Human v non-human officers

In the appointment list of an officer, there is a boolean field “is_corporate_officer”. We have found that this is often incorrect. We suspect the underlying cause is that the submitter or data-checker at CH asks themselves whether the officer is a corporate (a company) rather than whether the officer is human. This then excludes other non-corporate, non-human officers such as a partnership or a government office (an ex officio position). For examples:

A clearer field would be “is_human” which the filer (or whoever checks the data) can easily answer true or false.

I raised a similar point in March 2017 - I suspect the same underlying process is the issue. I recall finding that such misclassifications in both company officers and officer appointments (as you report) were not uncommon. I don’t have a list of examples to hand but if CH were minded to investigate I’m sure I could dig some out.

My query related to the categorisation of non-human entities in the company officers list using the “human” values for “officer_role” (e.g. a company being classified as “director” rather than “corporate-director”):

@mfairhurst advised that they’d investigate this as a data issue and they should differentiate between human and non-human roles. However I didn’t hear back and I just rechecked a couple of examples and the problem still exists.

For info: from my post above my first two examples have disappeared from Beta/API data set
( /company/06909966/officers, /company/06903160/officers) but the following are still present and incorrect (as they were in Mar 2017):

/company/03196419/officers - CORPORATE NOMINEES LIMITED - listed with “officer_role”: “director”

/company/03184321/officers - “BONDLAW SECRETARIES LIMITED” - “officer_role”: “nominee-secretary”

My post also references this earlier post:

May I suggest that Companies House does a quick clean-up as follows:

  1. Search for all non-corporate directors of companies and members of LLPs whose names contain any of these words: And, Associated, Atkiengesellschaft, Berhad, Company, Co., Corporation, Corp, Group, Holdings, International, Limited, Ltd, LLP, Private, Pte, Pty, Societe, Society. (and any other words that are often found in non-human names and almost never in human names).
  2. Convert all of those to the corporate equivalent, because very few humans would have those words in their name. If this results in a very few errors, it would be far better than the thousands of errors in classification currently existing.