WIPO’s 2026 index separates the conditions for innovation from its measured outputs. The scores reveal contrasts that a single league table conceals, but cannot establish which economy gets the best return on investment.
Switzerland leads the World Intellectual Property Organization’s Global Innovation Index for the sixteenth consecutive year, followed by Sweden and the United States. The United Kingdom is sixth. Those headline positions, released on 29 September, compress a more revealing comparison: the conditions that support innovation and the results captured by the index can point in different directions.1
Singapore has an input score of 70.0 and an output score of 52.0. China scores 57.5 and 56.5 respectively. Using the unrounded data, Singapore’s input advantage is 12.5 points, while China’s output advantage is 4.5 points. Switzerland combines an input score of 69.3 with an output score of 64.0. These are published composite scores, not amounts of money or counts of inventions.2
The overall GII score averages the two sub-indices equally. Five pillars describe inputs, including institutions, human capital and infrastructure; two describe knowledge and technology outputs and creative outputs. Indicators are normalised to a 0–100 scale. An output point therefore has no common physical unit with an input point.3
Our analysis: comparing economies within each score is useful. Subtracting one kind of score from the other and calling the result “efficiency” is much harder to defend. A low output-minus-input number is not a measured quantity of wasted research, and dividing the scores does not produce a financial rate of return.
Singapore’s profile places it first for institutions, second for human capital and research, and fourth for business sophistication. Its output ranks are lower: eighth for knowledge and technology outputs and eleventh for creative outputs. Nevertheless, it performs above its high-income group’s average in every pillar. A weaker output rank than input rank is compatible with a strong innovation system.4
China ranks twenty-second on inputs and fifth on outputs. WIPO links its broader performance to a research and industrial base that distinguishes it from other middle-income climbers. The report describes more selective strengths elsewhere: India in digital services and entrepreneurship, and Viet Nam and the Philippines in international manufacturing and service networks. These are WIPO’s interpretations of the indicator profiles, not experimental findings about which policies caused the results.56
The scores also preserve distances that rankings discard. The UK records 61.8 for inputs and 57.4 for outputs; India 42.4 and 37.8; Viet Nam 41.3 and 34.2. India ranks twenty-fifth on outputs against forty-ninth on inputs, yet its output score remains well below those of the leading economies. Relative standing and absolute score answer different questions.2
| Economy | Input | Output |
|---|---|---|
| Switzerland | 69.3 | 64.0 |
| Singapore | 70.0 | 52.0 |
| United Kingdom | 61.8 | 57.4 |
| China | 57.5 | 56.5 |
| India | 42.4 | 37.8 |
| Viet Nam | 41.3 | 34.2 |
For a reader assessing policy, the useful next question is which components explain the profile, and whether comparable economies share it. National averages cannot identify a particular successful programme. Nor does a strong export indicator, by itself, reveal how much value domestic workers or firms retain. Those questions require evidence beyond the composite score.
Only 4.1% of available observations are from 2026; 76.1% are from 2024 or 2025. Some are considerably older. Missing values account for 10.9% of possible data points and are omitted from aggregation rather than entered as zero.3
Our analysis: the timing mismatch cuts both ways. Recent spending may take years to produce a commercial product; current outputs may reflect capabilities built long ago. Pairing today’s published input and output scores does not track the same investment through its lifecycle. It cannot distinguish delayed success from persistent underperformance.
Financial returns need revenues, costs and a time horizon. Societal benefits need measures such as access, health, living standards and environmental effects. WIPO’s separate Innovation Tracker follows investment, technological progress, adoption and socioeconomic outcomes. The report records both wider technology adoption and uneven access, illustrating why an innovation league table cannot stand in for a welfare assessment.7
The European Commission’s Joint Research Centre tested alternative weights, aggregation methods and treatments of missing data in 5,000 simulations. For 86 of the 139 economies, the resulting 90% GII rank interval spans fewer than ten positions. Some are much wider. These intervals describe sensitivity to modelling choices, not every possible error in the underlying observations.8
The audit supports using the index to compare broad patterns while cautioning against taking every place literally. It does not independently establish the causal productivity of national innovation spending. Nor does this article estimate that productivity. The defensible finding is narrower: economies can reach strong overall positions through different combinations of measured inputs and outputs.
The selected examples illustrate contrasts; they do not define a global efficiency frontier. Sector mix, data coverage, timing and indicator design may all affect the comparison. Treat the scores as prompts for investigation, not verdicts on whose innovation policy works best.
WIPO’s 2026 workbook, Data sheet, supplies IN and OUT scores. Values are rounded once to one decimal; differences use full precision. All 139 overall scores were checked against the mean of their sub-indices. Selected ranks were cross-checked against WIPO profiles and results. These are the same underlying source, not an independent replication. JRC provides a separate methodological audit. No regression, efficiency ratio or causal effect is estimated. Observation-year shares are WIPO’s published figures.
All sources retrieved 29 September 2026.