Counting domestic patents applied internationally

Here you can post your opinions, ask questions and share experiences on the PATSTAT product line. Please always indicate the PATSTAT edition (e.g. 2015 Autumn Edition) and the database (e.g. PATSTAT Online, MySQL, MS SQL Server, ...) you are using.

Ucindami Mafeni
Posts: 9
Joined: Wed Oct 09, 2019 3:52 pm

Counting domestic patents applied internationally

Post by Ucindami Mafeni » Wed Oct 09, 2019 4:12 pm

Hello,

I wanted to ask for help countring the number of patents which came from a domestic inventor but have been applied overseas. I am wary of interpreting this to be the meaning of the appln_auth variable, as I fear it may the case that simply applying for a patent from a particular authority does not mean that the patent is only applicable within the country that the authority oversees. My best guess would be to use the tsl231 dataset and look at the variable 'designated_states', however, I am unsure. May you please advise me on how to find out which international countries a domestically produced invention has a patent in? Furthermore, mayou please help me to discover the covers; the number of patents issued by international investors that are applicable to the domestic country in question?

Many thanks


EPO / PATSTAT Support
Posts: 425
Joined: Thu Feb 22, 2007 5:33 pm
Contact:

Re: Counting domestic patents applied internationally

Post by EPO / PATSTAT Support » Tue Oct 15, 2019 11:39 am

Hello Ucindami,
your forum post touches upon a number of different issues.
The attribute: "application authority" (appln_auth) defines the territorial area where the patent is applied for. If the the code is a real country code: DE, US, JP,.... then the patent is applied for in that country and nowhere else. The application authority can also be WO (WIPO) , EP (EPO) or any of the other regional offices.

Specific for the EPO, the application can be "designated" to EPO member states, but the designation data is not (reasonably) suitable for statistical analysis because of past changes in the fee structure. The fee structure was steering applicant behaviour with regards to the designated countries and had (have) little to do with the intention of the applicant wanting protection in one of the member states. Therefore I recommend researchers to look at the renewal fee payments via the code PGFP. (There a couple of other forum posts on this topic.)

For other countries (patent authorities) I recommend users to look at the DocDB patent family if they need information on the geographical coverage of the protection of an invention.

A "domestic inventor" or "domestic invention" can be defined by looking at the country code of the inventor and the application authority. It gets a bit complicated then with EP applications. Would you consider a German inventor filing and EP application as being "domestic" ? That would be fine if your analysis would be about comparing "Europe" with the "US" (comparable sizes.?..) . But it would not be useful to compare DE with FR. Therefore researchers often look at the priority filing to "define" the origin of the invention.
Your question on "international investors that are applicable to the domestic country in question" would need some refinement. How would you identify / define "international investors" in a context of patent inventors and patent applicants ?
PATSTAT Support Team
EPO - Vienna
patstat @ epo.org


Ucindami Mafeni
Posts: 9
Joined: Wed Oct 09, 2019 3:52 pm

Re: Counting domestic patents applied internationally

Post by Ucindami Mafeni » Wed Oct 16, 2019 2:00 pm

Thank you for the prompt reply.

1) I think I made a typo when I used the word "investor". I meant inventor, apologies.

2) in refence to the question in your reply " Would you consider a German inventor filing and EP application as being "domestic" ?". My intention is to capture the level of innovation at the country level. I desire to capture the number of patents successfuly applied for by inventors of a particular country X and the number of foreign inventions that have patents that must be respected by X. Hence, domestic in this case would be country X and foreign would be everyone else. Going back to your Germany example, if I there was a patented invention produced by a German then it it would count as "domestic" for Germany. If the same patent was upheld in France, then it would be recorded as "Foreign" for France. My issue was inding a variable which could help me find these distinctions for every successful patent application. Would I use the priority filing for this?

3) with regards to identifying "designated states". You say that the Legal event code PGFP would be best for this. I verified from other querries in the forum that this was indeed a good proxy for whether or not an EP was enforced within a country and the identity of the said country. Many Thanks


Ucindami Mafeni
Posts: 9
Joined: Wed Oct 09, 2019 3:52 pm

Re: Counting domestic patents applied internationally

Post by Ucindami Mafeni » Sat Oct 19, 2019 10:25 pm

Another question: While the PGFP variable is the best proxy for whether an EP was enforced within a particular country, what is the best method to determine whether or not the EP has lasped yet/identify whether it is still valid?


EPO / PATSTAT Support
Posts: 425
Joined: Thu Feb 22, 2007 5:33 pm
Contact:

Re: Counting domestic patents applied internationally

Post by EPO / PATSTAT Support » Tue Oct 22, 2019 5:20 pm

"My intention is to capture the level of innovation at the country level."
Many researchers are using the first priority filing to measure innovative capacity. But it does not work for EP and WO applications (or other regional offices) where no priority was filed.
... In which case they mostly use the country code of the INVENTOR. But kindly consult the literature on pro's an con's.

Validity an lapsing: the code PG25 is used by member states to indicate a lapse. Though the "absence of the PGFP" (for the last 18 months) is just as well a good proxy.
I would combine and analyse the differences between the 2 approaches because the quality and coverage of PG25 can (does) differ depending the member state. (quality & coverage on pgfp is very good)
“WARNING: LAPSES OF ITALIAN PATENTS WITH EFFECTIVE DATE BEFORE 2007 MAY HAVE OCCURRED AT ANY TIME BEFORE 2007. THE CORRECT EFFECTIVE DATE MAY BE DIFFERENT FROM THE ONE RECORDED IN PATSTAT AND INPADOC DATA.”

Geert Boedt
PATSTAT Support Team
EPO - Vienna
patstat @ epo.org


Ucindami Mafeni
Posts: 9
Joined: Wed Oct 09, 2019 3:52 pm

Re: Counting domestic patents applied internationally

Post by Ucindami Mafeni » Fri Oct 25, 2019 3:23 pm

Hello again,

When I tried using the PGFP variable as a means of detecting patent enforcement, I find that the range of years which in which such payments were made for the particular kind of patent I am interested in ranges from 1981-2017. This implies that prior to 1981, there were no patents being applied for for this particlar class of patent. This makes me question:

1)does the validity of PGFP being a proxy for patent enforcement remain true for years prior to the 2000s, such that we can take this implication to be true?

Or

2) Is it the case that there were indeed successully enforced patents for this class of patent prior to 1981, but that we cannot accurately measure them for these years using PGFP?

If point 2 is true, is there a better way of identifying patent enforcement for the years prior to yhe 1980s?


EPO / PATSTAT Support
Posts: 425
Joined: Thu Feb 22, 2007 5:33 pm
Contact:

Re: Counting domestic patents applied internationally

Post by EPO / PATSTAT Support » Mon Nov 04, 2019 3:05 pm

1980 was the first year EP patents were granted.

Code: Select all

select year(publn_date) grant_date , count(appln_id)  #_applications
  from tls211_pat_publn
  where publn_kind = 'B1' and publn_auth = 'EP'
  group by year(publn_date)
  order by year(publn_date)
So it makes sense that no renewal fees records are available from before 1980.

You have to keep in mind that renewal fee payment records ONLY contain the record for the latest renewal fee being paid for a specific country.
Example
EP0000279B1 was granted on 1981-10-14 (filed 1978-06-27).
PGFP code indicates that payment for year 5 was paid on 1982-06-30 in FR,BE,DE and NL.
So no further fees were paid later on.

Code: Select all

SELECT publn_auth, publn_nr, publn_kind, publn_date, appln_filing_date
      ,[event_seq_nr]
      ,[event_auth]
      ,[event_code]
      ,[event_publn_date]
      ,[event_effective_date]
      ,[event_descr]
      ,[fee_country]
      ,[fee_payment_date]
      ,[fee_renewal_year]
      ,[fee_text]
FROM [tls231_inpadoc_legal_event]
  join tls211_pat_publn on tls231_inpadoc_legal_event.appln_id = tls211_pat_publn.appln_id
  join tls201_appln on tls211_pat_publn.appln_id = tls201_appln.appln_id
where event_code = 'pgfp' and tls211_pat_publn.appln_id = 16429299 and publn_kind = 'B1'
If you do the same for tls211_pat_publn-appln_id = 15724683 (granted 2009), you will see that the renewal fees were paid in NL and DE for year 15 on 2014-10-22.
Therefore one can conclude that renewal fees were also paid for the previous years. (2013 --> for year 14, 2012 for year 13, 2011 for year 12, etc...)
PATSTAT Support Team
EPO - Vienna
patstat @ epo.org


Ucindami Mafeni
Posts: 9
Joined: Wed Oct 09, 2019 3:52 pm

Re: Counting domestic patents applied internationally

Post by Ucindami Mafeni » Tue Feb 25, 2020 11:32 am

Hi again,

I am still yet to hear a response regarding how I can identify regional applications (i.e. situations in which appln_auth = EP, for example) enforced nationally (i.e. which EP countries the application is designated to) in the years prior to 1980 (i.e. before PGFP was used). How can I do this? Thus far, I have not used PGFP for fear of losing observations.


Ucindami Mafeni
Posts: 9
Joined: Wed Oct 09, 2019 3:52 pm

Re: Counting domestic patents applied internationally

Post by Ucindami Mafeni » Tue Feb 25, 2020 11:36 am

Furthermore, I wanted to disaggregate regional applications (i.e. those made to the EP or EA) so that I can identify the individual countries in which a particular application is enforced. I have disaggregated applications made to the EP and EA by, using TLS201 as the starting dataset:
1) merging the corresponding application ID's to TLS231
2) restricting the sample to applications in the regional and not the national phase
3) merging this combined dataset to TLS803 using application authority and legal event code as merging variables
4) Restricting the sample to observations with a non-regional application authority and with a positive event impact (i.e. those with event_impact="+").

Does this seem like a suitable strategy, or are there better ways to disaggaregate applications made to the EP/EA so I can know which individual countries the application is enforced in?


EPO / PATSTAT Support
Posts: 425
Joined: Thu Feb 22, 2007 5:33 pm
Contact:

Re: Counting domestic patents applied internationally

Post by EPO / PATSTAT Support » Wed Feb 26, 2020 11:59 am

Ucindami Mafeni wrote:
Tue Feb 25, 2020 11:32 am
Hi again,

I am still yet to hear a response regarding how I can identify regional applications (i.e. situations in which appln_auth = EP, for example) enforced nationally (i.e. which EP countries the application is designated to) in the years prior to 1980 (i.e. before PGFP was used). How can I do this? Thus far, I have not used PGFP for fear of losing observations.
The first European patents were granted in 1980, so no national renewal fee records are available before that date. Run this query to see the first granted EP applications:

Code: Select all

SELECT TOP (10) pat_publn_id
      ,publn_auth+publn_nr+publn_kind as number
      ,publn_date, publn_first_grant
FROM tls211_pat_publn
WHERE publn_auth = 'EP' and publn_first_grant = 'Y' 
ORDER BY publn_date asc
PATSTAT Support Team
EPO - Vienna
patstat @ epo.org


Post Reply