At the end of the year, I took a look at whether Alice really had a significant impact on patents as a whole.  The answer was that Alice simply doesn’t affect that many patent applications.  But several important questions were left unanswered.  I also wanted to know whether the affected applications are really being affected by Alice, or if there are other reasons those applications might have been rejected.

We can take a look at that using BigQuery.

Determining The Ultimate Outcome Of Applications Receiving A § 101 Rejection At Some Point In Prosecution

Applications Rejected Under § 101 And Eventually Abandoned

#standardSQL
SELECT DISTINCT peds.patentCaseMetadata.patentGrantIdentification.grantDate, rejs.app_id, peds.patentCaseMetadata.applicationStatusCategory
FROM `patents-public-data.uspto_oce_office_actions.rejections` rejs 
JOIN `patents-public-data.uspto_peds.applications` peds ON peds.patentCaseMetadata.applicationNumberText.electronicText = rejs.app_id
JOIN `patents-public-data.uspto_oce_office_actions.office_actions` oa ON oa.ifw_number = rejs.ifw_number
WHERE peds.patentCaseMetadata.patentGrantIdentification.grantDate IS NULL AND ((alice_in = '1') OR (bilski_in = '1') OR (mayo_in = '1') OR (myriad_in = '1')) AND oa.mail_dt > '2010-06-28' AND peds.patentCaseMetadata.applicationStatusCategory LIKE 'Abandoned%'
ORDER BY app_id

This doesn’t mean that these applications were abandoned because of the § 101 rejection—for example, a number were abandoned even though they were allowed because the applicant failed to pay the issue fee.  It doesn’t even mean that the § 101 rejection wasn’t overcome—in some cases, the § 101 rejection is overcome and rejections over prior art lead to abandonment.  All it means is that at some point a § 101 rejection was received, and the application was ultimately abandoned.

Still, this is the highest possible number of applications that could even arguably be considered to be abandoned for a reason that is in some way related to a § 101 rejection.

Let’s take a look at the opposite question—how many patent applicants overcome a § 101 rejection?

Applications Rejected Under § 101 And Eventually Patented

#standardSQL
SELECT DISTINCT peds.patentCaseMetadata.patentGrantIdentification.grantDate, rejs.app_id, peds.patentCaseMetadata.applicationStatusCategory
FROM `patents-public-data.uspto_oce_office_actions.rejections` rejs 
JOIN `patents-public-data.uspto_peds.applications` peds ON peds.patentCaseMetadata.applicationNumberText.electronicText = rejs.app_id
JOIN `patents-public-data.uspto_oce_office_actions.office_actions` oa ON oa.ifw_number = rejs.ifw_number
WHERE peds.patentCaseMetadata.patentGrantIdentification.grantDate IS NOT NULL AND ((alice_in = '1') OR (bilski_in = '1') OR (mayo_in = '1') OR (myriad_in = '1')) AND oa.mail_dt > '2010-06-28'
ORDER BY app_id

After receiving a § 101 rejection, 23,314 applications subsequently issued as patents, overcoming the § 101 rejection.

The final question is how many applications have neither been abandoned nor issued—how many applications have received a § 101 rejection at some point in time, and remain pending?

Applications Rejected Under § 101 And Still Pending

#standardSQL
SELECT DISTINCT peds.patentCaseMetadata.patentGrantIdentification.grantDate, rejs.app_id, peds.patentCaseMetadata.applicationStatusCategory
FROM `patents-public-data.uspto_oce_office_actions.rejections` rejs 
JOIN `patents-public-data.uspto_peds.applications` peds ON peds.patentCaseMetadata.applicationNumberText.electronicText = rejs.app_id
JOIN `patents-public-data.uspto_oce_office_actions.office_actions` oa ON oa.ifw_number = rejs.ifw_number
WHERE peds.patentCaseMetadata.patentGrantIdentification.grantDate IS NULL AND ((alice_in = '1') OR (bilski_in = '1') OR (mayo_in = '1') OR (myriad_in = '1')) AND oa.mail_dt > '2010-06-28' AND peds.patentCaseMetadata.applicationStatusCategory NOT LIKE 'Abandoned%'
ORDER BY app_id

An additional 19,286 applications remain pending.  Of these, the § 101 rejection has been overcome in some of them.  For example, the oldest pending case which received a Bilski/Alice/Mayo/Myriad rejection at some point in time only has prior art rejections under § 103 remaining.  (It’s still actively being prosecuted, 10 years after it was filed.)

So, from this data, we can tell that, of the 64,647 applications receiving a § 101 rejection at some point in prosecution, 36% of applications have issued as patents, 34.1% have subsequently been abandoned, and 29.8% remain pending.

Overcoming § 101 Rejections

But, as mentioned above, that isn’t enough; just because an application received a § 101 rejection and was subsequently abandoned doesn’t mean that it was abandoned due to the § 101 rejection.  For that, we need a more specific query.  First, we generate a table containing the most recent office action for each application ID in the database.

Sub-Table: Most Recent Office Action Per Application

#standardSQL
SELECT oas.*
FROM `patents-public-data.uspto_oce_office_actions.office_actions` oas 
LEFT OUTER JOIN `patents-public-data.uspto_oce_office_actions.office_actions` oas2 ON oas.app_id = oas2.app_id
  AND (oas.mail_dt < oas2.mail_dt OR oas.mail_dt = oas2.mail_dt AND oas.app_id < oas2.app_id)
WHERE oas2.app_id IS NULL

After saving this table to a local dataset (in my case, I named it alice_drizzle_data.most_recent_oas_by_app_id), we can combine that table with our earlier abandonment search, and look for those applications where the last office action prior to abandonment does not contain a § 101 rejection.

Abandoned Applications Where A § 101 Rejection Was Overcome

#standardSQL
SELECT DISTINCT oas.*
FROM `patents-public-data.uspto_oce_office_actions.rejections` rejs 
JOIN `patents-public-data.uspto_peds.applications` peds ON peds.patentCaseMetadata.applicationNumberText.electronicText = rejs.app_id
INNER JOIN `alice_drizzle_data.most_recent_oas_by_app_id` oas ON oas.app_id = peds.patentCaseMetadata.applicationNumberText.electronicText
WHERE peds.patentCaseMetadata.patentGrantIdentification.grantDate IS NULL AND ((alice_in = '1') OR (bilski_in = '1') OR (mayo_in = '1') OR (myriad_in = '1')) AND oas.mail_dt > '2010-06-28' AND peds.patentCaseMetadata.applicationStatusCategory LIKE 'Abandoned%' AND oas.rejection_101 = '0'
ORDER BY app_id

This query tells us that, of the 22,047 patent applications which were abandoned after receiving a § 101 rejection, in 4,047 cases, the final office action prior to abandonment did not contain a § 101 rejection.  In other words, 18% of those applications do not appear to have been abandoned due to § 101.

We can also perform a similar test to determine how many of the pending applications have overcome their § 101 rejections, so we can determine the possibility of overcoming an Alice rejection.

Pending Applications Where A § 101 Rejection Was Overcome

#standardSQL
SELECT DISTINCT oas.*
FROM `patents-public-data.uspto_oce_office_actions.rejections` rejs 
JOIN `patents-public-data.uspto_peds.applications` peds ON peds.patentCaseMetadata.applicationNumberText.electronicText = rejs.app_id
INNER JOIN `alice_drizzle_data.most_recent_oas_by_app_id` oas ON oas.app_id = peds.patentCaseMetadata.applicationNumberText.electronicText
WHERE peds.patentCaseMetadata.patentGrantIdentification.grantDate IS NULL AND ((alice_in = '1') OR (bilski_in = '1') OR (mayo_in = '1') OR (myriad_in = '1')) AND oas.mail_dt > '2010-06-28' AND peds.patentCaseMetadata.applicationStatusCategory NOT LIKE 'Abandoned%' AND oas.rejection_101 = '0'
ORDER BY app_id

This tells us that, of the 19,286 currently pending applications that received a § 101 rejection at some point in their prosecution, 2,896 no longer have a pending § 101 rejection.

Based on this, we can determine that, of the 64,647 applications that received Bilski/Alice/Mayo/Myriad § 101 rejections, the rejection was overcome in 46.8% of cases.  And again, referring back to my earlier post, most applications don’t receive a § 101 rejection in the first place.  Alice simply does not impact most patent applications.

Applications Where § 101 Was The Sole Reason For Abandonment

Finally, we can look at how many applications we can state with certainty were abandoned due to a § 101 rejection—the instances in which there were no other pending rejections in the last office action before abandonment.

Applications Abandoned Due Solely To A § 101 Rejection

#standardSQL
SELECT DISTINCT oas.*
FROM `patents-public-data.uspto_oce_office_actions.rejections` rejs 
JOIN `patents-public-data.uspto_peds.applications` peds ON peds.patentCaseMetadata.applicationNumberText.electronicText = rejs.app_id
INNER JOIN `alice_drizzle_data.most_recent_oas_by_app_id` oas ON oas.app_id = peds.patentCaseMetadata.applicationNumberText.electronicText
WHERE peds.patentCaseMetadata.patentGrantIdentification.grantDate IS NULL AND ((alice_in = '1') OR (bilski_in = '1') OR (mayo_in = '1') OR (myriad_in = '1')) AND oas.mail_dt > '2010-06-28' AND peds.patentCaseMetadata.applicationStatusCategory LIKE 'Abandoned%' AND oas.rejection_101 = '1' AND oas.rejection_102 = '0' AND oas.rejection_103 = '0' AND oas.rejection_112 = '0' AND oas.rejection_dp = '0' AND oas.objection = '0'
ORDER BY app_id

That tells us that there’s only 2,164 abandoned applications where the only rejection in the last office action was a § 101 rejection.  This lets us illustrate the breakdown of applications as follows

In other words, § 101 is the sole reason for abandonment in a grand total of 3.3% of cases receiving a § 101 rejection, and of only 9.8% of abandoned applications receiving a § 101 rejection.  § 101 rejections are actually overcome more than ten times as often as they represent the sole remaining block to patentability.

The Alice drizzle barely even merits an umbrella.