In setting Federal financial data standards, what industry standards should Treasury and OMB consider?

[philipashlock]

This is just a list of related efforts that I could come up with. There's also a good list of related efforts on the OpenSpending website.

Data Standards

Financial data standards

• XBRL
• LEDES

Government contracting and spending data standards

• Open Contracting Data Standard
• OpenSpending Data Standard

General government data standards

• NIEM

General purpose schemas

• Schema.org
• Dublin Core Metadata Terms

Unique Identifiers & Thesauri

IDs

• Federal program codes and bureau codes, including a crosswalk between OMB Codes, Treasury Codes, and GSA CGAC Codes
• Open Civic Data Division Identifiers (OCD-ID)
• PublicBodies.org/us
• OpenCorporates
• OpenLEIs

Thesauri

I'm not aware of a current or well maintained common vocabulary or taxonomy for categorizing things across the federal government (other than BRM noted below), but several other countries have such systems.

US:

• The Business Reference Model (BRM). A current version is available as an XLS on OMB MAX (requires login/permissions) and an older outdated version of which is available in XML
• Before 2013 metadata submitted to data.gov was categorized with a taxonomy based on the sections of the Census Statistical Abstract, sometimes referred to as the National Data Book Taxonomy.
• The thematic areas of performance.gov might also be worth referencing.
• FGDC Themes - for geospatial data

Elsewhere:

• EuroVoc - Multilingual Thesaurus of the European Union
• UK Integrated Public Sector Vocabulary
• Government of Canada Core Subject Thesaurus

[OMBOFFM]

@philipashlock thanks for those great resources.

[HerschelC]

Wow @philipashlock - great info!

Philip did a great job listing many resources - but it seems like the answer should be that Treasury/OMB consider the standards in place by other governments. Unless there is very compelling evidence, and "we're special" isn't compelling, XBRL seems like the de-facto choice for a data standard framework. It has off the shelf compatibility and an existing internal user base of experiences and process-related rules from which to draw. IMO, creating our own new standard is not a standard that should be considered. Further, most standards are extensible. We can likely find a way to handle any real special cases.

[CharlesHoffmanCPA]

A primary feature that Treasury/OMB should consider when determining which standard(s) to leverage is the expressiveness of the standard. Not all standards are created equal in terms of the sorts of business domain concepts and relations between concepts that they can express. Critical are the ability to express things like mathematical relations between reported information, dimensional relations, whole-part type relations, class-subclass type relations, and so forth.

At one end of the expressiveness scale is the ability to define little more than a "list" of terms or simply a dictionary. A little more expressiveness and you can articulate what would amount to a thesaurus with has basic relations between terms. A little more expressiveness and you can articulate a taxonomy of terms plus core relations. A little more expressiveness and you can express a conceptual model (or entity/relations diagram). The pinnacle of expressiveness is an ontology which can be used to formally express terms, classes of terms, and a very rich set of relations between terms.