If a finance ministry in Latin America could move one number, which number should it be? Our answer is the level of research spending. On the UNESCO Institute for Statistics 2026 R&D data release, Latin America and the Caribbean spend 0.57% of GDP on research and development against a global average of 1.92%, with Europe plus North America at 2.55% and East and Southeast Asia at 2.42%, all on a 2023 reference year. In this post: what the gap represents, who funds research here, which instruments already work, and what feeds the pipeline.

The Level Gap Is the Largest Number on the Table

We carry the UNESCO figures as reported rather than verified: the 2026 release covers 43 countries updated from the 2025 survey plus RICYT data, and we have not audited the submissions behind them. Even so, a spending level is a simpler object than a funding split: one budget line, it can be legislated, and it compounds.

A level decides capability. Under a third of the world average is the difference between a research base that can publish and one that can run a semiconductor pilot line. Our report puts regional GDP at roughly USD 6 trillion, a figure extracted and not yet re-verified, and 0.57% of that base pays for every laboratory in the region. One caution for readers building league tables from the ratio: it penalises commodity exporters, whose denominators carry output with little research content, so part of the distance to East and Southeast Asia’s 2.42% is industrial structure. Read the gap as an order of magnitude, not a scorecard.

What sits on the other side of the gap is worth the spend. The region holds 42% of the world’s biodiversity, a figure our report takes from Deep Tech Colombia and consistent with the roughly 40% that UNEP and the World Economic Forum report, and Ember’s Global Electricity Review 2026 puts the region at 65% clean electricity in 2025 against a 41% global average. Research capacity converts endowments like these into companies and exports.

Where the Room Is: Who Pays for Research Today

Composition is the second question, and it is where the room to grow shows. Our report gives two readings. Page 44, citing ECLAC, has businesses covering only about 35% of R&D funding in the region while governments provide roughly 60%. Page 107, citing the IDB in 2023, puts the private share at 43% of total R&D expenditure against 80% in Israel. Both sit in our ledger as extracted and not yet re-verified, they rest on different sources, and the ECLAC series is 2019 data, so read the private share as a range of roughly 35 to 43 per cent and check which source a ranking uses before you calibrate against it.

The range still tells you something. A state-dominated 0.57% buys academic publication and public-lab capacity; a business-dominated 0.57% buys process improvement and patents held by firms that can scale them. On either reading the private share has far to travel, and Israel is the case our report reaches for. Its 1985 R&D Law reimbursed up to 50% of corporate R&D, catalysed over 1,000 projects annually and lifted private R&D spending to 4% of GDP, on IDB figures also extracted and not yet re-verified.

Is the challenge one of increasing the amount of funding, or is it about improving the quality and efficiency of the available resources?

That is our report’s own question, put to the founders and investors it interviewed. Our answer: both, in that order. Composition is the variable an instrument touches, the level is what it aims at.

The Instruments Are Already Running in the Region

The encouraging part: the machinery does not need inventing. Four examples, with figures as our report published them and not yet re-verified:

Chile: CORFO has deployed matching funds since the early 1990s. Through FONTEC, now under the Innova umbrella, it provides reimbursable grants covering 40 to 65% of private R&D costs, while FONDEF co-finances pre-competitive R&D between academia and firms. Photio, whose nanoparticle additives turn urban surfaces into air purifiers, and Strong by Form, which builds structural wood by robotic manufacturing, both carry CORFO Alumni status.

Argentina: the 2017 Entrepreneurs Law established FONDCE, a fund of funds underwriting science-based accelerators and early-stage venture capital with reimbursable loans and partial operating grants, and let founders incorporate in under 24 hours. Within two years, five deep tech accelerators backed by FONDCE had invested in 79% of Argentina’s investor-funded deep tech ventures, on a 2021 source describing the law’s first two years.

Uruguay: Law 20.075, enacted in 2022 and formally approved in 2023, prioritises advanced digital platforms, biotechnology and green tech. The Uruguay Innovation Hub opened in May 2024 with US$10 million allocated for a 1:1 matching-funds programme it manages, in which vetted venture firms and angels co-invest through convertible notes tied to technical and commercial milestones.

Brazil: PIPE at FAPESP is modelled on the United States SBIR programme and funds R&D in small companies based in São Paulo. Mais Inovação, launched in 2024, offers non-reimbursable grants and credit to R&D-intensive companies in health, energy, ICTs and sustainability. FINEP, with BNDES, is offering 500 million dollars to induce multinational and national firms to set up R&D hubs in Brazil.

Our report’s verdict is that the instruments remain too fragmented and limited in scale, and that weaving them into a region-wide playbook is what turns pilots into a mature ecosystem. Domestic capital sits next to the problem: pension assets in Chile, Mexico, Brazil, Peru and Uruguay total roughly US$1 trillion on OECD figures we extracted and have not re-verified, and redirecting even 1% would mobilise USD 10 billion. That is recommendation 12, a fund of funds modelled on France’s Tibi initiative.

The Pipeline That Feeds Any Level of Spending

Money buys researchers, so the researcher stock is the constraint upstream of any level target. Page 23 of our report puts 865,000 STEM researchers in the region on IDB figures, while RICYT, the Ibero-American network of science and technology indicators, reports LAC researchers exceeding 700 thousand in its 2024 edition. Both are headcount rather than full-time equivalents and they differ in vintage and scope, so use the number as a headcount range. The same page calls the researchers outside deep tech a 100x runway of untapped talent, which is our own framing rather than a measurement.

Further upstream, our report cites the IDB on PISA 2022: teenagers in the region perform the equivalent of three to seven school-years behind their OECD peers in mathematics. That row is extracted and not yet re-verified, a newer PISA cycle may supersede it, and the range converts score gaps into school-year equivalents rather than counting years. Hedged that far it still binds: a research base is built from people who can do mathematics. Two training instruments in the report, also unverified, are worth copying: INFOTEP in the Dominican Republic runs on a 1% payroll levy, INEFOP in Uruguay co-finances up to 70% of company training plans.

The base the region already has is productive, the strongest argument for funding more of it. FabNS, a Federal University of Minas Gerais spinoff founded in Belo Horizonte in 2020, builds tip-enhanced Raman spectroscopy instruments at TRL 9 and was named a Deep Tech Pioneer in Hello Tomorrow’s 2024 Challenge for Industry and Machines. M4Life in Argentina is a CONICET spinoff. Symbiomics, in Florianópolis, is the third-largest Brazilian deep tech recipient of 2024 at US$2.7 million on EMERGE’s Radar. Puna Bio turns extremophile microbes from high-altitude deserts into agricultural inputs, backed by the Gates Foundation and SOSV/IndieBio.

What Would Sharpen This Reading

Three things would make this argument more useful, and all three are queued. The private share needs re-deriving from the ECLAC and IDB primary series, so a matching-fund ratio can be set against one number rather than a range. The researcher count needs a headcount-to-FTE reconciliation. And the region still lacks a shared evidence base on business R&D by sector, so a ministry cannot see whether its credits land in deep tech or in incremental product work.

That last gap is why two of our recommendations are infrastructure rather than advice: an open-access data commons with standardised statistics, deal and IP registries and a unified grants portal, and a live funders mapping of the capital active in regional deep tech.

What This Means for Ministries, Allocators and Founders

For ministries, sequencing matters more than the instrument. The level gap, 0.57% against a 1.92% global average, is stable enough to write into a multi-year budget line and large enough that partial progress is material. The instrument need not be new: CORFO’s cost-share grants, FONDCE’s fund-of-funds structure and the Uruguay Innovation Hub’s milestone-gated notes are documented, regional, and currently too small.

For allocators, thin deep tech deal flow is downstream of a research base spending under a third of the world average. That is why pipelines look shallow even where local venture capital has improved. For founders, grants are a real part of the stack here rather than a consolation prize: our report has Nintx assembling its Series A alongside a US$2.5 million FINEP grant.

Current changes in geopolitics can change the position of Latin America from a Deep Tech vassal to a stronghold.

That is the thesis of our September 2025 report, Accelerating Deep Tech in Latin America, and the research level is the part a government can act on inside one budget cycle. If you are designing an R&D instrument, allocating to the region, or holding data that sharpens the private-share range, we want to hear from you.