An International Comparison of Government Funding for R&D and IP Costs

Governments around the world are increasingly encouraging R&D, and particularly innovation, by supporting local companies as well as attracting foreign companies to invest locally in R&D. Equally there is a push for the IP generated from the R&D to be properly legislated for and protected. This article examines and compares ways that governments have been providing encouragement, for example the types and levels of direct support and tax relief offered by a number of key economies.

Direct support

Direct support in the form of grants or subsidies for R&D is very common in the jurisdictions listed. Direct support for resulting IP is less common in Europe and the US but is more widely available in a number of Asia Pacific countries as set out below. 

Direct support for R&D is often in the form of either research funding, (which may be done via collaboration with academia) or project type grants for a specified export product or market strategy. It will often partially cover such things as employing R&D technical staff, prototyping costs, market surveys and sometimes associated professional advice. For example In Singapore the direct funding support has been increased to 90% of IP costs for Covid-19 affected businesses.

As foreshadowed above, direct support for IP filings is less common than the direct support for R&D.  Sometimes funding may be available for capability development e.g. IP strategy work or FTO searching, but not IP filing per se. However, when it is available, it will usually be focused on the initial stage of filing a provisional followed by a PCT application. Usually by the end of the 30 Month PCT deadline, commercial arrangements should be in place and the project should be self-funding. China has long provided a comprehensive central policy for IP filing subsidies, which is then implemented by the provincial governments.

Tax relief

Tax relief is another tool used by governments to incentivise local innovation or investment. This support may be in the form of tax breaks (that may be refunded in cash for R&D or IP spending) or IP box regimes. 

Enhanced tax breaks on R&D and cash pay-outs for tax losses are often of most interest for start-ups who have a high burn rate and need to hoard every dollar they can. Increasingly for multinationals who are forum shopping for the most favourable regime to locate a R&D hub, the tax regime may be a deciding factor.

Tax relief on IP expenses, and more so on IP derived income, is also becoming a tool for encouraging innovation in some economies. 

Patent Boxes or IP Boxes are a developing tool used around the world and essentially prescribe a lower tax rate on income derived from patents (and in some cases from other forms of intellectual property).  The UK has a good example of a patent box, where the tax on patent derived income is prescribed at to 10% but it can go as low as 2.5% such as Cyprus.

Tax is obviously a more complex issue to describe and compare between different jurisdictions. The table below (and accompanying footnotes) is a snapshot of the different types of support available in several patent jurisdictions at the time of writing, but is subject to significant change in the current environment. 

Country IP direct support R&D direct support Tax Break IP box
NZ Yes1 Yes2 Yes3 No
Australia Yes4 Yes5 Yes6 No
Singapore Yes7 Yes8 Yes9 Yes10
China Yes11 Yes12 Yes13 No14
US No15 Yes16 Yes17 No
EU Yes18 Yes19 Yes20 Yes21


A wide range of subsidies and tax relief provisions are currently available, and in the predicted recession, countries are likely to offer even more generous schemes (for those that haven’t already). This is an important factor for corporate counsel or leadership teams considering where to base their R&D efforts. As always, expert advice will be necessary to assess this for individual circumstances.

1 Callaghan Innovation IP development programme up to $22,000 (40% subsidised). Co-investment of up to $900,000 available for international growth. and

2 Grants of 40% of eligible R&D costs. See

3 15% R&D tax credit, 28% cash-out of R&D losses. and

4 15% R&D tax credit, 28% cash-out of R&D losses. and

5 Up to $100,000 for research costs. and

6 Up to 43.5% R&D tax offset.

7 Grant of up to 80% of product development costs.

8 Grant of up to 80% of product development costs and 70% of costs up to $100,000 SGD for entering new markets. and

9 Grant of up to 80% of product development costs and 70% of costs up to $100,000 SGD for entering new markets. and

10 IP income taxed at 5% or 10%.

11 Up to $7100 USD per application.

12 Up to $7100 USD per application.

13 Tax credit up to 75% of R&D costs. Corporate tax rate reduction to 15% under High and New Technology Enterprise (HNTE) programme. and

14 Although see the HNTE programme above..

15 Although companies can benefit from IP developed using federal R&D funding.

16 Although companies can benefit from IP developed using federal R&D funding.

17 Tax credit up to 20% of R&D costs.

18 Tax credit up to 20% of R&D costs.

19 Funding awards of several million Euros at the EU level, grants or awards available in UK, France, Germany and others.,,,

20 Funding awards of several million Euros at the EU level, grants or awards available in UK, France, Germany and others.,,,

21 Patent box systems in UK, France, Italy and others.

Strange Bedfellows: Marrying Deep Tech with IP Strategy

In a drama worthy of Shakespeare, players in deep tech, such as artificial intelligence, block chain, big data, robotics, electronics, and photonics, often feel like they are arrayed against an army of foes.

In this series of articles, we will investigate some of the issues encountered by deep tech in trying to integrate IP strategy into their battle plans (or courting strategy as the case may be). In this first issue we will look at the Agile development approach, what challenges this brings, and how an Agile IP strategy can be transformed into a perfect bedfellow.

Agile Development

Agile development is sometimes described as a cult or manifesto for coders. Even though it dates back to 2001, it is still in widespread use today in the deep tech arena, and is gaining more popularity in wider business processes.

It typically relies on short sprints or iterations where everybody in the organisation is focused on implementing a feature or revision within a set timebox. In this mode, it is very hard to find focus or time to spend on IP or other “trivialities”: everybody is focused on the end goal for that iteration.

At the beginning of each sprint there is a goal, but by the very nature of this approach, goalposts shift drastically over the project. This means pinning down what the product will look like at the end is difficult if not impossible. Sprinting is very focused on getting customer feedback at the end of each sprint and feeding that back into the plan for the next sprint.

Overall IP strategy

The overall business goals in deep tech will typically be related to a successful exit or further investment rounds. In such a case, the IP strategy can often be focused on adding value to future investors or other stakeholders. Deep Tech IP is often shrouded in mystery as if there is some secret crystal ball. However, IP strategy is usually very simple:

  • What aspects will add the most value to your revenue projections?
  • Can those aspects be protected (and how)?
  • Who will try to copy it?
  • How will you stop the copiers?
  • Do you need to worry about patent trolls or infringing other patents?

Freedom to Operate

Freedom to operate (FTO) can sometimes be the last thing entrepreneurs want to know about. They are focused on getting their product to market and burn rate. However, stakeholders such as investors, and even government funding can often be tied to this thorny issue.

While stakeholders may insist on getting freedom to operate searching done, depending on the nature of the tech, the markets and even on the investors, it may make sense to delay or rethink what searching is most useful.

Sprints can change the nature of the product drastically, so figuring out what to search is tricky.  Because FTO searching can be expensive, and relies on comparing the launched product with live patents existing at launch, it is often left until late in the process. However, this need not be the approach.

Early on in the project an overall FTO search can be done on the very broad premise. This will draw out any major roadblocks but won’t focus on finer details. This can be used to guide each sprint on general areas or features to avoid.

For each sprint, instead of an FTO search, a short landscape search may be better. This is more targeted at giving an idea of who has patented what in the area of the iteration. This can be used together with a SWOT analysis of your strengths, weakness, opportunities, and threats to help align IP value with the R&D development. At the end of the day, the vast majority of value that successful businesses accrue is intangible (whether measured in the balance sheet or through expert intangible valuations) and this should remind us all of the risks at stake.


Deciding what to protect, when and how, may cause endless hand wringing. This is particularly acute with an Agile approach since nobody knows what the product will look like at the end of the project. Alternatively, the founders may have heard through contacts that trade secrets are the way to go, or some rumour that software isn’t patentable.

The best approach (as with most things) is to be pragmatic and prioritise with your best guesses. For each sprint, a decision can be made about what might be patented during the sprint and at what stage. This will change according to where that sprint is, in the overall timeline. The organisation can include things to patent in the burn chart/story map but this can be updated as part of the Agile process.


During internal development, before the product has been launched, there is minimal practical risk of patent infringement. However once your product is out in the market, especially in the US, is it common in deep tech to get infringement warnings in the mail. Whether the threat comes from a trading business or a non-practising entity (patent troll), a pre-planned approach to a range of scenarios is a good start.

Often the key is acting quickly and confidently. This may mean calling their bluff straight away (where the patent in question was already known previously to be of dubious validity), not replying, entering a dialogue about collaboration instead, or even entering an early settlement at a much lower royalty rate.


The brand may be the last thing on engineers’ minds during each sprint, but while customer perception may be the focus of marketing or sales, it should be on the forefront of the whole team.  Sometimes a pseudo internal brand for each sprint can help, which encompasses or represents the personas for the project or sprint. 

At fulfilment or closer to product launch, these internal brands might actually provide some candidates for focus groups. Even at that stage it is important to think through perceptions of different cultures and implications of different languages or pronunciations of major target jurisdictions.

A final word

Buzzwords may be king, but in the end what matters is value. Simplicity is essential in the IP strategy and the simpler and shorter it is the easier it is for all the team to buy into it and integrate it into their tasks. The IP strategy can be a perfect bedfellow of Agile development.

A Final Innovative Step: Australia Plans to Phase Out Innovation Patents (Update)

On 26 February 2020, the Intellectual Property Laws Amendment (Productivity Commission Response Part 2 and Other Measures) Act 2019 came into force in Australia. This begins the abolition of Australia’s innovation patent system.

Any innovation patent application must be filed before 26 August 2021. From that date, you will only be able to pursue a standard patent in Australia.

This will mark the end of Australia’s second-tier patent. Our earlier post discusses the background to this change.

This change will affect a number of New Zealand and Australian businesses who protect their incremental innovations which do not quite fit in the standard patent system. Get in touch with us to discuss how this change will affect your business.

A Final Innovative Step: Australia Plans to Phase Out Innovation Patents

On 25 July 2019, the Australian Government introduced a bill to make a number of changes to the patents system. The changes to Crown use of patents and designs and to compulsory licences will be of interest to some. But the biggest change is phasing out innovation patents, Australia’s equivalent to a second-tier patent or utility model.

Innovation patents have a lower requirement for inventiveness than standard patents, are granted without substantive examination, and have an 8-year term. This is ideal for incremental improvements which may not justify the expense of a standard patent, but still benefit from a level of protection.

This will align Australia with jurisdictions like the United States and New Zealand which do not have a second-tier patent.

The proposed law will prevent new innovation patent applications being filed. It will not affect existing innovation patents.

This follows the Australian Productivity Commission’s report in 2016 that recommended abolishing innovation patents. Their reasoning is that the innovation patent system probably does not benefit Australian SMEs.

It is not clear when (or if) the bill will be enacted. However, there is no indication that this will be a contentious bill, and it may well pass through both houses relatively quickly.

Once the law comes into force, there will be an 18-month transition period in which new innovation patents can be filed. After that, no new applications will be allowed. With the 8-year term of innovation patents, this means that the final innovation patent, and so the innovation patent system as a whole, will likely reach the end of its life in 2028.

Reducing Reuse: Australia’s Full Court Develops the Law on Refurbishing Patented Products

The Full Court has confirmed that refurbishing patented products in Australia may result in patent infringement.

The Cast

Seiko is a Japanese company which is a large player in the printer market. Seiko sold ink cartridges in Australia under the Epson brand which were covered by the claims of an Australian patent.

Ninestar is a Malaysian company who would acquire used Epson cartridges. These would be refilled with ink and have their memory updated or replaced to allow the cartridge to be reused in Epson printers.

Calidad is an Australian group of companies, which imported the refurnished cartridges from Ninestar and sold them in Australia.

The Background

Seiko asserted that Calidad’s actions in Australia infringed its Australian patents.

In 2017, the Federal Court of Australia found partly in favour of Seiko. Burley J ruled that cartridges which had the memory replaced were infringing, but those that were merely refilled were not infringing. Both parties appealed to the Full Court of the Federal Court of Australia.

The Law

The Full Court’s judgment in Calidad Pty Ltd v Seiko Epson Corporation [2019] FCAFC 115, issued on 5 July 2019, found entirely in favour of Seiko. All of Calidad’s cartridges were found to be infringing.

The Full Court confirmed that the sale of a patented product gives a third party user an implied licence to have “all the normal rights of an owner”, which includes using or selling. However, the third party user cannot rely on this implied licence to make an infringing product.

Jagot J’s position (supported by Greenwood J) was that, because Seiko sold a single-use cartridge, the implied licence could not cover multiple uses. Ninestar’s actions made the cartridge act beyond how Seiko intended, and therefore was not covered by the implied licence:

The product which Seiko sold, in the form in which it was sold, was capable of use until the memory chip showed that the ink in the container was exhausted. By re-programming the memory chip, Ninestar enabled an embodiment of the invention, which could no longer be used, to be used. It was not repairing the cartridge. The cartridge was not damaged or worn in any way. It had simply reached the end of its intended life as a printer cartridge. Ninestar, by re-programming the memory, putting a new hole in the cartridge to enable it to be re-filled with ink, re-filling the cartridge with ink, and sealing the new and the existing hole created by the original use, manufactured a new embodiment of the invention, an act which could never have been authorised by the implied licence and could never be the subject of an exhaustion of patent rights by reason of sale. Calidad, in importing those cartridges, necessarily exploited the invention and thus infringed the patents.

[177] per Jagot J.

Yates J had a slightly different reasoning. The modifications were such that Ninestar’s refurbished products could not be said to be of Seiko’s making:

The correct approach was to ask whether, in each particular category, the modifications which Ninestar made to the original Epson cartridges altered them in such a way that they were, in substance, different articles to those which Seiko had put into the market and thus into the hands of the original owners. If so, they were not the articles in respect of which Seiko, as patentee, had given the implied licence and Calidad could not rely on that licence in respect of what were, in substance, different articles.

[293] per Yates J.

Under either reasoning, Ninestar’s action went beyond the implied licence that comes attached to the sale of patented products. Calidad’s importation of these refurbished products therefore was an infringement of Seiko’s patents.

This is a more restricted implied licence than that of the primary judge’s approach. The primary judge looked at whether Ninestar’s modifications amount of material changes to the features that embodied the claimed invention, and found that merely refilling the ink and updating the existing memory was not a material change.

The Future

Any modification of a patented product which does not strictly replace damaged parts or which was not envisioned by the original seller may be an infringement of an Australian patent.

This could have a significant effect on third party repairers or modifiers, who will now need to consider whether their actions could be patent infringement.

Calidad may try to appeal this to the High Court, which would be free to depart from the earlier decisions (including the 1911 Menck decision) which the Federal Court and Full Court felt bound by. If the High Court decides to hear the appeal, the doctrine of implied licence in Australia may yet be refurnished.