24 August 2014

Australia’s Patent Application Backlog Revealed

2013 IPAU BacklogEarlier this month, IP Australia released a ‘report on patent backlogs, inventories and pendency’.  The cover page states that it is ‘an independent report commissioned by IP Australia’, and that ‘[f]indings and opinions are those of the researchers, not necessarily the views of the IP Australia or the Government.’  My understanding is that the author, Ahmer Iqbal Siddiqui, worked on the report as an intern within IP Australia for around three months during 2013.

The research is based upon historical data of pendency for applications disposed of (i.e. either granted or abandoned) between January 2000 and January 2013, i.e. prior to the commencement of the Intellectual Property Laws Amendment (Raising the Bar) Act 2012.  During this period IP Australia’s inventory (i.e. number of pending applications at any stage between filing and grant or abandonment) rose from just under 45,000 to a peak of almost 100,000 in October of 2009, before falling back to hover around 90,000.

This represents an inventory of around 278 applications per examiner, as compared with the UK Intellectual Property Office with 198, and the USPTO with just 169.  Patents granted in January 2013 had been pending for an average of about three-and-a-half years, up from just two years in 2000, but down from a peak of four years in 2010.


The report uses an analysis framework that was developed by Benjamin Mitra-Kahn (now IP Australia’s Chief Economist) an co-authors for a joint UK-IPO/USPTO report entitled ‘Patent backlogs, inventories, and pendency: An international framework’ [PDF, 4.0MB].  According to this approach, overall pendency is divided into three stages:
  1. the period between filing (or national phase entry of a PCT application) and when the application is ‘ripe’ for examination;
  2. the time then spent waiting for examination to commence; and
  3. the period during which the application is under examination, ending with disposal of the case.
In Australia, we have a ‘deferred examination’ system, i.e. examination does not commence until the applicant makes a request and pays the applicable fee.  This can be done voluntarily at any time (up until an absolute deadline of five years after filing), but most commonly (in around 80% of cases filed in 2008, for example) examination is requested following a direction issued by the Commissioner of Patents.  During the period of the study, a direction gave the applicant six months within which to request examination, or the application would lapse (this has since been reduced to two months).

The direction process enables the Patent Office to control its operations: directions are issued to limit the backlog of older applications, while keeping sufficient examination work in the pipeline.

Comparison With US System

The deferred examination system makes it very difficult to compare figures with the USPTO, for example, even though the same analysis framework has been employed.  In Australia, the first stage of application pendency is driven primarily by applicant choice of when to request examination, and can last anywhere from virtually no time at all (if examination is requested at filing) to many years (if the applicant awaits a direction).  In the US, on the other hand, the first stage is generally much shorter on average, and may be dominated by administrative processing in the USPTO.

Similarly, the duration of the third stage is strongly influenced by national laws and regulation.  In Australia, during the period of the study, once a first examination report was issued the applicant had 21 months to gain acceptance or the application would lapse (this has since been reduced to 12 months).  While this could involve any number of rounds of examination, the period is effectively capped.  In the US, however, an application may be kept pending for as long as the applicant is willing to keep paying for further rounds of examination, by filing a Request for Continued Examination (RCE) after any ‘final’ Office Action.  So the examination period is uncapped.

As a result, when compared with the US, Australia has a greater proportion of pending applications in the first stage.  In the US, about 80% of all pending applications are either in the second or third stage, with a fairly even distribution between the two.

Total Pendency Reasonably Consistent

Overall, however, and despite the extended time some Australian applications spend waiting for examination to be requested, the differences in total pendency between Australia and the US are not so great.  In mid-2010 (the last time at which US data was available) patents being granted by IP Australia had been pending for only around six months longer than patents being granted by the USPTO (just under four years for Australia, compared with just under three-and-a-half for the US).

I expect that this is because, in many cases, the examination process proceeds far more quickly in Australia.  By waiting to request examination, applicants who have corresponding applications pending in other jurisdictions, such as the US and Europe, can wait until they have dealt with prior art and other objections arising elsewhere before turning their attention to Australia. 

Australian examiners, for their part, can employ the search and examination reports produced in other jurisdictions to reduce their own workload.  And it is for this reason that the relatively high number of pending applications per examiner in Australia is probably not of great concern.  Anybody who deals with both US and Australian patent applications would be well-aware that an Australian examiner rarely needs to put in the same amount of time and effort as a US counterpart – partly due to the common availability of foreign examination reports, and partly due to significantly less onerous examination reporting requirements imposed upon Australian examiners.

Conclusion – Recommendations for Future Analysis

The report concludes with a number of recommendations to enhance future analysis and operations at the Australian Patent Office.

For example, it suggests that additional data should be captured to distinguish between time spent awaiting office action, and time spent awaiting action by the applicant.  As matters currently stand, the first stage of the process is dominated by the applicant’s power to choose when to request examination (or to await a direction), while the second stage basically reflects the length of the ‘queues’ of examination awaiting attention by examiners.  In the third stage, however, responsibility passes back and forth between the examiner and the applicant, and IP Australia currently does not capture this data (unlike the USPTO, which keeps meticulous records for the purpose of computing patent term adjustments).

It also suggests capturing additional data relating to the complexity of examination of each application, based on such factors as the number of claims, and whether foreign search and examination reports are available.  Such data could be used for the planning and allocation of examination resources.  On a related point, the report recommends putting in place better mechanisms for measuring the time required by examiners for examination of applications, and suggests that the present system of ‘self-reported hours’ does not ‘appear robust for analysis’.

It will be interesting to see whether these recommendations are put into practice.  While it is always possible to collect and analyse more data, this does introduce additional overhead.  Determining in advance whether the benefits will exceed the costs is not straightforward.  As I noted at the outset, the ‘[f]indings and opinions are those of the researchers, not necessarily the views of the IP Australia or the Government.’


Mark Summerfield said...

You're undoubtedly more knowledgeable about IP Australia's practices than I am, but it seems to me that they do actually capture data distinguishing between time spent awaiting action and time spent awaiting applicant response, since the "date out" of an examiner's report and the "date in" of the response are present in the file and indeed visible on-line in AusPat for any given case. That isn't to say that the data are neatly or easily usably captured for analysis, but captured they are.
I was underwhelmed by the report, which seemed to be the uncritical accumulation of data with no clear attempt to establish its significance either to IP Australia or to applicants. In particular, comparing the inventory per examiner between the US PTO and IP Australia without taking into account that in the US essentially all Stock 1 inventory (applications that are formally complete) ultimately has to be worked on at least to first examination report whereas in Australia only Stock 2 inventory (formally complete applications on which examination has been requested - and annuities paid, if they have come due) has to be worked on effectively overstates the Australian inventory, since only around three-quarters of applications have examination requested (Figure 4).

Mark Summerfield said...

Hi Derek,

Thanks for your comment.

I assume you are referring to the eDossier records, which show when each examination report was issued, and when each response is filed. There are two issues with this. Firstly, applications filed prior to about 2006 are not covered by eDossier. Secondly, I believe that the eDossier section of the AusPat record is a view into a document management system, and does not form part of the main database. The only data readily available for analysis (or available at all, prior to 2006) is the first report date, and the (most recent) further report date.

From Figure 4, it looks to me like nearly 80% of all applications filed in 2008 had examination requested (60.5% following a direction and 19.4% voluntarily), while there were 0.3% (when the data was captured) that were still awaiting a direction or request. But I do see your point. However, there is probably little value in discounting for the 20% of applications that will never reach examination when the cost of examination is itself essentially unknown!

It seemed to me that the point of the recommendations at the end of the report was an acknowledgement of precisely the issue you raise. If IP Australia wants to extract practical outcomes from this type of analysis, it needs to capture the right kind of data. The author cannot be blamed for the shortcomings of his inputs. As I said in the opening paragraph, I understand that he was an intern on a three-month stint at IP Australia, so I do not want to be too critical.

Post a Comment