Skip to content

Sponsor type needs a richer taxonomy than the current Organization_Type value #46

Description

@nate-layman

Context

Organization.Organization_Type currently takes one of: Department, College, School, Sponsor, Subrecipient, Vendor, Institute, Center. A single bucket called Sponsor collapses everything from NIH to a small family foundation to an industry partner.

For most research admin use cases the type of sponsor drives more downstream behavior than almost any other Organization attribute:

  • Indirect rates and base differ (federal F&A vs. non-federal capped rates vs. negotiated industry rates).
  • Reporting requirements differ (federal cascade of Uniform Guidance, RPPR, financial reports vs. foundation custom letters).
  • Compliance and audit thresholds differ (single-audit threshold applies to federal pass-through; doesn't apply to industry).
  • Pricing and IP terms differ (industry agreements vs. federal grant terms).
  • Diversity-of-funding reporting that universities run for strategic planning needs sponsor-type slices.

Today none of that is queryable from the schema. A consumer has to maintain its own sponsor-type lookup outside the UDM.

What an admin actually needs

At minimum a sponsor-type taxonomy with these distinctions:

  • Federal (with optional federal-agency sub-classification: NIH, NSF, DOE, DOD, etc.)
  • State
  • Local / Municipal
  • Foundation / Private non-profit
  • Industry / For-profit
  • Higher Education (pass-through)
  • Foreign / International
  • Tribal
  • Internal (institutional seed funding)
  • Other

Sub-types of "Federal" matter enough that an institution may want them stored explicitly rather than buried in Organization_Name text.

Possible designs

  1. AllowedValues field on Organization. Add Sponsor_Type_Value_ID referencing AllowedValues, populated by institutions. Lightweight; per-institution flexibility; no CHECK enforcement.

  2. Dedicated CHECK constraint. Add Sponsor_Type directly to Organization with a CHECK-constrained universal taxonomy (the list above). Hard to evolve; consistent across deployments.

  3. Separate SponsorClassification table. A bridge table linking Organization to one or more classification dimensions (sector, level of government, federal agency, etc.). Most expressive; heaviest schema surface.

  4. Hybrid. CHECK-constrained top-level type ("sector") + AllowedValues sub-type ("federal agency", "foundation category") for institution-specific refinement.

Open questions

  • Is sponsor type universal enough (option 2/4) or genuinely institution-specific (option 1)?
  • Does federal agency need to be modeled (NIH vs. NSF vs. DOD) or is that fine as a separate Organization row with a Parent_Organization_ID pointing at "U.S. Federal"?
  • How does this interact with Prime_Sponsor_Organization_ID for pass-through funding (subaward from a foundation that's passing federal money has two relevant sponsor types)?
  • Should AllowedValues be used or is this CHECK-constrained because the categories are externally standardized (NSF SED, IPEDS, etc.)?

Related issues

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions