Global Cooperation in Science, Technology & Innovation for Development

The cooperation imperative

Scientific research increasingly takes place among global teams of researchers, creating networks that extend well beyond national institutions and single disciplines (Wagner et al., 2015). The growing technological complexity, the speed of change and the sheer scale of transformation brought by recent waves of innovation all demand a collaborative approach. Choices about climate change, artificial intelligence and biotechnology will affect every country on earth, yet the capacity to shape those choices is concentrated in a handful of economies.

The numbers are stark. Between 2018 and 2021, developed countries saw their total exports of green technologies grow from roughly $60 billion to $156 billion, an increase by a factor of 2.6. Developing countries managed a factor of 1.3, from $57 billion to $75 billion. Developed countries are capturing most of the opportunities created by Industry 4.0, AI and green technologies, putting developing countries at risk of being left further behind, as happened in every previous wave of technological revolution. The 17 frontier technologies examined in UNCTAD's Technology and Innovation Report 2023, from AI and IoT to solar PV, green hydrogen and electric vehicles, represent markets projected to grow from $1.5 trillion in 2020 to over $9.5 trillion by 2030. Who captures that value will depend on who builds the scientific and industrial capacity to participate.

Four pillars of STI development

The traditional National Innovation System (NIS) framework, while useful for conceptualising the factors that shape a country's innovative capacity, no longer suffices as a practical guide for STI policy. The current technological paradigm demands approaches that look beyond national actors and account for the increasingly dense web of international linkages through which knowledge, technology and talent flow. The paper proposes a framework built around four key elements, each with distinct internal dynamics and external connections.

Strategic planning concerns how governments set direction: national STI agendas, multi-year plans, frameworks for technology foresight and assessment. The paper draws on examples like Ghana's "STI for SDGs Roadmap" and UNCTAD's own STI Policy Review of Angola, which maps and benchmarks the innovation ecosystem by combining data analysis with stakeholder interviews. The key insight is that strategic planning requires two inputs many developing countries lack: an understanding of global STI trends, and a clear-eyed assessment of the country's own strengths and weaknesses.

STI prerequisites cover the tangible and intangible resources that must be in place before cooperation can work: physical and digital infrastructure (stable electricity, broadband, cloud computing), human capital (STEM skills, digital literacy, managerial capability) and financial resources. The current economic paradigm requires not just mobile networks but affordable internet with high bandwidth and low latency. The digital revolution makes mastering STEM skills crucial, from coding and data analytics to the interpersonal and collaborative competences needed in hybrid work environments (Marr, 2022; WEF, 2016). The lack of appropriate skill sets in government directly results in insufficient representation of technical expertise in legislative and regulatory processes (UNCTAD, 2021c), limiting the capacity to design effective STI policy.

Research and development involves systematic work to increase knowledge and devise new applications. R&D is characterised by high uncertainty, long time horizons and cumulative complexity. The average R&D investment of low middle-income countries sits at about 0.53% of GDP, far below the world average of 2.63% (UNCTAD, 2023a). For many developing countries, closing this gap without external support is essentially impossible. Only about 31% of innovating firms in developing countries invest in R&D activities (Cirera et al., 2020); in developed countries too, a large share of innovative firms do not introduce products new to the market.

Innovation is the practical implementation of ideas into goods, services or processes that reach the market. It is distinct from invention. Many promising R&D projects never survive the "Valley of Death" (Hudson and Khazragui, 2013), the gap between lab success and commercial viability, for two main reasons: the disconnect between public research funding and private investment, and the skills needed to identify market needs and translate technical ideas into business propositions. Stronger public-private partnerships, test environments mimicking real-life conditions, and international open testing platforms could all help bridge this gap, especially for less endowed countries.

Strategic planning: who sets the agenda?

The formulation of the international STI agenda has been historically skewed towards the perspective of developed countries. This is not a conspiracy; it is the natural consequence of superior technological capabilities and the capacity to manage extensive networks (Lundvall et al., 2002). The cumulative effect is a system where the research questions that get funded, the standards that get set and the regulations that get adopted overwhelmingly reflect the priorities of countries already at the frontier.

The paper identifies three dimensions where international cooperation in strategic planning matters most. The first is building an inclusive international STI agenda. Collaborative activities can take many forms, but a key insight from the literature on cognitive distance (Nooteboom, 2008) is that the more focused the topic, the lower the cognitive gap among participants and the easier it is for less advanced countries to engage meaningfully. This is why regional organisations have been important: ASEAN's APASTI 2016-2025 developed policies through a common vision with specific timelines and key performance indicators, establishing four new science and technology networks (including an ASEAN Foresight Alliance) and a shared high-performance computing facility. The African Union's STISA-2024, anchored on six priority areas aligned with the AU Vision, offers a regional blueprint, though progress in implementation has been generally slow due to insufficient monitoring, low investment and inadequate budgets.

Two models of inclusive global collaboration stand out. CERN, the European Organization for Nuclear Research, adopts a partnership-oriented approach in which members are elected solely on scientific merit and without reference to nationality. It employs a light leadership style and consensual governance to manage multi-polarity and avoid gridlocks in international cooperation (Robinson, 2019). The scientific spirit, rather than political alignment, drives the agenda. The Consultative Group on International Agricultural Research (CGIAR), operating through 15 research centres with over 3,000 partners in nearly 90 countries, moved in 2019 toward a unified governance structure and pooled funds after recognising that its fragmented institutional design was limiting its impact as food, land and water issues became increasingly interconnected.

The second dimension is technology foresight and assessment. Small and less developed countries often lack the critical mass to engage in the technological forecasting exercises that inform strategic planning. The paper calls for a multilateral system, coordinated through the CSTD, that could consolidate different foresight exercises and make their findings accessible to countries that cannot produce them independently. The accelerating pace of AI development, with concerns about misinformation, IP infringement, data privacy and bias (UNCTAD, 2023c), makes this particularly urgent: the ethics guidelines, regulatory frameworks and governance structures being debated today will shape how frontier technologies are deployed worldwide.

The third dimension is supportive international rules. The TRIPS agreement's Article 66.2 requires developed countries to provide incentives for technology transfer to least developed countries, and the 2022 WTO Ministerial Decision allowing eligible members to produce COVID vaccines without patent holder consent demonstrated what flexible IP can achieve. Extending similar flexibilities to environmentally sound technologies would make the multilateral trade regime more consistent with international climate agreements. The Trade and Environmental Sustainability Structured Discussions at WTO have acknowledged that weak infrastructure and lack of access to technology are barriers to greater adoption of green technologies in developing economies.

Prerequisites: infrastructure, skills, money

Cooperation requires capable partners. The paper dedicates substantial attention to the foundational conditions without which no amount of networking will produce results. Digital infrastructure comes first. It is no longer enough to have mobile networks; the current paradigm demands affordable internet with high bandwidth and low latency to support IoT, AI and cloud computing. Private firms are investing heavily: Google's "Firmina" subsea cable connects the US East Coast to Argentina to improve Latin American connectivity; Microsoft's Airband initiative has helped more than 51 million people globally gain internet access. Regional programmes like ESCAP's Asia-Pacific Information Superhighway and the Programme for Infrastructure Development in Africa are fostering coordinated approaches. The Broadband Commission for Sustainable Development, established by ITU and UNESCO, has promoted universal connectivity, while the Alliance for Affordable Internet has driven internet prices down in low- and middle-income countries through policy and regulatory reforms.

Skills matter at every level. Between 2014 and 2020, the global share of female researchers increased from 28.8% to 31.2%, still far from parity. The paper highlights programmes that work: Portugal's FCT partnership with American universities (involving over 1,000 teachers and researchers and 1,500 graduate students), the World Bank's Africa Higher Education Centers of Excellence (43 centres across the continent), the Marie Skłodowska-Curie Actions in Europe, and UNCTAD's own Young Female Scientist and Young Scientist PhD programmes with Okayama University. The common thread is that successful capacity-building programmes are designed around the specific needs of participants, not around what donors find convenient to offer. The UN interagency task team on STI for the SDGs (IATT) has been particularly important in designing training courses on STI policy and maintaining a global repository of training materials.

Then there is money. In 2022, total ODA from DAC member countries amounted to $204 billion, about 0.36% of their combined GNI and well below the UN target of 0.7%. The share dedicated to STI has been fluctuating around 1% for two decades and actually declined in both 2020 and 2021 after a peak of 1.4% in 2019, even as the COVID pandemic demonstrated how critical scientific capacity is. Medical research ODA fell from 0.35% to 0.26% during a global health emergency. The paper contrasts the ODA-for-STI share of major donors with their domestic R&D intensity: the US and Japan spend less than 0.2% of their ODA on STI while investing over 3% of GDP in domestic R&D. The UK (2.76% of ODA for STI) and France (1.16%) show considerably stronger engagement. The main takeaway is blunt: relatively small reallocations of ODA budgets could make a significant difference. If channelled toward collaborative projects, they would simultaneously strengthen the inclusion of developing countries in international research and innovation networks.

Research and development: who invents, who borrows

The patent landscape speaks volumes about the distribution of inventive capacity. In low-income countries, more than two out of every three patents are filed by foreign actors. In lower-middle-income countries the figure is about 60%. High-income countries, by contrast, see 56% of patents filed by residents. And the volume gap is staggering: high-income countries filed 1.55 million patents in 2020 compared to just 1,900 in low-income countries. This is not merely a measure of openness to foreign investment. It reflects the weakness of national actors in developing patentable inventions and the degree to which the knowledge base in these economies is shaped from outside.

The paper reviews a wide range of research funding mechanisms. Horizon Europe, the EU's 9th framework programme, allocates roughly EUR 95 billion for 2021-2027 and has an explicit collaborative design: a large share of the budget goes to projects with at least three partners from three different EU countries, and recent editions have opened access to non-EU countries. The European Cooperation in Science and Technology (COST) funds bottom-up, open and inclusive networks for four-year periods, providing a reasonable time horizon for research. The Green Climate Fund aims to channel resources toward climate adaptation in developing countries but struggles to attract funding from donors who prefer bilateral channels (Kumar, 2015). The Bill and Melinda Gates Foundation offers a different model: flexible IP arrangements allowing pharmaceutical companies to profit from drugs in developed markets if they sell at marginal cost in developing ones, combined with a milestone-driven oversight approach rather than annual reporting.

International research networks range from EUREKA (the world's biggest public R&D cooperation network, present in 45 countries, with over EUR 560 million in public-private investment in 2022) to the Ibero-American CYTED programme (25,000 researchers and 1,000 firms across 21 Spanish- and Portuguese-speaking countries) to the targeted CEPI and ACTIV platforms launched during COVID. The International Science Council has been recognised as one of the initiatives most successfully mobilising the academic community toward the SDGs (Dibbern and Serafim, 2021). Open data platforms like the Global Biodiversity Information Facility provide standardised data-sharing infrastructure with clear rules for authorship and attribution. But the paper repeatedly stresses that openness without capacity is empty: if a country's researchers can access the literature but lack laboratories, computing infrastructure or trained personnel to act on what they read, open access becomes a symbolic gesture.

Alternative modes of technology creation are also gaining traction. The open-source paradigm (Linux, MySQL, Firefox, WordPress) has expanded beyond software: the Human Genome Project's landmark pre-publication data release policy established a greater awareness of data sharing in biomedical research. Crowdsourcing platforms like InnoCentive allow companies to post R&D challenges to a global talent pool. The NASA Space Apps Challenge, the largest annual global hackathon, registered over 30,000 participants from 162 countries in 2022 with more than 3,000 projects submitted.

Innovation: from lab to market

The "Valley of Death" between invention and commercial innovation is where most promising technologies fail, and it is steeper in developing countries. The paper identifies three categories of mechanism that can bridge it: technology and knowledge transfer, test beds, and incubators and accelerators.

Technology transfer has traditionally been conceived as a linear flow from advanced to less advanced economies. The paper argues this model is outdated. Transfer is now interactive and networked, and its effectiveness depends critically on the receiving country's absorptive capacity (Cohen and Levinthal, 1990). Wide disparities in capability impede effective transfers, especially in frontier technologies like AI, where talent training programmes and cooperative laboratories are needed to smooth the process. Three UN initiatives are highlighted. The Technology Bank for LDCs works through three pillars: assessing STI ecosystems, identifying appropriate technologies for transfer, and building capacity to ensure sustainability. The Global Environment Facility has allocated more than $22 billion in grants and mobilised $120 billion in co-financing across 5,000 projects in 170 countries. The Climate Technology Centre and Network responds to country-driven requests across five systems transformations (water-energy-food nexus, buildings, mobility, energy systems, industry).

Test beds allow technologies to be refined under controlled conditions before deployment. The EU's Open Innovation Test Beds provide open access to nanotechnology and advanced materials testing. The Industry IoT Consortium offers a global test bed programme and has developed the Industrial Internet Reference Architecture to ensure compatibility across sectors. These platforms could be particularly valuable for developing countries that cannot afford to build their own testing infrastructure, offering access from their geographical location.

Incubators and accelerators help navigate the early-stage journey from idea to business. Y Combinator, the most successful startup accelerator by key metrics (exits, unicorn creation rate, combined valuation), demonstrates the importance of a dynamic feedback environment and robust networks: its co-founder matching platform has made over 100,000 matches across major global cities. Start-up Chile, a public accelerator, allocated over 80% of its budget to foreign entrepreneurs in 2021, showing how a developing country can attract talent to nurture its national innovation ecosystem. The UNDP Accelerator Labs (91 teams covering 115 countries) represent the largest global network dedicated to sustainable development challenges, with a focus on grassroots innovation and local insights. InfoDev's Climate Technology Program operates through Climate Innovation Centers, with the Kenya CIC recognised as the "Most Promising Business Incubator in Africa." UNCTAD's own STI Parks project builds expertise in four developing countries across Africa and Asia through South-South cooperation.

The CSTD's role

The Commission on Science and Technology for Development, established in 1992, is the United Nations focal point for STI for sustainable development. It acts as a forum for strategic planning, foresight, best-practice sharing and consensus building, and is closely linked to the Technology Facilitation Mechanism which champions multi-stakeholder cooperation. The CSTD has generated concrete programmes from its deliberations: the CropWatch Innovative Cooperation programme for agricultural monitoring using satellite data, PhD programmes for young scientists from developing countries with Okayama University, geospatial technology projects for disaster resilience in SIDS, a Bio-Circular-Green economy model partnership with Thailand, and a workshop on harnessing STI for disaster risk reduction.

The paper argues that the CSTD is uniquely positioned to coordinate international STI cooperation because it is the only UN body with a specific mandate on STI for development, it convenes an unusually broad range of stakeholders (governments, international organisations, academia, civil society and the private sector), and its intersessional panels provide a mechanism for sustained dialogue between annual sessions. What it needs is greater institutional weight, stable funding and the capacity to deliver technology foresight at scale. A multilateral system of technology foresight and assessment, coordinated through the CSTD, could offer comprehensive analysis of global STI development, provide directionality to technological change, promote alignment of national, regional and international STI agendas with the SDGs, and foster international collaboration. Strengthening the CSTD would be a high-value, relatively low-cost investment.

Six recommendations

The full issues paper is available at unctad.org. The final publication (UNCTAD/DTL/INF/2024/1) was presented at the CSTD's 27th session in April 2024.

Tools and methods