepared for 2014 to 2020 and is not covered here.

• 🇱🇻Latvialinked to University-industry collaborationstarted 2011delivered and closed

  Kompetences centri, the industry competence centres programme

  Latvia's six industry competence centres, funded through the Investment and Development Agency with 53.2 million euros of ERDF money, had run 231 research projects involving 164 companies and 33 scientific partners by 31 December 2014, of which 78 were completed and 44 percent involved cooperation with scientific institutions.

  231 research projects, research projects supported in the six competence centres (table 3, total), 2014-12-31. 164 companies, industry partners in the six competence centres (table 3, total), 2014-12-31. Ekonomikas ministrija (Ministry of Economics of Latvia), ex-ante evaluation of SAM 1.2.1.1 'Atbalsts jaunu produktu un tehnoloģiju izstrādei kompetences centru ietvaros', Riga 2015, chapter 2 (evaluation of the 2007 to 2013 programme), retrieved 2026-10-02.

  How it worked: The economics ministry funded industry-led competence centres, companies grouped by sector, which commissioned or ran research projects for their members, partly with universities and institutes, from EU structural funds.

  Limits: The ministry evaluated a programme its own agency ran, and the figures are in a 43-page PDF: the 164 companies, 231 projects and 78 completed are on printed page 10 (PDF page 10) and Table 3 with 33 scientific partners on printed page 10 to 11 (PDF page 11). Most projects were firms' own research: the same evaluation says only 44 percent of projects involved cooperation with private or public scientific institutions, with 24 registered institutions, so the 231 overstates university-firm collaboration, and the 33 scientific partners count partners of the centres, not distinct universities. The programme was suspended from December 2011 to May 2012 and the centres only started full operation in 2013; by the end of 2014 they had drawn 56 percent of their funding and completed 34 percent of projects. Patents (3) and publications (25 against 107 planned) were few. The data are eleven years old. The evaluation states the 2007 to 2013 centre projects were to run until 31 December 2015, so this delivery is concluded; a successor competence centre programme for 2014 to 2020 under the same agency is what this ex-ante evaluation prepares, and this record does not show its scale.

• 🇳🇴Norwaylinked to University-industry collaborationstarted 2007still operating

  Senter for forskningsdrevet innovasjon (SFI), Centres for Research-based Innovation

  The Research Council of Norway allocated 354 million kroner in 2025 to its Centres for Research-based Innovation, eight-year consortia of research institutions, companies and public bodies whose business and public-sector partners must co-finance at least half the cost, and in December 2025 named eight new centres with 98 business partners and 22 public-sector partners.

  354 million Norwegian kroner, funding allocated to the sfi centres, 2025. 98 business partners, business partners in the eight sfi centres announced in december 2025, 2025-12-15. Norges forskningsråd (Research Council of Norway), programme page 'Sentre for forskningsdrevet innovasjon (SFI)', retrieved 2026-10-02.

  How it worked: The research council funds eight-year centres in which a university or institute and research-active firms run a shared programme, with the council paying up to half and the partners the rest, and renews the portfolio in generations through open competitions.

  Limits: The council reports on its own scheme. The 354 million kroner is a funding figure shown as a number block on the programme page, not a count of collaborations, and the page gives no partner count for the centres running in 2025; the same page lists 22 centres under SFI-IV (2020-2028) and 17 under SFI-III (2015-2023), counted by us from the lists. The partner figures come from a separate council news item of 15 December 2025 (https://www.forskningsradet.no/nyheter/2025/nye-forskningssentre-innovasjon/), describe centres that were only being contracted, and the centres also include 16 research institutions, 17 international research partners and 13 foreign companies. Host institutions include research institutes such as SINTEF, not only universities. The 2007 start is from the history of the scheme (first centres appointed 2006, operating 2007) and is not stated on the cited page. Nothing here measures the innovation the centres produced.

• 🇩🇰Denmarklinked to Public-private collaborationstarted 2021still operating

  MitID, the digital identity built by the state and the banks

  Denmark's Agency for Digital Government, representing the state, regions and municipalities, and the banks, through a Finance Denmark subsidiary, jointly own MitID, launched in May 2021 to replace NemID and migrated in 2021 and 2022; by July 2026, 97.2 percent of residents over 15 had an active MitID, and it averaged 94.4 million transactions a month in 2025.

  97.2 % of population over 15, share of the population over 15 with an active mitid, 2026-07. 94.4 million transactions per month, mitid transactions per month, average for 2025, 2025. Digitaliseringsstyrelsen, MitID statistik (mitid.dk); partnership terms on digst.dk, Om MitID, retrieved 2026-10-02.

  How it worked: The state and the banks tendered and co-own one identity solution that works for both online banking and public services, contracted one supplier to build and run it, and moved about five million users off the previous shared solution, NemID.

  Limits: The page is run by the partnership the record describes, so this is the owner grading its own system; it says the figures were updated in July 2026 and the 94.4 million is labelled an average for 2025, while the 97.2 percent and the 5.3 million registered users carry no separate date. The page refused WebFetch with HTTP 403 and was read with a browser user agent. The partnership terms (established in 2016 after a tender, the state represented by the agency, the banks by an FR I subsidiary of Finance Denmark, the partnership owning the rights to the solution, name and logo) and the timeline (go-live May 2021, migration through 2021 and 2022) are on https://digst.dk/it-loesninger/mitid/om-mitid/. The build and operation were contracted to one private supplier, Nets DANID, for DKK 939 million plus ten years of operation, so the private side is both the banks as co-owners and a vendor. Near-universal coverage follows from mandatory digital post and bank login in Denmark, not only from the partnership's design, and the share says nothing about people who cannot use the app; the minister sent parliament a project evaluation in September 2024 (ft.dk, DIU Alm.del bilag 3), which this record has not read.

• 🇮🇹Italylinked to Public-private collaborationstarted 2016still operating

  SPID, the public digital identity issued by accredited private providers

  Italy's Agency for Digital Italy runs SPID, the public digital identity system in which accredited identity providers, private firms such as Aruba, InfoCert, Namirial and Register alongside state-controlled Poste Italiane and public companies such as Lepida and InfoCamere, issue the credentials under agreements with the agency, and by 8 October 2025 41 million SPID identities had been issued, when the agency extended the providers' agreements to 2027.

  41,000,000 identities, spid digital identities issued, 2025-10-08. 55,000,000 cards, electronic identity cards (cie) issued, the state-only alternative (of which 9 million cie credentials activated), 2025-10-08. Agenzia per l'Italia Digitale (AgID), SPID, rinnovate le convenzioni con gli Identity Provider fino al 2027 (news, 8 October 2025), retrieved 2026-10-02.

  How it worked: The digital agency set technical rules and signed agreements with accredited identity providers, who verify citizens and issue credentials that every public administration must accept, so private firms carried the cost of enrolment while the state set the standard.

  Limits: The 41 million is a rounded figure in a news item, not a statistics series, and counts identities issued, not active users; one person can hold SPID from more than one provider, so the count overstates people. The mix of providers is read from the agency's list of accredited identity providers (https://www.agid.gov.it/it/piattaforme/spid/identity-provider-accreditati), which shows the first accreditations in December 2015, the first agreements in February 2016, an additional act signed with each active provider on 8 October 2025, and one provider (In.Te.S.A.) that ceased in April 2023; the agency publishes no split of identities by provider, so the private share of issuance is not known, and the largest provider, Poste Italiane, is state-controlled. The future of the partnership is uncertain: the news item frames the extension as accompanying a transition toward a single national identity built on the state-issued CIE and the IT Wallet. If the government moves logins to CIE the partnership will shrink; this record does not count that as erosion because the item reports no loss yet. The item gives no transaction figure, so the record shows reach, not use. The 55 million CIE in the second metric are cards issued; the same item says only 9 million CIE digital credentials had been activated, so the two counts in the item's 96 million total are not like for like.

• 🇨🇿Czechialinked to Public-private collaborationstarted 2021still operating

  D4 Příbram to Písek motorway, the first Czech transport PPP

  The Czech Ministry of Transport procured the completion of the D4 motorway between Příbram and Písek as the country's first transport public-private partnership, and the concessionaire Via Salis opened 32 km of new motorway and 16 km of modernised sections on 17 December 2024 after 43 months of construction, as the ministry had announced in June 2021, with the state paying an availability payment until 2049.

  32 km (plus 16 km modernised), new motorway built and opened under the ppp, 2024-12-17. 29.5 CZK billion, availability payment approved by the chamber of deputies, against the audit office's inflation-adjusted estimate of about czk 33.7 billion, 2023-12-11. Ministerstvo dopravy ČR, Dálnice D4 se otevřela řidičům, první dopravní PPP projekt zrychlí cestu na jih (press release, 17 December 2024), retrieved 2026-10-02.

  How it worked: The ministry tendered a design, build, finance, operate and maintain concession for one motorway corridor, the concessionaire took over the existing sections in 2021 and built the gap, and the state pays monthly only once the road is open, with deductions when availability or quality falls short.

  Limits: The opening, the kilometres and the 43 months are the ministry's own press release; the planned December 2024 opening is in the ministry's release of 9 June 2021 as listed on its D4 tag page, so on-time delivery is the ministry's claim checked against its own earlier announcement. The cost figures come from the Supreme Audit Office (NKÚ, audit 22/29, press release of 11 December 2023, https://www.nku.cz/cz/pro-media/tiskove-zpravy/nku-upozornuje:-vystavba-dalnice-d4-formou-projektu-ppp-mozna-pro-stat-nebude-tak-vyhodna--jak-se-predpoklada-id13640/), which warned that the value-for-money case rested mainly on the adviser's foreign experience because the ministry could not state its own maintenance costs for the 1,276 km of motorway it ran, and that with inflation the state's availability payment would be about CZK 4.2 billion above the CZK 29.5 billion approved. One project is not a programme: the record shows that the Czech state can close and deliver a PPP, not that it does so often. The concessionaire is French-owned (VINCI Highways and Meridiam), and the construction was done by VINCI's Czech subsidiary.

• 🇱🇻Latvialinked to Public-private collaborationstarted 2021still operating

  Ķekava bypass, the first major road PPP in the Baltic states

  Latvia's Ministry of Transport and the state roads company signed a design, build, finance and maintain contract for the Ķekava bypass on the A7 on 16 July 2021, the private partner Kekava ABT built the roughly 18 km road from 2021 to 2023 and it opened on 13 October 2023, with the state paying EUR 250.13 million in availability payments over 20 years.

  250.13 EUR million, total availability payments over the 20-year maintenance period, 2023-10-13. 27 months, months from contract signature (16 july 2021) to opening (13 october 2023), 2023-10-13. Centrālā finanšu un līgumu aģentūra (CFLA), Atklāj vērienīgāko PPP ceļu projektu Baltijā: Ķekavas apvedceļu (news, October 2023), retrieved 2026-10-02.

  How it worked: The transport ministry and the state roads company procured a bypass as one design, build, finance and maintain contract, the private partner built it in about two years, and the state pays quarterly for 20 years once the road is available.

  Limits: CFLA supervises PPP projects in Latvia and is part of the delivery it reports, so this is the supervisor's announcement on opening day, not an audit. The 27 months are our arithmetic from the two dates on the page; the page gives no planned opening date, so it does not show whether the road opened on schedule, and it gives no construction cost separate from the availability payments. The same page says the state roads company had to obtain several Eurostat opinions to keep the project off the government balance sheet, and that CFLA is working to remove that step: the procedure was slow before construction began. One project is not a programme; the page calls it the first large road PPP in the Baltic states, which also means Latvia has no record of repeating it. Representatives of the Nordic Investment Bank and the European Investment Bank took part in the opening; the page does not describe the financing.

• 🇱🇹Lithuanialinked to Public-private collaborationstarted 2009still operating

  Public-private partnership contracts monitored by CPVA

  Lithuania's state and municipal institutions had concluded 76 public-private partnership contracts by 31 December 2025, 49 of them in force at the end of 2025 across culture and sport, waste, energy, law enforcement, education, transport and defence, with only one new contract signed in 2025.

  76 contracts, ppp contracts concluded, cumulative, 2025-12-31. 49 contracts, ppp contracts in force, 2025-12-31. Centrinė projektų valdymo agentūra (CPVA), PPP Lietuva, Viešojo ir privataus sektorių partnerystė Lietuvoje: 2025 m. rezultatai ir 2026 m. projektų planai, retrieved 2026-10-02.

  How it worked: A central agency advises and monitors, and individual ministries and municipalities procure PPP contracts for buildings, lighting, waste and defence facilities, reporting each year so the agency can publish a national portfolio.

  Limits: CPVA compiled the figures from data submitted by state and municipal institutions; it is the national PPP centre, not an auditor, and the underlying annual report is linked from the page but was not opened. Most contracts are municipal (24 municipalities in 2025), so the count mixes national objectives such as defence and law enforcement, where the public sector made its largest payments in 2025 (EUR 32.5 million and EUR 10.3 million), with local sports halls and street lighting. A contract count says nothing about value for money; the National Audit Office has published a critical audit of PPP project management (valstybeskontrole.lt), which this record has not read. The gap between 76 concluded and 49 in force is not explained on the page (expiry, termination or both). The pace is slow: one contract in 2025, and 18 projects planned for 2026 that are still in preparation. The start year 2009 is the year of the finance minister's reporting rules (order 1K-489 of 24 December 2009) cited on CPVA's statistics page, not a documented first contract.

• 🇸🇰Slovakialinked to Public-private collaborationstarted 2016still operating

  D4/R7 Bratislava bypass PPP

  Slovakia built the D4 motorway and R7 expressway around Bratislava, 59.0 km at EUR 1,670 million, as a public-private partnership between 2016 and 2021, the only PPP among the major road projects the Supreme Audit Office listed in its 2025 audit of road investment.

  59 km, length of the d4/r7 ppp project, 2021. 1,670 EUR million, cost of the d4/r7 ppp project as listed by the audit office, 2025. Najvyšší kontrolný úrad Slovenskej republiky (NKÚ SR), Správa o výsledku kontroly: Investičné projekty v cestnej doprave (2025), retrieved 2026-10-02.

  How it worked: The transport ministry procured the Bratislava bypass as one availability-payment concession covering design, construction, financing and decades of operation, so a single private partner delivered 59 km at once rather than in budget-limited sections.

  Limits: The figures are one row of a table on printed page 12 of the audit office's PDF, which lists the period 2016 to 2021, 59.0 km, EUR 1,670.00 million and 'PPP' as the funding source; the audit is about road investment generally and does not evaluate the PPP itself, its cost against a public alternative or its schedule against plan. The table gives no definition of the cost (construction cost, capital value or total availability payments), so the EUR 1,670 million cannot be compared with the other rows without that caveat. The opening dates of the sections are not on this page; 2021 is the end of the period the table gives. The same table shows every other major road project financed from EU or state funds, so the record documents one delivered PPP, not a programme, and this source says nothing about Slovakia's earlier PPP attempts.

• 🇳🇴Norwaylinked to Public-private collaborationstarted 2005still operating

  OPS road projects of the Norwegian Public Roads Administration

  The Norwegian Public Roads Administration has delivered four road sections as public-private partnerships (OPS) now in operation, opened in 2005, 2006, 2009 and July 2020, each run by a private company for 20 to 25 years, and had two more under planning or construction in 2026, Rv. 555 Sotrasambandet and E10/Rv. 85 Tjeldsund to Langvassbukt.

  4 projects, ops road projects completed and in operation, 2026-10. 2 projects, ops road projects in planning or under construction, 2026-10. Statens vegvesen, Våre OPS-prosjekter, retrieved 2026-10-02.

  How it worked: Parliament designates a road section for OPS, the roads administration tenders it, and a private company plans, builds, finances, operates and maintains the section for 20 to 25 years against state payments, after which it returns to the state.

  Limits: The page is the roads administration's own list with no date of update, so the asOf is the retrieval date. It gives no cost, schedule or value-for-money figure, so the record shows frequency and duration, not performance. Six projects in two decades is a small programme for a country that builds many roads, and it moved in fits: three sections opened between 2005 and 2009, then nothing opened until 2020, after the transport ministry asked the agency in 2015 to use the model on three new projects (stated on the agency's OPS page, https://www.vegvesen.no/vegprosjekter/prosess/ops-ppp/); the list does not say what became of each of those three. The page says parliament decides which projects use OPS, so the count reflects political choices project by project. Projects by Nye Veier, the state road company, are not on this list.

• 🇭🇷Croatialinked to Public-private collaborationstarted 2006operating below its peak

  Public-private partnership contracts in Croatia's PPP register, and their dwindling

  Croatia's register of public-private partnership contracts lists 17 contracts with an estimated capital value of EUR 338.3 million, 13 of them signed between 2006 and 2008, mostly Varaždin county schools, one in 2013 for Zagreb Airport, one in 2016 and two street lighting contracts in 2018, and none since October 2018.

  0 contracts, ppp contracts entered in the register with a signature date after october 2018, 2026-10. 13 contracts, ppp contracts signed 2006 to 2008, the peak, 2008. Ministarstvo gospodarstva, Registar ugovora javno-privatnog partnerstva (Invest Croatia), retrieved 2026-10-02.

  How it worked: Counties and cities signed design, build, finance and maintain contracts for schools, halls and a bus station in a burst before the 2008 crisis, the state concessioned the Zagreb Airport terminal in 2013, and then PPP contracting all but stopped while responsibility for PPP passed to the economy ministry.

  Limits: The counts are ours: the register page lists 17 contracts and a total of EUR 338,260,027, and the signature dates are on each contract's own page under /jpp-project/ (for example Zagreb Airport, 2 December 2013; Novi Vinodolski street lighting, 25 October 2018, the latest), which we opened one by one and tallied. The register holds only contracts entered under the PPP framework; concessions under the separate concessions law, such as motorways, are not in it, so the record does not show that private participation in Croatian infrastructure stopped, only that the PPP model did. The ministry's project list (https://investcroatia.gov.hr/jpp/projekti/) also shows approved projects not yet contracted, such as Trg pravde in Zagreb (EUR 179.2 million) and county schools, so the pipeline is not empty. The peak itself was narrow: ten of the 13 early contracts are in Varazdin (eight school extensions, the county building and a sports hall). Zagreb Airport alone is EUR 188.6 million, more than half the register's value. The register page has no date of update, so the zero is as of retrieval.

• 🇬🇷Greecelinked to Public-private collaborationstarted 2006still operating

  Greece's PPP programme under the 2005 PPP law

  Greece's Inter-Ministerial PPP Committee and the finance ministry's PPP Unit have taken 33 public-private partnerships to signed contracts since 2006, from Attica school buildings and fire stations to waste plants, the Kalamata to Rizomylos and Crete northern road sections and ultrafast broadband, with another 34 projects listed in procurement or preparation in 2026.

  33 contracts, ppp contracts signed since 2006, 2026-10. 34 projects, projects in the ppp unit's list still in preparation or procurement, 2026-10. Ministry of Economy and Finance, Public Private Partnerships Unit (pppunit.minfin.gov.gr), home page and project list, retrieved 2026-10-02.

  How it worked: A 2005 law set up an inter-ministerial committee that approves which projects may be procured as PPPs and a central unit in the finance ministry that prepares and tenders them for ministries, regions and municipalities, mostly as availability-payment contracts.

  Limits: The home page states 33 signed PPP contracts and 34 approved PPP projects since 2006, without a date of update. We read the 67 project pages in the unit's list (https://pppunit.minfin.gov.gr/en/sdit/erga, seven pages) and tallied their status fields ourselves: 33 'Contract signed', 13 in competitive dialogue, 13 at request for proposals, 6 at expression of interest and 2 approved and in preparation, so the 34 in the second metric is our count. The list counts the ultrafast broadband programme as seven lots, each showing the programme's EUR 868 million cost, so a sum of project costs would be wrong and the record gives none. 'Contract signed' is the unit's last populated status; no project is marked completed, so the list does not say how many assets are in operation or whether they came in on time and budget, and several roads and waste plants took three to ten years from committee approval to signature (for example the Peloponnese waste system, approved December 2011, signed June 2018, and the Nafplio marina, approved November 2013, signed April 2023). The unit is part of the ministry that runs the programme, and its home page cites a World Bank ranking as an endorsement; the ranking is not the evidence here. 2006 is the first committee approval in the list (Attica schools, April 2006).

• 🇧🇪Belgiumlinked to Public-private collaborationstarted 2023still operating

  De Kade prison in Antwerp, a federal DBFM contract

  Belgium's federal Buildings Agency had the new 440-place prison de Kade in Antwerp, part of the federal prisons masterplan, designed, built and financed by the consortium Hortus Conclusus under a DBFM contract, with works from 2023 to 2026 at a construction cost of EUR 195.5 million and an annual availability fee of EUR 23.1 million for 25 years of maintenance, after which the prison returns to the federal state free of charge.

  195.5 EUR million, construction cost of the dbfm prison, excluding vat, 2025-09. 23.1 EUR million per year, annual availability fee, including vat, 2023 prices, 2025-09. Régie des Bâtiments / Regie der Gebouwen, Nouvelle prison « de Kade » (project page, updated September 2025), retrieved 2026-10-02.

  How it worked: The federal Buildings Agency tendered a design, build, finance and maintain contract for a new prison, a consortium of an infrastructure investor and contractors built and financed it, and the state pays an annual availability fee for 25 years while the Justice ministry runs the prison.

  Limits: The page is the contracting agency's own project sheet. It says the prison 'a été réalisée' under a DBFM contract and that Antwerp 'now has two prisons', but it states the works period as 2023 to 2026 and gives no opening date or planned date, so the record cannot show whether the prison opened on time or on budget; status operating rests on the page's own past tense, updated September 2025. The construction cost and the availability fee are the agency's figures, not an audit; 25 years of fees at EUR 23.1 million in 2023 prices would be about EUR 578 million before indexation (our arithmetic), well above the construction cost, which is the normal shape of a DBFM but means the cost to the state is the fee stream, not the build cost. One project is not a programme: the page places the prison in the federal masterplan for detention and internment, but this record documents only this contract and says nothing about the masterplan's other projects. The consortium's members are named on the page (TINC, Jan De Nul, EEG, with ISS for services).

• 🇲🇩Moldovalinked to University-industry collaborationstarted 2024still operating

  ANCD innovation vouchers (Vouchere Inovationale) for SME research services

  Moldova's National Agency for Research and Development (ANCD) financed 13 innovation voucher projects in 2025, in which small and medium firms buy research and development services from research organisations, after 28 applications to the 2025-2026 call yielded 10 funded projects and four funded projects were stopped when the firms declined to sign contracts.

  13 projects, innovation voucher projects financed in the year (10 from the 2025-2026 call plus continuing 2024-2025 projects, less those stopped), 2025. 28 proposals, proposals submitted to the 2025-2026 innovation voucher call, 2025. Agentia Nationala pentru Cercetare si Dezvoltare (ANCD), Raport anual ANCD pentru anul 2025, section 1.4 Proiecte de Inovare si Transfer Tehnologic, retrieved 2026-10-02.

  How it worked: The national research agency gives small firms a voucher that pays a public research organisation to do a defined R&D or testing job for them, with the firm co-financing a share.

  Limits: The agency reports on itself, on pages 28 and 29 of a 137-page PDF. The arithmetic in the report does not close on its face: 10 new projects plus 7 continuing ones is 17, and the report also says four projects were stopped when three firms refused to sign financing contracts, and states the total of 13 without showing the subtraction. The volume is very small: 1,073.0 thousand lei of state money and 153.4 thousand lei of firm co-financing in 2025, roughly 60,000 euros, and the same report counts five technology transfer projects approved in 2025 under a separate call, not added here. The 2024 report gives seven voucher projects funded in 2024, so the instrument is two years old and its scale is pilot-like even if national in reach. Counts are projects financed, not completed or commercialised. The started year is the first funded call read in the 2024 report, not a programme launch date.

• 🇷🇴Romanialinked to University-industry collaborationstarted 2014still operating

  Proiect de transfer la operatorul economic (PTE) research-to-firm transfer grants

  Romania's research funding agency UEFISCDI accepted 35 firm-led projects with research organisation partners for funding in the 2024 PTE (transfer to the economic operator) competition, plus one more on a second list, after 47 in the 2021 competition.

  35 projects, pte 2024 projects on the list accepted for funding, PTE 2024 competition. 47 projects, pte 2021 projects on the list accepted for funding, PTE 2021 competition. Unitatea Executiva pentru Finantarea Invatamantului Superior, a Cercetarii, Dezvoltarii si Inovarii (UEFISCDI), Lista proiectelor PTE 2024 acceptate la finantare (PNCDI IV, Programul Parteneriate pentru Inovare), retrieved 2026-10-02.

  How it worked: The national research funding agency runs a periodic competition in which a company leads a project to take a research organisation's result to market, with the state grant paying both sides under state aid rules.

  Limits: The count is ours: the four-page PDF numbers its rows 1 to 35 and is a list of projects accepted for funding, not contracts signed or completed; a separate 'Lista nr. 2' adds one more project (PN-IV-P7-7.1-PTE-2024-0336), and a reserve list exists, so the contracted number may differ. The 2021 figure is the row count of a different file on the same agency site (resource-862977, 'Lista proiectelor PTE 2021 acceptate la finantare'), not at the cited URL. Each project is coordinated by a company with one or more universities or national R&D institutes as partners, and the maximum grant was about 1.5 million lei, so this is a few dozen projects a competition in a country of 19 million. PTE competitions are not annual (2019, 2021, 2024, and a 2026 call that closed on 16 September 2026 with 378 proposals submitted and none funded yet), so the gaps between rounds are part of the record. The started year is approximate: PTE ran under PNCDI III from about 2014 to 2016, which is not stated in the cited file. The figures say nothing about technology actually transferred or sold.

• 🇷🇸Serbialinked to University-industry collaborationstarted 2019still operating

  Innovation Fund Collaborative Grant Scheme (Program saradnje nauke i privrede)

  Serbia's Innovation Fund financed 79 joint research and development projects of private small and medium firms with public research organisations across eight calls of its Collaborative Grant Scheme, the eighth call, decided in 2024, funding eight.

  79 consortium projects, collaborative grant scheme projects listed as funded, all eight cycles (counted from the fund's per-cycle lists), cycles one to eight, to 2024. 8 consortium projects, projects funded in the eighth cycle, 2024. Innovation Fund of the Republic of Serbia (Fond za inovacionu delatnost), Collaborative Grant Scheme, funded projects, all cycles, retrieved 2026-10-02.

  How it worked: The national innovation fund runs competitive calls in which a private SME leads a consortium with a public research organisation, pays part of the budget, and must route at least 15 percent of the grant to the research partner.

  Limits: The Fund reports on its own programme and publishes no total: 79 is our count of the project entries on its per-cycle pages (14, 9, 13, 12, 11, 6, 6 and 8 projects in cycles one to eight; each page renders every project twice, and the count is of distinct titles). The pages carry no dates, so the year of each cycle is not read there; that the eighth call funded eight projects in 2024 is from the Ministry of Finance CFCU release of 15 August 2024 ('8 under the Collaborative Grant Scheme Program'), and a Fund release of 20 March 2024 reports six joint projects in the call before. The counts are grants awarded, not completed projects or commercial results. The scheme is co-financed by the European Union through IPA and began under World Bank financing, so the delivery is the Fund's but the money is partly external. Annual volume has fallen from 11 to 14 projects a call in the early cycles to six to eight, and the programme page shows the call as closed with no ninth cycle listed on 2026-10-02, so 'operating' rests on the absence of any announced end. The started year is approximate: the first cycle is not dated on the cited page. Each consortium needs only one firm and one public research organisation, and requires firms to co-finance 30 to 40 percent, so 79 projects is a small national volume.

• 🇷🇴Romanialinked to Public-private collaborationstarted 2009operating below its peak

  Prima Casa / Noua Casa state-guaranteed mortgages through banks, and their wind-down

  Romania's Prima Casa programme, renamed Noua Casa in 2020, had the National Credit Guarantee Fund for SMEs (FNGCIMM) issue 334,048 state guarantees and guarantee promises worth RON 31.68 billion on mortgages made by participating banks between 2009 and December 2024, but only 811 guarantees in 2024, and the ratio of guaranteed lending to banks' unguaranteed housing loans fell from about 70 percent in 2017 to about 10 percent in 2023.

  10 % (circa), ratio of prima casa / noua casa guaranteed lending to banks' unguaranteed housing loans to individuals, 2023. 70 % (approx.), same ratio in 2017, the first year of the 2017-2021 strategy and the earliest year the source gives, 2017. Ministry of Finance of Romania, Nota de fundamentare for the draft Government Decision on the Noua Casa guarantee ceiling for 2025, 4 March 2025, retrieved 2026-10-02.

  How it worked: The finance ministry, through a state guarantee fund acting as its agent, guaranteed half or more of each first-home mortgage that participating banks underwrote, set a lower interest cap and annual guarantee ceilings, and then cut the ceilings year by year so private lending took over.

  Limits: The live file at mfinante.gov.ro did not answer a script on the day of retrieval, so the record cites the Internet Archive copy; the figures are on pages 1 and 2 of a nine-page PDF, prepared with FNGCIMM, which runs the programme, so the ministry is grading its own scheme. The decline was largely chosen: the note says the 2017-2021 medium-term strategy set falling guarantee ceilings to decelerate state intervention and that its objectives were met, while private mortgage lending without a state guarantee grew, so the erosion is a planned withdrawal as much as a loss of capacity. The 70 and 10 percent are approximate ('aprox.', 'circa') and describe a ratio of guaranteed to unguaranteed lending as the note words it, which is not the same as a market share. 2017 is the first year of the strategy, not a demonstrated peak: the note gives no earlier year, and the programme's share was probably at least as high in its first years after 2009. 811 guarantees in 2024 is a calendar-year count against a cumulative total, and the note does not give the annual peak; the 1,449 guarantees the Ministry of Finance paid out to banks since launch show the state carried real credit risk. Whether the programme raised home ownership or mainly house prices is not shown.

• 🇭🇺Hungarylinked to Public-private collaborationstarted 2013delivered and closed

  Funding for Growth Scheme (Novekedesi Hitelprogram), central bank refinancing lent by commercial banks

  Hungary's central bank refinanced commercial banks at zero interest on condition that they lent to small and medium firms at a capped margin under the Funding for Growth Scheme, and across its phases from June 2013 to 2022 some 75 thousand enterprises received about HUF 6,400 billion of loans, more than half of them investment loans.

  75,000 enterprises (some), enterprises that gained access to fgs loans across all phases, 2013-2022. 6,400 HUF billion (approximately), loans granted under all fgs phases, 2013-2022. Magyar Nemzeti Bank, From the turnaround in lending to the green transition: collection of studies on the nine years of the Funding for Growth Scheme to date, 2013-2022, retrieved 2026-10-02.

  How it worked: The central bank offered banks zero-cost refinancing for loans to small and medium firms on condition that they cap the client rate, so banks did the origination, credit assessment and servicing while the central bank bore the funding cost.

  Limits: The central bank designed, financed and here evaluates its own scheme; both headline numbers are rounded ('some', 'approximately') and appear in a PDF on page 7 and again on pages 9 and 20; the MNB also publishes a Power BI dashboard of the loans that a script cannot read. The MNB carried the funding cost and banks the credit risk, so the partnership was a public subsidy delivered by private lenders, with the policy cost sitting off budget on the central bank's balance sheet. The book's estimate that the scheme added more than 5 percentage points to GDP between 2013 and 2021 is the MNB's own model result and is not used here. Enterprises are counted across phases (FGS 1 to 3, FGS fix, FGS Go!, Green Home), so the 75 thousand cannot be compared with the stock of firms in any one year, and the volumes include refinanced foreign currency loans in the first phase. Some lending would have happened without the subsidy; the record does not measure additionality.

• 🇷🇸Serbialinked to Public-private collaborationstarted 2020operating below its peak

  Moja prva plata, state-paid first-job training placed with employers, and its 2025 halt

  Serbia's Government paid a monthly allowance to young people without work experience while employers trained them for nine months under the Moja prva plata programme run by the National Employment Service, which placed 9,237 young people with 5,471 employers in the 2021 cycle and 7,214 with 4,520 employers from the 2023 call in January 2024, but ran no call for 2024 and started no activity under the measure in 2025 against a plan of 10,000 places.

  0 young people, young people included in moja prva plata in the year (table 6, measure 5.8), against 10,000 planned, 2025. 9,237 young people (at 5,471 employers), young people selected by participating employers in the 2021 cycle, the highest dated count read, 2021-12-03. Nacionalna sluzba za zaposljavanje (National Employment Service), Izvestaj o radu Nacionalne sluzbe za 2025. godinu, Table 6 and section 5.8, retrieved 2026-10-02.

  How it worked: The Government let firms post training positions on a national portal, let young people apply, had employers pick candidates, and paid the trainees directly through a tripartite contract with the National Employment Service, so firms trained at no wage cost.

  Limits: The agency reports on its own programme. The zero is in Table 6 (PDF page 23), measure 5.8, planned 10,000 and realised 0, and section 5.8 (PDF page 41) states that no activities to implement the measure were started in 2025; the report gives no reason. The 2021 figure is from a Government portal release (srbija.gov.rs/vest/595588, 3 December 2021), not at the cited URL, and counts young people employers selected at the start of a cycle, not completions; the 2022 cycle was not read, so 2021 is the highest dated count found, not a demonstrated peak. The 2024 report (Izvestaj o radu NSZ za 2024. godinu, Table 6 and page 45, nsz.gov.rs) gives 7,214 young people at 4,520 employers from the 2023 call, contracted between 1 and 19 January 2024, of whom 4,950 were with 3,892 private employers and 2,264 with 628 public employers, and says no public call for 2024 was carried out; the first cycle in 2020 also reserved 2,648 places for public-sector employers, so the programme was never wholly public-private. The 35 percent of first-cycle participants who signed an employment contract is stated in the 2021 release without a denominator or follow-up period. In 2025 municipalities co-financed 36 placements under a local variant (Table 18 of the 2025 report), which this record does not count. The programme portal (mojaprvaplata.gov.rs) still announces a new cycle of up to 10,000 places with no date, so the halt may be a gap between calls rather than an end; the record documents two years without a national call, not abolition. The state paid the allowance and the employer supplied the training and the workplace, so it is a wage subsidy for trainees as much as a training programme, and some employers may have used it in place of hires they would have made anyway.

• 🇲🇩Moldovalinked to Public-private collaborationstarted 2013dismantled

  Chisinau International Airport concession, and its rescission in 2020

  Moldova's Public Property Agency concessioned the assets of Chisinau International Airport and its land to Avia Invest under contract 4/03 of 30 August 2013, with investment obligations of EUR 244 million over the life of the concession, and in 2020 notified the concessionaire that the contract was rescinded for failing to post the performance guarantee, while Avia Invest claimed EUR 900 million from the state at arbitration in Stockholm.

  244 million EUR, total investment the concessionaire was to make over the life of the concession, 2013-08-30. 900 million EUR, damages avia invest claimed from the state at arbitration in stockholm, 2020-07-22. Agentia Proprietatii Publice (Public Property Agency of Moldova), 'Comunicat: unele detalieri privind rezolutiunea Contractului de concesiune a activelor aflate in gestiunea I.S. Aeroportul International Chisinau si a terenului aferent, nr. 4/03 din 30 august 2013', 22 July 2020 (Internet Archive copy of 21 September 2021), retrieved 2026-10-02.

  How it worked: The state property agency handed the airport's assets and land to a private concessionaire for investment and operation in exchange for fees and investment obligations, then used the contract's performance-guarantee clause to terminate it once a new government contested the deal.

  Limits: The source is one party to a dispute stating its case: the agency wrote the communique to answer press coverage, and the EUR 900 million is the concessionaire's claim as the agency reports it, not an award. The live page now returns 404 on app.gov.md and the record cites the Internet Archive copy. The EUR 244 million is a contractual obligation over the whole concession, not an amount invested; the communique gives no figure for what Avia Invest actually invested in 2013 to 2020, so the record shows the partnership's scale on paper and its end, not its delivery. The communique shows a rescission notice that the concessionaire contested, not the end of its operation: no official page opened shows when the airport returned to the state, and secondary sources (not used as evidence) date the final annulment to a Chisinau Court of Appeal ruling of November 2022. The state enterprise I.S. Aeroportul International Chisinau that appears in the agency's procurement notices held the conceded assets throughout the concession, so its presence does not by itself show a return of operations. The outcome of the arbitration is not covered. The concession was the country's flagship infrastructure PPP: the agency's own list of PPP contracts it signed (archived April 2021) showed a single other contract, for bus stations, signed on 19 December 2018, so Moldova's national PPP portfolio was thin before and after this rescission.

• 🇰🇿Kazakhstanlinked to University-industry collaborationstarted 2019delivered and closed

  Innovation Consortia Matching Grant Program (Fostering Productive Innovation Project)

  Kazakhstan's innovation ministry (MDDIAI), through the Innovation Consortia Matching Grant Program of the World Bank financed Fostering Productive Innovation Project, had supported 33 technology consortia of research teams and firms, including 12 competence centres, by September 2024, whose first sales and signed contracts totalled about 17 billion tenge.

  33 consortia, technology consortia supported under the innovation consortia matching grant program, September 2024. 17,000,000,000 tenge, first sales and signed contracts of productive sector consortia subprojects, September 2024. World Bank, Kazakhstan: Fostering Productive Innovation Project (P150402), Implementation Support Mission Aide Memoire, September 16 to 24, 2024, retrieved 2026-10-02.

  How it worked: The ministry ran a competitive matching-grant window that paid consortia of research teams and firms to take a technology to first sales, with firms co-financing and a project unit monitoring sales and contracts.

  Limits: The figure is the World Bank's mission account of a programme it financed, not a ministry statistic, and the number is in a PDF (paragraph 6 of the aide memoire). The sentence 'supported 33 technology consortia, including 12 competence centers, and 9 IICs' does not make clear whether the nine inclusive innovation consortia are inside or beside the 33. The aide memoire does not define a consortium's membership; the 2020 restructuring paper describes the window as a tool 'for fostering industry-science linkages' with an average 40 percent private co-financing in the first six grants, and the annex shows competence centres hosted by universities (Astana IT University, the ENU-Lab centre), but not every consortium is shown to include a university. The 17 billion tenge is first sales and contracts as reported by beneficiaries, rounded ('around'), and is not audited. The programme ran on loan money: the 2020 restructuring cut the consortia component from US$35 million to US$25.1 million, and in 2024 the ministry asked to cancel a further US$5.4 million of unused consortia grants because no new round could be run before closing (extended to June 30, 2025), so the status is concluded and a follow-on (FPIP-2) was only in preparation. The 2019 start is the year the first consortia grant agreements were signed according to the 2020 restructuring paper.

• 🇦🇲Armenialinked to Public-private collaborationstarted 2000still operating

  Water management contracts and leases, and the national lease from 2017

  Armenia ran a sequence of water public-private partnerships from 2000, a Yerevan management contract, a 2006 to 2016 Yerevan lease and management contracts for the national and three regional utilities, which raised daily supply from about 6 hours in 1998 to 24 hours in most of Yerevan by 2015, and from January 2017 consolidated them into a single 15-year national lease covering about 2.2 million people.

  2,200,000 people (about two-thirds of the population), people in the area covered by the 15-year national water lease from january 2017, 2017. 6 hours per day (24 in most of Yerevan by 2015), average daily water supply in most parts of the country, including yerevan, before the ppps, 1998. World Bank and PPIAF, Review of Armenia's Experience with Water Public-Private Partnerships (Marin, Muzenda and Andreasyan), 2017, retrieved 2026-10-02.

  How it worked: The government contracted private operators in stages, first to manage the Yerevan utility for a fee with performance incentives, then to lease it, then to manage the national and regional utilities, and finally to lease all of them as one national concession paying a lease fee to the state, while the state kept financing capital works.

  Limits: All figures are in the executive summary of a PDF (printed pages ix to xi). The review was written by World Bank staff with support from PPIAF, a facility that promotes infrastructure PPPs, and credits the operators generously, and the 24-hour figure is 'most of Yerevan', not the country. The same review says non-revenue water barely improved, that tariffs outside Yerevan stayed below cost recovery when the management contracts ended in 2016, and that capital investment was financed by the government, so the private operators ran networks without financing them. The record's newest verified fact is the start of the national lease in January 2017; we did not find an official account of its performance since then at an open URL, so 'operating' rests on its 15-year term, and any later dispute, renegotiation or early termination is not covered. The national lease introduced a single national tariff, a cross-subsidy from Yerevan to the rest of the country, whose effect on supply hours outside Yerevan is not shown here.

• 🇦🇿Azerbaijanlinked to Public-private collaborationstarted 2023still operating

  Private solar and wind plants under government investment agreements

  Azerbaijan's government brought its first utility-scale private solar plant, Masdar's 230 MW Garadagh plant built with US$262 million of foreign investment, into operation on 26 October 2023, followed on 16 September 2026 by the 100 MW Gobustan plant won at a 2024 auction, with further private wind and solar plants under investment agreements and power purchase contracts.

  230 MW, capacity of the garadagh solar power plant, the first private utility-scale plant, in operation, 2023-10-26. 262 million US$, foreign investment in the garadagh plant, 2023. Ministry of Energy of the Republic of Azerbaijan, 'Use of renewable energy sources in Azerbaijan', retrieved 2026-10-02.

  How it worked: The government signed investment agreements, land leases, power purchase, grid connection and balancing agreements with foreign developers, first bilaterally with Masdar and then through an auction, for private plants connected to the national grid.

  Limits: The ministry describes its own programme on an undated web page that has been updated at least to September 2026; the investment figure and capacities are the ministry's, with no audited cost and no published tariff. Two private plants are operating, 330 MW together by our sum, against 2,170 MW of renewable capacity the same page reports in operation, most of it state hydropower; the rest of the private pipeline (ACWA Power's 240 MW Khizi-Absheron wind farm, bp's 240 MW Shafag plant, Masdar's 1 GW 'mega project', two 50 MW plants in Jabrayil) is under construction or planned, and the page gives no completion dates. 'Implemented jointly with local and international partners' covers different structures: Garadagh and Gobustan are built and owned by foreign developers, though the page does not name their buyer or contract terms, Shafag serves bp's own terminal through a 'virtual' transmission swap, and the page does not say which plants were competitively tendered beyond Gobustan. The collaboration is recent, so the record shows a start, not a sustained programme. The ministry's page describes Garadagh as 'put into operation as a pilot project'; it is a 230 MW grid-connected plant, so we read the word as first-of-kind rather than as a trial, but the source uses it.

• 🇯🇴Jordanlinked to Public-private collaborationstarted 2015operating below its peak

  Utility-scale renewable power through direct proposals and IPP rounds, and the stop after 2021

  Jordan's Ministry of Energy and Mineral Resources added utility-scale solar and wind plants selling to the grid every year from 2015 to 2021, mostly private projects under direct proposals and competitive rounds, peaking at 345 MWac commissioned in 2020, and then added none: its annual reports put grid-selling renewable capacity at 1,498 MW in 2022 and 2023 and 1,478 MW in 2024.

  0 MW (stock 1,498 MW in 2022 and 2023, 1,478 MW in 2024), utility-scale renewable capacity selling to the grid added from 2022 to 2024 (computed from the reported stock), 2024. 345 MWac, utility-scale renewable capacity added in 2020, the peak year, 2020. Ministry of Energy and Mineral Resources of Jordan, Annual Report 2024, retrieved 2026-10-02.

  How it worked: The ministry offered sites and long-term power purchase agreements with the state-owned single buyer NEPCO to private developers through direct proposals and competitive rounds, alongside donor-financed public plants, and added no utility-scale plant after 2021.

  Limits: The 2024 annual report's year-by-year table, 'Renewable Energy Power Generation Projects: Direct Proposals and Grants', runs over PDF pages 13 to 15 with annual totals on page 15 (134.5, 270, 14, 339.1, 235.1, 345 and 135.5 MWac for 2015 to 2021, under a note that the 2015 to 2021 list 'remain[s] as stated'); the 2021 total is a misprint for 153.5, the sum of its three projects and the figure in the 2022 report. The zero is ours, not printed: it rests on the grid-selling capacity the ministry reports, 1,498 MW in the 2022 report (https://www.memr.gov.jo/ebv4.0/root_storage/en/eb_list_page/annual_report_2022.pdf) and the 2023 report (same path, annual_report_2023.pdf) and 1,478 MW on PDF page 12 of the 2024 report, and on the 2022 report's project table also stopping at 2021; the 20 MW fall in 2024 looks like reclassification (a new 10 MW biogas and hydro line) rather than a closure. The 345 MWac peak is a commissioning year for plants contracted in earlier rounds, so it overstates the pace at which new partnerships were being signed in 2020. The table mixes private independent power producers (Masdar's Baynounah 200 MW, ACWA's Mafraq plant, the Tafila wind farms) with plants financed by Gulf, EU and Spanish grants and loans, including an Air Force plant, so not every megawatt is a partnership; we did not separate them. The reports do not say why additions stopped. What kept growing was subscriber-owned capacity under net metering and wheeling, 1,079 MW in 2022 and 1,320.8 MW in 2024, which is private investment without a public counterpart contract. The erosion is in the frequency of new joint deliveries, not in the stock, which still operates. The ministry reports on its own programme, and the stop may be a pause rather than an end.

• 🇰🇿Kazakhstanlinked to Public-private collaborationstarted 2015still operating

  National and local PPP and concession contracts

  By the end of 2025 Kazakhstan's central and local governments had signed 1,257 PPP and concession contracts worth 3.6 trillion tenge, of which 697 were in implementation, 1,240 were local projects mostly in social services, health, education, energy and utilities, and 17 national projects accounted for more than 1.8 trillion tenge.

  1,257 contracts, ppp and concession contracts signed, cumulative, 2025. 697 projects, ppp and concession projects in implementation, 2025. Kazakhstan Public-Private Partnership Center (national operator for PPP), briefing on PPP results for 2025, 'From the number of contracts to the quality of projects', retrieved 2026-10-02.

  How it worked: The state set up a national PPP operator and let regional akimats sign PPP contracts for local social infrastructure, with the operator appraising projects, ranking regions each year and standardising contract documents.

  Limits: The figures come from the state-owned PPP Center reporting on the portfolio it promotes, at a briefing of the government's Central Communications Service; the page carries no date beyond 'results for 2025' and no table, and the Center's own homepage gives a different, later count (1,438 contracts and 3.8 trillion tenge as of 1 September 2026) that we could not reconcile with the briefing's 1,257. The gap between 1,257 signed and 697 in implementation is not explained: it mixes completed, not yet started and terminated contracts, so the share of contracts that failed is unknown. Most contracts are small local projects in social sectors, and the Center's PPP White Book describes availability payments from the budget as a standard repayment route, so many contracts may be deferred public procurement rather than shared risk; the briefing gives no split by payment mechanism. The briefing itself says the number of new contracts is falling while average project size rose from about 6 billion tenge before 2022 to 18 billion in 2022 to 2025, a policy shift away from the small local contracts that built the count. The start year is that of the Law on Public-Private Partnership of 31 October 2015, as cited in the Center's PPP White Book; the cumulative count also includes concessions signed under earlier law.

• 🇷🇺Russialinked to Public-private collaborationstarted 2019still operating

  Faster Payments System (SBP), run by the Bank of Russia with the banks

  The Bank of Russia's Faster Payments System, which commercial banks connect to so that customers can pay any account by phone number or QR code, processed more than 3 billion operations worth 14.4 trillion rubles in 2022 with 213 banks connected, and 18.3 billion operations worth 103 trillion rubles in 2025.

  18.3 billion operations, operations made through the faster payments system in the year, 2025. 3 billion operations (more than), operations made through the faster payments system in the year, with 213 banks connected, 2022. Bank of Russia, Faster Payments System key indicators, Q4 2025, retrieved 2026-10-02.

  How it worked: The central bank built a national instant-payment switch addressed by phone number, capped what banks may charge individuals, and had commercial banks connect it to their own apps, so that a transfer or QR payment works across all connected banks.

  Limits: The 2025 figures (18.3 billion operations, 103 trillion rubles, 1.4 and 1.5 times 2024) are in the text of the Bank of Russia's quarterly indicator page; the 2022 figures, the 213 connected banks and the comparison with 2019 to 2021, which dates the start, are on the Q4 2022 page (https://www.cbr.ru/analytics/nps/sbp/4_2022/). The pages are HTML text over charts, with no downloadable series. The central bank owns the system, sets its tariffs and reports on it, so this is the operator grading itself. The collaboration is asymmetric: the Bank of Russia and its payment-card subsidiary NSPK built and run the infrastructure, and banks connect and offer it in their apps; the indicator pages do not say how far connection was voluntary, and the record does not claim it was. Growth after 2022 coincided with the exit of Visa and Mastercard and with sanctions on Russian banks, which pushed users towards domestic rails, so the count reflects a shock as much as the design. The values are nominal rubles in a period of high inflation. Russia's infrastructure PPP and concession programme is not covered by this record.

• 🇹🇳Tunisialinked to Public-private collaborationstarted 2024still operating

  ONAS regional sanitation concessions, and their seven-year procurement

  Tunisia's national sanitation utility ONAS launched procurement of its first two ten-year regional concessions in September 2017; the south-east concession started in June 2024 and served 797,000 people by July 2026, and the North Tunis concession entered into force only on 9 July 2026 after the original consortium had to be reorganised.

  797,000 people, people receiving wastewater collection and treatment from the concessionaire (south-east lot only), 2026-07. 0 agreements, concession agreements signed and effective at the june 2023 status report, against a target of two by december 2020, 2023-06-05. World Bank, Implementation Status and Results Report, Tunisia Sanitation PPP Support Project (P162957), sequence 7, 23 September 2026, retrieved 2026-10-02.

  How it worked: The national sanitation utility bundled the operation, maintenance and rehabilitation of whole regional networks and treatment plants into ten-year performance-based concessions, procured with IFC transaction support and paid from a World Bank loan, replacing short task-based management contracts.

  Limits: The 797,000 and the North Tunis entry into force are on pages 2 and 3 of the September 2026 status report, which says the count derives from the south-east lot only and is 92 percent of that lot's end target; the end target for both lots is 2,028,000 people by July 2030. The south-east concession with SCAST was 'scheduled to start on Jun 10, 2024' in the June 2024 status report, and the 783,000 beneficiaries the September 2026 report gives for February 2026 confirm it did. The zero signed agreements against a December 2020 target are in the June 2024 status report (https://documents.worldbank.org/curated/en/099061224041597450/pdf/P162957-61a112cc-f5f3-4ef2-ac1e-0e3c14a54b06.pdf, which also attributes the North Tunis delay to the reorganisation of the ADP-led consortium). The 2017 launch is in the project appraisal document of November 2022, paragraph 13. The World Bank finances the operators' payments through this loan (the status report lists the operation and maintenance component at a cost of 415.30, in US$ millions), so the concessions are public money paying private operators under performance contracts, not private finance, and the lender is reporting on its own project. 'People receiving services' counts the population in the concession area, not households newly connected: only 3,400 people had gained access to improved sanitation through the project by July 2026. The record shows a partnership delivered at national-utility scale but nine years after procurement began, which is as much evidence of slow delivery as of collaboration.

• 🇳🇿New Zealandlinked to Public-private collaborationstarted 2010still operating

  Ultra-Fast Broadband programme

  The Crown's infrastructure investment company, now National Infrastructure Funding and Financing, co-funded a fibre network built and owned by Chorus, Northpower, Enable and Tuatahi First Fibre, which was completed in December 2022 covering 87 percent of the population in 412 towns and cities, with NZ$1.75 billion of government funding in a programme costing more than NZ$5.5 billion.

  87 % of population, population covered by the ufb fibre network at completion, 2022-12. 1.75 billion NZ$, government funding invested in ufb since 2010, of a total cost exceeding nz$5.5 billion, 2025. National Infrastructure Funding and Financing (formerly Crown Infrastructure Partners), Fibre programme page, retrieved 2026-10-02.

  How it worked: A Crown-owned company invested in regional fibre companies on the condition that they build an open-access, wholesale-only network to a fixed rollout plan, and recovered its investment as the companies sold connections, recycling the repayments into further extensions.

  Limits: The Crown company that ran the programme reports on itself, and the page is promotional ('world leading', 'on time and under budget'); no audit figure for cost against the original budget was checked. The 87 percent is coverage (premises passed), not use: the same page gives uptake as 73 percent of UFB premises passed at 30 June 2025. The total cost above NZ$5.5 billion includes partner co-funding, so the private share is inferred, not stated, and NZ$390 million of the Crown's investment was repaid and recycled into extensions, which the page treats as reinvested rather than spent twice. The page does not describe how partners were chosen or on what terms; the corporate history (Crown Fibre Holdings, then Crown Infrastructure Partners, then National Infrastructure Funding and Financing) and the role of the incumbent's structural separation are background not shown on this page. The page has no date stamp; the uptake figure dates it to after June 2025.

• 🇧🇩Bangladeshlinked to Public-private collaborationstarted 2010still operating

  Public Private Partnership Authority project portfolio

  Bangladesh's Public Private Partnership Authority listed 81 projects in its portfolio as last updated on 19 August 2026, of which four were in operation (including Patenga Container Terminal, the NIKDU and Chittagong haemodialysis centres and two Info-Sarkar 3 rural broadband components), three were partly operating while still under construction (among them the Dhaka Elevated Expressway), and 10 were under construction.

  7 of 81 portfolio projects (4 operational, 3 partial operation and construction), ppp projects in operation, fully or partly, 2026-08-19. 10 projects, ppp projects under construction, 2026-08-19. Public Private Partnership Authority Bangladesh, PPP projects list, retrieved 2026-10-02.

  How it worked: A dedicated authority under the head of government screens line-ministry projects, runs procurement with technical assistance and viability gap funds, and takes them to the economic cabinet committee for approval and award.

  Limits: The counts are ours, tallied from the stage column of the authority's project list, which the page says was last updated on 19 August 2026 and which was read on 2026-10-02; the page gives no totals, and it reports stages, not dates of financial close or of operation. Two of the four operating entries are components of one rural broadband project (Info-Sarkar 3), so the operating count overstates distinct partnerships by one. The authority promotes its own portfolio. The headline is how little has reached operation: 7 of 81 projects after some 16 years, with 24 still at feasibility study and 6 at conditions-precedent stage, so the record documents scale of activity more than delivery. Several listed projects are land leases for commercial development (shopping malls, textile mills on state land), not public services. The Dhaka Elevated Expressway opened in parts against long delays. A 2026 Invest Bangladesh Act appears on the same site and may reorganise the authority; its Bangla text could not be read in this session, so the record does not say whether the authority continues in its current form.

• 🇱🇦Laoslinked to Public-private collaborationstarted 2005still operating

  Nam Theun 2 hydropower concession

  The Government of Lao PDR granted a build-operate-transfer concession for the 1,070 MW Nam Theun 2 plant to the Nam Theun Power Company, in which it held 25 percent through Lao Holding State Enterprise; the plant was completed in 2010 in 58.5 months, 4.5 months over plan, at US$1,308 million against a US$1,250 million base estimate (US$1,450 million with contingencies), and paid the government US$180 million in net revenues from 2010 to 2017.

  180 million US$, net revenues from nt2 to the government of lao pdr, 2010 to 2017, 2017. 1,070 MW, installed capacity delivered (995 mw for export to thailand, 75 mw for the lao grid), 2010. World Bank Independent Evaluation Group, Implementation Completion Report Review, Lao PDR Nam Theun 2 Social and Environment Project (2019), retrieved 2026-10-02.

  How it worked: The government granted a concession running to 2034 to a special-purpose company, took a minority stake through a new state holding company financed by multilateral loans, and relied on a long-term export power purchase agreement with a Thai buyer that made the debt bankable, with royalties, taxes and dividends earmarked to poverty programmes under a revenue management agreement.

  Limits: The World Bank financed and guaranteed the project (an IDA grant and a US$50 million partial risk guarantee), so the review is the lender's evaluation unit grading a flagship; it rates the outcome Satisfactory. The figures are in a PDF (cost and financing in the project data section, completion time and capacity and the revenue figures in the efficacy section). The US$180 million is net of nothing the review can confirm: it notes that the US$30 million a year target should be reduced for the government's repayments of the ADB and EIB loans that bought its 25 percent stake, information the Bank team could not provide, so the true net figure is uncertain. Revenue was about US$25 million a year, below the appraisal's US$30 million for the early years, and the much larger US$110 million a year expected after 2020 is a projection, not a result. Equity came from EDF, Italian-Thai Development and EGCO, debt from 14 commercial banks and multilaterals; the delivery is a foreign-led consortium that the state contracted and part-owns, and most of the output is exported to Thailand rather than serving Lao consumers. The resettlement and environmental programme, which the project was also judged on, is not covered by these numbers. The start year, 2005, is our inference from completion in 2010 after 58.5 months. The review also records that by 2014 a significant portion of NT2 revenues were being deposited directly into the Treasury rather than through the agreed special account, which weakened the earmarking of revenue to poverty programmes.

• 🇹🇿Tanzanialinked to Public-private collaborationstarted 2007dismantled

  Central railway concession to the RITES-led Tanzania Railways Limited, and its collapse in 2011

  Tanzania concessioned the central railway network for 25 years to Tanzania Railways Limited, 51 percent owned by a consortium led by Rail India Technical and Economic Services, which took over operations in October 2007; freight fell from 470,000 tons in 2008 to 237,000 tons in 2010, and after a dispute the RITES shares were transferred to the government in 2011, leaving the operator fully state owned four years into the term.

  237,000 tons, freight carried by tanzania railways limited on the central (metre gauge) network, 2010. 470,000 tons, freight carried by tanzania railways limited in its first full year of concession operation, 2008. World Bank, Project Appraisal Document, Tanzania Intermodal and Rail Development Project (P127241), Report No. 78925-TZ, April 2014, paragraphs 14 and 16, retrieved 2026-10-02.

  How it worked: The government unbundled the railway into an asset-holding company and an operating company, sold 51 percent of the operator to a foreign rail consortium on a 25-year concession, and kept the track; when traffic, investment and maintenance fell short, the state bought back the private shares and ran the operator itself.

  Limits: The figures are in a PDF, on pages 16 and 17 of the appraisal document (printed pages 4 and 5), written by the World Bank as it prepared a new loan to the now state-owned operator, so the publisher had an interest in describing the concession as a failure. The PAD gives no date in 2011 for the share transfer and calls it amicable, and it is not consistent on the year: paragraph 16 (page 17) puts the transfer in 2011 and an annex (PDF page 116) says TRL took over in 2011, but another annex paragraph (PDF page 117, Annex paragraph 28) says the concession was terminated in 2010; the concession was not formally terminated by a court or an arbitral award in this source, so what ended was private participation, not the railway. The earlier completion report for the Central Transport Corridor Project (ICR00001299, June 2010, page 9) records the concession as performing poorly and the parties discussing termination, and adds (paragraph 94) that most of the traffic had already moved to road during the two years of bidding delay, from 1.2 to 0.5 million tons a year, before the concessionaire took over: the decline therefore began under public operation. The 2008 to 2010 fall also includes the 2009 to 2010 floods in the Kilosa to Gulwe section, which the PAD says halted services between Dar es Salaam and Dodoma, so not all of it is the concessionaire's doing. 2008 is the first full year of concession operation, not a demonstrated peak. The 1.5 million tons TRC carried in 2002, which the PAD also gives, is the public railway before the concession and is not used as the peak here.

• 🇿🇲Zambialinked to Public-private collaborationstarted 2003dismantled

  Zambia Railways concession to Railway Systems of Zambia, and its cancellation in 2012

  Zambia concessioned its national railway, after a decade in which traffic halved from 1,072 million net ton kilometres in 1993 to 540 million in 1999, to the private operator Railway Systems of Zambia, which took over on 3 December 2003 and raised traffic by about 30 percent in net ton kilometres in 2004 against 2003; the World Bank later reported that the network and service kept deteriorating and that the concession was cancelled in September 2012.

  30 %, increase in rail freight traffic (net ton kilometres) in the first year of the concession, 2004 against 2003, 2004. 3.7 million US dollars, concession fees paid to government by the concessionaire since takeover, per the concessionaire's auditors, 2005 (third quarter). World Bank, Implementation Completion Report, Zambia Railways Restructuring Project, Report No. 32520, December 2005, retrieved 2026-10-02.

  How it worked: After two decades of falling traffic under the state railway, the government retrenched staff with World Bank money and let one private consortium run all freight and passenger services and the track under a vertically integrated concession; the operator lengthened hauls and paid fees for a few years, then a later government judged the network was still deteriorating and took the railway back.

  Limits: The cited completion report was written two years into the concession and supports the start and the early gains only; it says the 30 percent rise came with a fall in tonnage as the operator dropped short hauls, and that the borrower disputed the concessionaire's fee compliance and Parliament set up a review in December 2005. The cancellation rests on a second World Bank document, the Country Partnership Strategy for FY13 to FY16 (Report No. 75089-ZM, February 2013, http://documents.worldbank.org/curated/en/182591468178437607/pdf/750890CORRIGEN0IC00IDA0R20130002703.pdf, PDF page 22, paragraph 49), which says the rail network condition and service continued to deteriorate and the concession was cancelled in September 2012 with a commitment to recapitalise the railway, and elsewhere (page 11) calls it a reversal of the railway concession that stakeholders saw as policy incoherence. Neither document gives traffic for the later concession years, so the record shows a delivered early gain and a cancellation, not a measured decline from a peak. Both documents come from the lender that financed the concession process and argued for privatisation. The figures are in PDFs (the 1993 and 1999 traffic on page 5, the fees on page 10, the 2004 traffic gain on page 19, the 20-year term on page 43). The concession was originally 20 years with possible extensions (Annex 10), so it ran about nine years of a 20-year term, a figure computed here.

• 🇬🇳Guinealinked to Public-private collaborationstarted 1989dismantled

  Urban water lease (affermage) with SEEG, and the return to public management after 2000

  Guinea ran its urban water service from 1989 under a ten-year lease in which the public asset company SONEG owned and financed the network and the privately led operator SEEG ran it, but after renegotiation failed the private operator withdrew in December 2000, the government decided in July 2003 to keep the service public, and revenue collected from private customers fell from US$5.8 million in 1998 to US$1.7 million in 2005.

  1.7 million US dollars, revenue collection from private customers, urban water services, 2005. 5.8 million US dollars, revenue collection from private customers in the last full years of the lease, 1998. World Bank, Implementation Completion Report, Guinea Third Water Supply and Sanitation Project, Report No. 36083-GUI, June 2006, section 4.1, retrieved 2026-10-02.

  How it worked: The state split the urban water utility into a public asset-holding company that owned and financed the network and a private-led operating company that ran it under a lease and was paid from tariffs; when the lease came up for renewal, tariffs were frozen, the private shareholder saw too much risk and left, and the public company took over operations it was not prepared to run.

  Limits: The figures sit in a PDF (page 9, printed page 4; the ten-year lease ending in 1999 is on page 6). The World Bank financed the lease from its first project in 1989 and rated this follow-on project and the borrower unsatisfactory, so it has a view on who was to blame. The revenue figures are in US dollars, and the report says tariffs were frozen until 2005 and fell from US$0.86 to US$0.21 per cubic metre in real dollar terms, so much of the fall in dollar collections is a currency and tariff effect, not only worse collection; the same paragraph says SONEG's net income went from US$2.9 million in 1998 to a loss of US$0.5 million in 2005. 1998 is a year chosen by the report, not a demonstrated peak. The report also says 24 hour service and compliance with water quality standards, achieved under the earlier operation, were lost and service became intermittent, with no figure for when. SEEG's private shareholder is named only as SAUR, in passing, and the share structure is not given. The 1989 start is inferred from the ten-year lease ending in 1999 and the first supporting credit of May 1989; the report gives no signing date. The lease ended at the end of its term after two years of failed renegotiation and contract extensions, so this is a partnership that was not renewed rather than one terminated for breach.

• 🇲🇿Mozambiquelinked to Public-private collaborationstarted 2000dismantled

  Maputo water lease to Aguas de Mocambique, and the public buyout of the operator in 2010

  Mozambique's water asset fund FIPAG leased Maputo's water service from 2000 to the private operator Aguas de Mocambique, whose lead shareholder SAUR pulled out in December 2001 and offered its stake to Aguas de Portugal for one euro; under the lease Maputo connections rose from 77,793 at the end of 2000 to 109,932 in March 2009, and in 2010 FIPAG bought out the Portuguese shareholder, leaving the operator primarily publicly owned.

  109,932 connections, maputo water connections served by the lease operator at project completion, 2009-03-31. 77,793 connections, maputo water connections at the start of the lease (baseline), 2000-12-31. World Bank, Implementation Completion and Results Report, Mozambique Second National Water Development Project, Report No. ICR0000700, September 2009, retrieved 2026-10-02.

  How it worked: The state kept the water assets in a public fund and leased operation of the capital's system to a private consortium paid from tariffs, with a regulator able to order tariff reviews; when the lead foreign operator left and its successor underperformed, the asset fund bought the operating company itself and kept the lease structure with a public owner.

  Limits: The connection counts are in a PDF (page 7, the results table; the SAUR withdrawal on page 18, paragraph 22), reported by the lender that designed and financed the lease. The end of private participation rests on a second World Bank document, the completion report for the Greater Maputo Water Supply Expansion Project (Report No. ICR00005527, February 2022, http://documents.worldbank.org/curated/en/306811653587372748/pdf/Mozambique-Greater-Maputo-Water-Supply-Expansion-Project.pdf, page 22, paragraph 37), which says the operator's performance fell well below expectations, relations with FIPAG deteriorated, FIPAG bought out the Portuguese partner's roughly 85 percent stake in 2010, and the renamed company, Aguas da Regiao de Maputo, has run the lease since then as a primarily public company, with the lease extended to 2019 and private partners for the regional utilities only scheduled for 2022. The two reports do not agree on the lease length (15 years in 2009, 14 years in 2022). The record shows growth in connections under the lease and the end of private ownership, not a fall in service: connections were not shown to decline after the buyout, so what eroded is the partnership, not access. The 2009 report rates the private operator's performance poorly through much of the period: the 2000 floods, SAUR's underbidding, and two years of renegotiation delayed the programme by about four years, and supply in 2009 averaged 14 hours a day against a target of continuous supply. Lease revenue to FIPAG stayed far below expectation (2022 report, paragraph 52).