The Frascati Code: decoding the global R&D tax credit definition

In 1963, an international group of experts gathered at an Italian villa to solve a puzzle: how to define and measure research and development (R&D) in a consistent way. The result was the Frascati Manual, named after the meeting location. Drafted with input from British economist Christopher Freeman, who provided a key background paper, the first edition laid out a common framework for R&D measurement. This framework not only standardised R&D statistics worldwide but also unwittingly created the “code” for what counts as R&D – a code that decades later underpins R&D tax credit schemes across the globe. A definition that lays at the heart of an ever increasing number of challenges by tax authorities worldwide into R&D tax credit claims.

A brief history in Frascati, 1963

The Frascati Manual’s origin story is a lesson in international collaboration. In June 1963, the OECD’s National Experts on Science and Technology Indicators (NESTI) convened at Villa Falconieri in Frascati, Italy, to draft guidelines for surveying R&D. Freeman’s vision and expertise in innovation studies helped shape the manual’s content. What they produced was officially titled The Proposed Standard Practice for Surveys of Research and Experimental Development, but it’s affectionately known as the Frascati Manual. Over the years, this manual has been updated (the latest major edition is 2015 with an update scheduled in 2025) and has become the international reference for R&D definitions and statistics. By the 21st century, around 75% of countries were using Frascati’s methodology as a common language for R&D policy. In other words, what started as an OECD document from a meeting in Italy evolved into a global standard.

Defining R&D: Novelty, uncertainty, and systematic work

At the heart of the Frascati Manual is a clear definition of what constitutes R&D. This definition rests on a few key pillars that distinguish genuine R&D from routine work. In simplified terms, an activity may qualify as R&D if it involves:

Novelty: It aims to achieve something new – an advance in science or technology that goes beyond the current state of the art. There should be a measurable element of originality or innovation involved, not just the application of existing knowledge.
Uncertainty: The outcome is uncertain at the outset. R&D tackles problems where it’s not initially clear if a solution is possible or how to achieve it. In other words, there are scientific or technological uncertainties that require resolution. If you already know exactly how to do it, it’s probably not R&D.
Systematic Investigation: It is carried out in a structured way. Frascati emphasises a planned, structured approach – formulating hypotheses, experimenting, testing, and learning in a reproducible manner. R&D is not a one-off eureka moment or trial-and-error tinkering; it’s an organised and documented inquiry into the unknown.

These criteria (along with attributes like creativity and the potential for knowledge transfer) form the “Frascati definition” of R&D. Only work that meets these tests should be counted as research and development in surveys – and, by extension, for tax incentive purposes. This clarity is crucial: it draws the line between true R&D and activities that may be innovative but fall short of R&D (such as routine product updates, quality control, or market research).

Global ripple effect on R&D tax credits

Because the Frascati Manual became the gold standard for defining R&D, governments worldwide have based their R&D tax credit regimes on its principles. The idea is simple: if you’re going to give tax breaks for R&D, you need to define R&D consistently. Many jurisdictions therefore crib directly from the Frascati definition (novelty, uncertainty, systematic approach) in their tax laws or guidelines. Here’s how the “Frascati code” influences some major R&D tax regimes:

United Kingdom: The UK’s R&D tax relief follows Frascati. To qualify, a project must seek an advance in science or technology and resolve scientific or technological uncertainty. This is directly descended from the novelty and uncertainty criteria.
United States: The U.S. tax code’s definition of qualified research uses a four-part test that aligns with Frascati principles. It requires a technological innovation (new or improved business component) and a process of experimentation to overcome technical uncertainty. In essence, the IRS demands the same two elements – something new, and a systematic attempt to solve an uncertainty – even if the phrasing differs.
Canada: Canada’s SR&ED program (Scientific Research & Experimental Development) is famously stringent. It insists on a “systematic investigation or search” carried out through experiment or analysis to resolve unknowns. This language could be lifted straight from Frascati. Canadian claimants must document their hypothesis and tests to show they followed the scientific method.
Australia: Australia’s R&D Tax Incentive defines core R&D activities in almost identical terms to Frascati. It must be an experimental activity whose outcome “cannot be known or determined in advance” and can only be determined by “applying a systematic progression of work” from hypothesis to observation and conclusion. The work must be aimed at generating new knowledge – a direct nod to the novelty criterion.

It’s remarkable that whether you talk to HMRC in the UK, the IRS/CRA in North America, or AusIndustry in Australia, you’ll hear the same fundamental requirements. This is no coincidence – it’s the Frascati Manual’s legacy. Even where tax authorities don’t name the manual explicitly, they rely on its concepts to distinguish eligible R&D. By establishing a common baseline, the Frascati definition allows multinationals and policymakers to speak a common language about innovation incentives.

Of course, there are local twists. For example, the US focuses on business components and excludes social sciences; Canada requires the advancement to be in the realm of science/tech and not just company-specific; Australia splits core vs. supporting R&D activities. But these are variations on a theme – the theme set in Frascati. CFOs and finance directors should recognise that the same DNA runs through all these regimes. Understanding the Frascati principles thus gives a head start in grasping any country’s R&D credit rules.

When misunderstanding turns into misuse

Despite this clear “code”, many companies (or their advisors) misinterpret it – leading to erroneous, inflated, or even fraudulent R&D tax credit claims. Tax authorities worldwide have flagged a surge in claims that stretch the definition of R&D beyond its intent. Often, the root cause is a failure to apply the Frascati fundamentals:

Including non-R&D work in claims: A common error is claiming routine upgrades, data collection, or tweaking of a process as if it were groundbreaking R&D. As HMRC notes, many claims include “project activities outside the scope of R&D for tax purposes”. For instance, developing a new website interface or conducting market testing might be innovative, but unless there’s a scientific/technological uncertainty being resolved, it’s not R&D by the Frascati definition.
Lack of Novelty: Some claims list projects that simply apply existing know-how with no appreciable advance. If dozens of companies have already implemented a similar solution, it fails the novelty test. Claimants sometimes misunderstand that the “advance” must be relative to the global state of the art, not just new to their firm (except in cases where that knowledge was truly inaccessible to them). This misinterpretation can lead to ineligible claims that get rejected on audit.
Overlooking the Uncertainty requirement: If a project’s success was a foregone conclusion, it isn’t R&D. Yet, some abusive claims retrofit an R&D narrative onto work that had no real scientific uncertainty. Tax authorities are on the lookout for projects where the solution was obvious (to a competent professional) from the start. Genuine R&D involves risk of failure – and indeed failed projects can still qualify, whereas guaranteed-success projects often do not.
Not being Systematic: Documentation is the Achilles’ heel of many claims. Companies that did hackathon-style innovation without a plan or failed to document their experiments struggle to prove the work was systematic. For example, Canada’s CRA and others may deny credits if you can’t show a hypothesis, experiment records, and results analysis. Simply stumbling onto a fix during routine work isn’t the same as deliberately researching a problem. Claimants who don’t maintain R&D project logs, methodologies, and outcomes can find their claim labelled as “not in the spirit of R&D”.
Boundary-pushing advisors: In recent years, firms offering aggressive R&D credit services have exacerbated misuse. They might encourage companies to claim business-as-usual costs as R&D, or use overly broad interpretations of eligible activities. This has led to crackdowns. The UK, for example, paused certain R&D tax credit payments in 2022 amid concerns of widespread abuse, and introduced stricter filing requirements in 2023. Similar vigilance is rising in the US, where IRS scrutiny of the credit (and required documentation) has grown after some high-profile misuse cases.

The consequences of getting it wrong are serious. Erroneous claims can trigger audits, repayments, penalties, and reputational damage. For CFOs and business owners, the lesson is clear: know the Frascati definition and ensure your claims stick to it. It’s not just a bureaucratic technicality – it’s the very filter that separates legitimate innovation from ordinary activity.

Decoding the code across borders

What makes the Frascati Manual so enduring is that it provides a clear, authoritative benchmark that transcends national boundaries. For time-pressed executives, the takeaway is this: if you understand Frascati’s three key criteria – novelty, uncertainty, systematic inquiry – you can navigate R&D incentives anywhere. Think of it as the “Da Vinci Code” of R&D tax policy: an Italian-origin code that, once deciphered, unlocks secrets (in this case, tax benefits) around the world.

In practice, this means crafting your R&D project documentation to highlight what’s new, why it was challenging, and how you approached it methodically. By doing so, you not only maximise your eligible tax credits in the UK, US, Canada, Australia or wherever you operate, but you also protect your company from compliance risks. The Frascati Manual may have been written over 60 years ago, but its principles remain the compass for modern R&D tax relief. In a landscape of ever-evolving technology and tax rules, sticking to the fundamentals defined in Frascati is the smartest way to ensure your R&D claims are credible, defensible, and truly innovative.