How to deal with redistribution of a grant in the standard (AKA - regranting)?


#1

Currently, the 360Giving standard takes into consideration the simple case where a grant is granted in a direct connection to a grantee:
Funding body ----> Grantee

However, we are seeing more and more cases where a grant is granted to an organisation that then re-grant the amount. For example, a foundation can grant money to an institituion that then can re-grant the money to an individual or other institutions.
Funding body ----> Another institution —> Grantee

Following up from Github, I have the following questions for this disucssion:

  1. How critical is the regranting issue to publishers/users?
  2. How should we deal with regranting? Should we create a new field in the schema?

#2

Moving this on a bit from the discussion in Github…

Original proposal (stevieflow)

The simple was to deal with this is that we have a field "regranting" which can be set to true or false.

The problem I can see with that is who uses that field? If BIG FUNDER makes the original grant to SMALLER FUNDER then does BIG FUNDER use "regranting": true against their award?

If SMALLER FUNDER does the same in their published data then it would be likely that both grants are excluded from analysis (with the intention of avoiding double-counting, it is in effect nil-counting).

The alternative to this might be two new (non-required) boolean fields "For redistribution?" and "Regrant?". I think with that it is clearer who uses which field.

Another proposal - linking regranting

A more complex solution might be that the initial grant is flagged as being for redistribution, with any publication of actual regrants identified and (optionally) linked to the original flow of money.

So, e.g. From Big Lottery to Local Trust,

  • Big Lottery would use a field "For redistribution?" :true
  • Local Trust, in publishing their regrants of this money, could use the field "Regrant?": true, and could also use a field "Regrant:Identifier" with the string value "360G-blf-0030116147". I think in this case the money that Local Trusts has is going to be redistributed again, so it is possible for them to use "For redistribution?": true as well
  • Any organisation that Local Trust funds which subsequently makes another regrant of that money can use "Regrant?": true and then link to the previous Local Trust grant with that identifier, allowing a “chain” of grants to be followed.

If I were doing the analysis on this I could exclude any totals which are "For redistribution?": true

Advantages of this:

  • It’s easier to break down the responsibility for flagging money that will be regranted by putting it on the larger funder/publisher to identify as being for redistribution.
  • By dividing into "For redistribution?" and "Regrant?" the data can show where grants are redistributed more than once (as in the above example).

Disadvantages:

  • Less straightforward. Regranters need to do more work to find the identifer for the grant that they are redistributing (is that a big burden?)

#3

Simplified example of a more complex regranting chain