Wrong trend in patent filings?

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.
Post Reply

SMaurischat
Posts: 1
Joined: Thu Dec 24, 2015 3:37 am

Wrong trend in patent filings?

Post by SMaurischat » Thu Dec 24, 2015 3:46 am

Hey Guys,

When i try to count patents over the years in Germany, there is a clear trend visible which does not make any sense to me.


select count(distinct a.appln_id) , year(a.appln_filing_date)
from tls201_appln a

where a.appln_auth = 'DE'
and a.appln_kind in ('A')
and a.granted = 1
group by year(a.appln_filing_date)
order by year(a.appln_filing_date)


Image

any quick idea here? I really start to mistrust the data quality provided here..
Thanks and rgrds,
Seb


mkracker
Posts: 114
Joined: Wed Sep 04, 2013 6:17 am
Location: Vienna

Re: Wrong trend in patent filings?

Post by mkracker » Mon Jan 04, 2016 4:17 pm

Hi Seb,

The downward trend of German granted applications in your result in the last years is clearly visible, but I still think the data is fine. You must be aware that it takes years (in some cases more years than you would expect) for an application to get granted. So for a young application chances are good that it is not yet granted, but will be granted in the future - or not at all. All these not-yet-granted applications lead to this typical downward trend in the last years.

To see the "loose" correlation between filing year and grant year (= publication year of the granting publication), you may try this query:

Code: Select all

select year(appln_filing_date), year(publn_date), count(*)
from tls201_appln a
join tls211_pat_publn p on a.appln_id = p.appln_id
where a.appln_auth = 'DE'
and a.appln_kind in ('A')
and a.granted = 1
and p.publn_first_grant = 1  -- indicates the granting publication
and year(publn_date) between 1995 and 2015  -- some sensible restriction
group by year(appln_filing_date), year(publn_date)
order by year(appln_filing_date), year(publn_date)
When you load the result into Excel and create a pivot table, you get this matrix, showing how many applications of a certain year have been granted in a certain year. For the sake of readability, I removed outliners. The numbers in the yellow line are exactly the numbers of your table.
pivotTable3.PNG
pivotTable3.PNG (202.26 KiB) Viewed 987 times
In Germany this delay effect is intensified, because examination does not take place automatically. The applicant must specifically request an examination, and he can wait up to 7 years to do so. Some other offices, e. g. Japan, have a similar policy for deferred examination.

There could be additional effects, like changes in the German procedures, but I am not aware of these.

It would be good to verify the numbers with original German data. I tried this with DEPATISnet, but did not succeed - which is likely my fault.

Probably somebody with better insight into German grant numbers could add to this post?
-------------------------------------------
Martin Kracker / EPO


Post Reply