I am interested in knowing for each EP where and how many years they have been renewed. I am using the latest version of PATSTAT online.
From what I read, the best way to proxy for the renewal decisions is to look at the fee payment. So more specifically to look at the relevant code ‘PGFP’.
My SQL query for year 1999 is:
Code: Select all
SELECT TOP 10 t1.appln_id, event_code, fee_country, fee_payment_date, fee_renewal_year
FROM tls201_appln t1
JOIN tls231_inpadoc_legal_event t2
ON t1. appln_id = t2.appln_id
WHERE event_code = 'PGFP'
AND appln_filing_year = 1999
ORDER BY t1.appln_id
But now if I look at lapse event with the relevant code ‘PG25’, results are quite different.
Code: Select all
SELECT t1.appln_id, event_code, lapse_country, lapse_date, lapse_text, year(lapse_date)-year(appln_filing_date) as renewal_year
FROM tls201_appln t1
JOIN tls231_inpadoc_legal_event t2
ON t1. appln_id = t2.appln_id
WHERE event_code = 'PG25'
AND t1.appln_id = 773
ORDER BY t1.appln_id
I understand that lapse events are ‘problematic’ because some countries have long period of grace so a lapse event does not mean necessarily that the patent is not in force anymore. Nevertheless, it seems that using only renewal (PGFP) underestimates the territorial scope of many EPs (another example is appln_id = 775 with only 3 country with PGFP and 19 countries with PG25).
What would be the right approach/trade-off to construct the most accurate renewal variable? I would be tempted to combine renewal and lapse to proxy for renewal. For example, I would say that appln_id = 773 was renewed:
In DE, GB, IT, FR for 16 years AND in NL for 10 years.
What is your view on that?
Thank you for all the work you are doing!
Alexis