Evidence Shows that 2013 Australian Patent Law Reforms Have Had No Long-Term Effect on Application Acceptance Rates

On 15 April 2013, the majority of provisions in the Intellectual Property Laws Amendment (Raising the Bar) Act 2012 came into effect.  Among other things, these reforms were designed to lift the inventive step standard, and to impose more-stringent requirements on the disclosure in a patent specification.  As a consequence of this, you would think that some inventions that would have been patentable prior to commencement of the reforms would be unpatentable under the new standard, and that some patent specifications that would have passed muster previously could fail to meet the higher standards for disclosure.  And, if so, then it might logically follow that the proportion of applications that successfully pass examination (i.e. are accepted for grant) would be lower under the new law than under the old.

Yet, as logical as that reasoning may seem, just recently I published an analysis indicating that acceptance rates across almost all examination technology sections at the Australian Patent Office were, in fact, consistently increasing on an annual basis between 2014 and 2016.  That result is surely worthy of a closer look!

A more-detailed analysis of Australian patent examination data covering a period of over three years either side of the Raising the Bar reforms, from 2010 up until the end of 2017, indeed confirms that a naïve assumption that lifting standards should reduce acceptance rates is incorrect.  In fact, acceptance rates had been rising for a number of years prior to commencement of the reforms, and initially this trend continued under the supposedly higher standards.  And although there was subsequently a brief decline in acceptance rates, this has been followed by a period during which acceptances have once again been on the rise.  Indeed, right now it appears as though the proportion of applications that are accepted for grant following examination is around 80%, which is almost exactly where it stood immediately prior to commencement of Raising the Bar.

Distinguishing Applications ‘Before’ and ‘After’ Reforms

Under the transitional provisions of the Raising the Bar Act, the amended patent laws apply to all applications, and resulting granted patents, for which a request for examination was filed on or after 15 April 2013.  Of course, once a request has been filed, there is a delay before examination actually commences, and then the applicant was allowed (under the former regulations) up to 21 months to gain acceptance after any initial adverse examination report was issued.  Furthermore, in exceptional cases extensions of time may be granted, leading to even longer delays in finalising an application.  As a result, applications where examination requests were filed prior to 15 April 2013, to which the old law applies, were still being accepted in 2017.  According to the latest 2018 Intellectual Property Government Open Data (IPGOD) release, there were 10 such applications for which a notification of acceptance was published in the second half of 2017.

Under revised regulations, the maximum period for an applicant to overcome any examination objections has been reduced from 21 months to 12 months.  Even so, there are still many cases in which more than two years may pass between filing of an examination request, and a final resolution (i.e. either acceptance, lapsing or refusal of the application).

These delays do not present an insurmountable problem in comparing acceptance rates over time, however they do create some challenges in conceptualising the relevant comparisons.  To properly understand the following analysis, you will need to keep these time scales in mind – I apologise if this makes anybody’s head hurt, that is not my intention!

The method I have employed to enable a fair comparison is to relate all of the data back to the date on which a request for examination was filed.  If this date is before 15 April 2013, then the old laws apply.  If it is on or after 15 April 2013, then the amended laws apply.  So far, so straightforward.

Determining the Outcome of Examination

For the purpose of comparing acceptance rates, I consider the relevant question to be: what is the ultimate outcome of examination for each application?  I assume that how long it takes to arrive at this outcome is unimportant (this is not quite true – subsequent changes in case law and/or examination practice could well affect the outcome in some cases – but as an initial assumption it is reasonable).  Once examination has been requested, there are essentially three possible ‘fates’ that are relevant here:

1. the application may be examined and, either immediately or following one or more adverse reports and responses by the applicant, be accepted;
2. the application may be examined, receive at least one adverse report and, either due to inaction by the applicant or specific decision by the Commissioner, lapse or be refused; or
3. the applicant may decide to abandon the application, by non-payment of an annual maintenance fee, before examination commences.

Note that I am not interested in what happens after an application is accepted.  Of course, in most cases a patent would be granted.  However, the application might be opposed (see this earlier article for some statistics on how often this happens – the short answer being ‘hardly ever’), and/or the applicant might abandon the accepted application by non-payment of fees before grant (again, a relatively rare event).  For present purposes, however, I am interested only in the outcome of examination – did the examiner end up accepting the application, or was it refused or allowed to lapse following an adverse examination report?  For this reason, I am also not interested in the third situation listed above – if an examiner never looks at an application, because it has already lapsed, then we have no way of knowing what the outcome would have been.

Finally, given the delays inherent in the examination process, there will always be many pending applications – mostly those for which an examination request was more-recently filed – that are still awaiting, or undergoing, examination, and for which the outcome is therefore not yet known.

Examination Outcomes, 2010-2017

To recap, then, for every application of interest we will have:

1. an examination request date; and
2. a status/outcome chosen from the set {Accepted, Refused/Abandoned, Pending}.

In the chart below I have plotted the percentage of applications having each status/outcome (as at the end of 2017) as a function of the month in which examination was initially requested.  I have also included a line labelled ‘Conditional Acceptance Rate’, which represents the number of accepted applications as a percentage of all applications for which examination has been finalised (i.e. excluding all still-pending applications).  In order to align these figures with the commencement of the Raising the Bar reforms on 15 April 2013, I have treated the 15^th day as the relevant ‘start date’ for each month.  Thus, for example, where the horizontal axis reads ‘Dec-09’, this refers to the period of one month falling between 15 December 2009 and 14 January 2010, inclusive.  All of the data is taken from the 2018 IPGOD release.

The following points are notable from the above chart:

1. month-to-month variations notwithstanding, acceptance rates under the ‘old laws’ steadily increase (from just over 70% to around 80%) over a number of years prior to commencement of the Raising the Bar reforms;
2. there is, at least initially, absolutely no evidence of any discontinuity in this upward trend of acceptances following commencement of the ‘new laws’;
3. there is an apparent drop in acceptance rates corresponding with examination requests filed one year after commencement of Raising the Bar;
4. since this drop, however, acceptance rates have again been on the rise, and within a further year are back up at the pre-Raising the Bar rate of around 80%;
5. for examinations requested more recently than the second half of 2015, the proportion of applications that remain pending increases due to the time-frames involved in the examination process and, as a result, the proportions of acceptances and refusals/lapsings decline;
6. however, the proportion of acceptances declines less rapidly than the proportion of lapsings, because an application can be accepted at any time during the examination process (including right at the beginning), whereas most applications that are going to lapse only do so once the full 12-month acceptance period has expired;
7. for the same reason, the ‘conditional acceptance rate’ rises with the proportion of pending applications, because most of the cases that are finalised earlier in the examination process are acceptances, whereas it takes longer for an application to be refused or allowed to lapse.

It should therefore be apparent that the data on acceptance rates is reliable for the monthly sets of applications for which examination requests were filed prior to the end of 2015, or thereabouts, where the proportion of pending cases remains low.  After this, however, the final acceptance rates will be dependent upon the outcomes of examination of the remaining pending cases.  They must, however, be higher than the currently-indicated acceptance rate, and in practice are almost certain to be lower than the currently-indicated conditional acceptance rate, so these two lines generally represent bounds on the final rate. 

To the extent that the (recent) past is likely to be the best available predictor of the (near) future, it looks very much as though the acceptance rate is likely to remain at or around 80%.  Given that this was the acceptance rate of applications for which an examination request was filed immediately prior to commencement of the Raising the Bar reforms, my preliminary conclusion is that the changes in the law have ultimately had virtually no long-term impact on acceptance rates of Australian patent applications.

Explaining the Pattern in Acceptance Rates

Some of the features of the previous chart can perhaps be understood by looking at the numbers and nature of examination requests filed over time.  The chart below shows the total number of examination requests filed each month (again using the 15^th day as starting point), along with the number of those requests that were filed voluntarily by the applicant, i.e. without the Commission of Patents having issued a direction requiring an examination request to be filed.

Not surprisingly, the number of voluntary examination requests soared just prior to commencement of the Raising the Bar reforms, as applicants sought (mostly on the advice of their Australian patent attorneys) to ensure that their applications, and resulting patents, would be subject to the former, less stringent, requirements.  Indeed, in anticipation of, and response to, this behaviour the Patent Office issued significantly reduced numbers of directions to request examination in November and December of 2012, none at all in January and February of 2013, and just 107 for the entire remainder of 2013.

As a result, the vast majority of examination requests filed during the first year following commencement of the revised laws were voluntary requests.  And the first month in which directed examination requests returned to pre-Raising the Bar levels was April 2014 – i.e. exactly the same month for which the notable drop in acceptance rates occurred.

Based on this data, my hypothesis (and that is all it is at this stage) is as follows.  Applicants that file voluntary examination requests can generally be assumed to have a specific commercial or legal reason for doing so, and can perhaps be assumed to be more than ordinarily confident about the outcome of examination – or at least about their willingness to compromise on claim scope in order to achieve their objectives.  It is interesting that Raising the Bar does not appear to have had any impact on the underlying rate of voluntary examination requests, suggesting that the specific legal requirements are, at best, of secondary concern to these applicants.  It is therefore perhaps to be expected that these voluntarily-examined applications, taken in isolation, should continue to follow the pre-reform trend in acceptance rates.

It appears, then, that it was only when applicants were first compelled to request examination under the revised laws – bearing in mid that these applications would mostly have been filed prior to commencement of the reforms – that any significant change in acceptance rates can be observed.  However, this effect was transient, and it seems that it took applicants only a further year to adapt to the changes, with acceptance rates returning to the same level as before the drop.  For example, the acceptance rate had recovered to 81% for cases with examination requests filed in April 2015.

Conclusion – Is Consistency In Acceptance Rates Surprising?

Personally, I am not at all surprised to find that the Raising the Bar reforms have had no long-term impact on acceptance rates.  The proportion of applications that ultimately result in granted patents is not what matters most – not to applicants, not to examiners, not to the Patent Office, and not to the Australian economy.  What matters is the scope and quality of the claims that are granted.  By ‘scope’, I mean the legal and technical extent of the patentee’s right to exclude others from exploiting the invention defined in the claims.  And by ‘quality’ (a notoriously ill-defined term, when it comes to patents) I mean essentially the ability of those claims to withstand more-exacting scrutiny, e.g. in the course of litigation, so as to deliver the promised rights to the patent owner.

With regard to scope, it is difficult to say whether raising the patentability standards has had a significant impact.  It is reasonable to assume that there must be recent cases in which narrower claims have been granted than would have been allowable under the former laws.  However, even before Raising the Bar many claims granted in Australia were based on claims that had already been allowed in other jurisdictions, such as the US and Europe, the laws of which the reforms were intended to emulate.  Such claims should equally be allowable following the reforms.  From a policy perspective, however, the objective was to eliminate the possibility of broader claims being granted in Australia than elsewhere, regardless of whether or not this was actually happening at any scale.

As for ‘quality’, it is even more difficult to assess whether Raising the Bar has made any difference.  The objective of patent examination, surely, is to endeavour to ensure, as far as possible given inherent resource limitations, that invalid claims – and thus illegitimate property rights – are not granted, while also not denying rights on applications that meet the legal requirements.  This objective is the same, regardless of the underlying leniency or stringency of the standards required by the law.  So while a drop in acceptance rates might mean that a greater proportion of applications are being legitimately rejected due to higher legal standards, it could just as easily indicate that over-zealous, or poorly-trained, examiners are rejecting applications that should be accepted (note that I am not saying that this necessarily is, or ever has been, the case – just making the point).  And the converse argument applies to rising acceptance rates.  In the absence of substantive case law involving patents issued under the revised laws, we currently have limited information on whether the Patent Office is applying them appropriately, or on the ‘quality’ of the patents being granted.

Over time, applicants (mostly with the assistance of their patent attorneys) will adapt to whatever national patent laws, and examination practices, happen to be in-place.  A well-advised party does not waste time and money filing an application for a clearly unpatentable ‘invention’.  They do not knowingly file patent specifications that do not meet the description requirements.  And they do not, ultimately, abandon an otherwise commercially-valuable patent application merely because they might not be able to secure claims that are as broad in scope as they would like, or could historically have obtained, as opposed to allowable claims that are as broad as they need.

If Australian patent claims now are narrower in scope than they could have been under the former law, then that might be a ‘good thing’ as a matter of principle and policy.  However, if the potential infringing activity that is of concern to a patent holder falls just as well within the narrower claims as the broader ones, then there may be no difference in practice.  In the same way, there is no good reason to expect that changing the legal standards of patentability will have any long-term impact on application acceptance rates.  And the data bears this out.

Tags: Australia, Examination, IPGOD, Law reform, Patent analytics, Patent law