Planning: The Nation

The fourth module climbs to the top rung of the ladder. Part II worked at the level of the city, Part III at the county, and Part IV at the state. Part V works at the level of the federal government, the largest producer of public data in the United States and the source that city, county, and state analyses depend on for denominators, baselines, and exposure estimates. It is also the level at which the public record is most exposed to wholesale change: a single transition can retire a dataset that thousands of local analyses cite.

Pressure and value. The pressure is erosion: the mundane and sometimes deliberate decay of the infrastructures that let the public return to its own evidence. Portals are reorganized, programs lose appropriations, data products are “retired” in one-paragraph notices, and series that ran for decades stop updating. The value that pushes back is ownership, which Keegan (2026) defines not as property but as collective, accountable governance over how public-relevant data is preserved, accessed, and used. Here ownership takes four forms: stewardship, collective governance, sovereignty (the claim of Indigenous nations to govern data about themselves), and refusal (the decision that some data should not exist in the form proposed).

Lineage and genre. The lineage is planning, joined by civic technology and by the critical traditions that question the framework itself. Planning learned, through the wreckage of mid-century urban renewal, to ask who was in the room when a decision about shared infrastructure was made, who was not, and whose knowledge counted. Those are the questions to ask when a federal dataset disappears. The genre is the public comment: a written, evidence-backed argument submitted into a federal rulemaking docket, which the agency must consider and which becomes part of the record a court will review.

Cases. The national anchor is the removal of federal environmental and public-health datasets across the 2017, 2021, and 2025 transitions, and the volunteer preservation response of the Environmental Data and Governance Initiative, DataRefuge, and their successors. The local face of federal memory is Boulder’s federal research infrastructure (NOAA’s David Skaggs Research Center, NIST’s Boulder laboratories, and the National Center for Atmospheric Research), where federal records are made a few miles from campus and where federal erosion lands as local loss. The non-US counter-cases are Māori data sovereignty as articulated by Te Mana Raraunga, the CARE Principles for Indigenous Data Governance, and the First Nations principles of OCAP in Canada, each of which questions whether stewardship by an outside institution is the right relationship at all. The refusal case is the Stop LAPD Spying Coalition’s “Algorithmic Ecology.” The federal case is a federal dataset at risk and a live rulemaking on regulations.gov, framed by Case 3 of Keegan (2026), “sustaining archives in the Anthropocene.”

Portfolio piece. Piece 4 has the same four parts as every piece. The technical artifact is an erosion audit of a federal dataset (link rot, Wayback recovery, datasheet) plus either a stewardship plan or a refusal specification, as a reproducible notebook or small repository a classmate can run. The public text is a substantive public comment on a live federal rulemaking, ready to submit on regulations.gov. The installed-base note (about 300 words) names the pressure you met and the installed-base elements your artifact built and left missing, with provenance, license, and an AI-use disclosure. Graduate students in INFO 5871 add a methods memo (750 to 1,000 words) that situates the piece in at least five scholarly sources and defends one methodological choice.

Course weeks. This part covers Weeks 13 through 15. Week 13 pairs erosion with planning and ends in a lab that audits a federal dataset. Week 14 pairs ownership with critical data studies and refusal and ends in a lab on stewardship plans and refusal specifications. Week 15 teaches the public comment and ends in a studio for Piece 4.

Roadmap.

16 Erosion: What Can Be Remembered? names the pressure. You audit data.gov records for link rot, recover losses from the Wayback Machine, and write a datasheet for a federal dataset you did not make.
17 Planning introduces the lineage through the Central 70 highway project and its federal environmental review: a stakeholder map, a “who was missing” analysis, and a geopandas map, then the same tools applied to your dataset.
18 Ownership develops the value: tiered access to a federal campaign finance mirror, a threat model, and a stewardship plan, set against Indigenous data sovereignty.
19 Critical Data Studies, Data Justice, and Refusal turns the framework on itself with a category audit of a federally defined homelessness data system and a refusal specification.
20 Public Comments teaches the genre with the regulations.gov API and ends in Piece 4.

Keegan, Brian C. 2026. “Public Interest Data Infrastructuring.” Under Review.

# Planning: The Nation The fourth module climbs to the top rung of the ladder. Part II worked at the level of the city, Part III at the county, and Part IV at the state. Part V works at the level of the **federal government**, the largest producer of public data in the United States and the source that city, county, and state analyses depend on for denominators, baselines, and exposure estimates. It is also the level at which the public record is most exposed to wholesale change: a single transition can retire a dataset that thousands of local analyses cite. **Pressure and value.** The pressure is **erosion**: the mundane and sometimes deliberate decay of the infrastructures that let the public return to its own evidence. Portals are reorganized, programs lose appropriations, data products are "retired" in one-paragraph notices, and series that ran for decades stop updating. The value that pushes back is **ownership**, which Keegan [-@keegan2026] defines not as property but as collective, accountable governance over how public-relevant data is preserved, accessed, and used. Here ownership takes four forms: stewardship, collective governance, sovereignty (the claim of Indigenous nations to govern data about themselves), and refusal (the decision that some data should not exist in the form proposed). **Lineage and genre.** The lineage is **planning**, joined by civic technology and by the critical traditions that question the framework itself. Planning learned, through the wreckage of mid-century urban renewal, to ask who was in the room when a decision about shared infrastructure was made, who was not, and whose knowledge counted. Those are the questions to ask when a federal dataset disappears. The genre is the **public comment**: a written, evidence-backed argument submitted into a federal rulemaking docket, which the agency must consider and which becomes part of the record a court will review. **Cases.** The *national anchor* is the removal of federal environmental and public-health datasets across the 2017, 2021, and 2025 transitions, and the volunteer preservation response of the Environmental Data and Governance Initiative, DataRefuge, and their successors. The *local face of federal memory* is Boulder's federal research infrastructure (NOAA's David Skaggs Research Center, NIST's Boulder laboratories, and the National Center for Atmospheric Research), where federal records are made a few miles from campus and where federal erosion lands as local loss. The *non-US counter-cases* are Māori data sovereignty as articulated by Te Mana Raraunga, the CARE Principles for Indigenous Data Governance, and the First Nations principles of OCAP in Canada, each of which questions whether stewardship by an outside institution is the right relationship at all. The *refusal case* is the Stop LAPD Spying Coalition's "Algorithmic Ecology." The *federal case* is a federal dataset at risk and a live rulemaking on regulations.gov, framed by Case 3 of Keegan (2026), "sustaining archives in the Anthropocene." **Portfolio piece.** Piece 4 has the same four parts as every piece. The **technical artifact** is an erosion audit of a federal dataset (link rot, Wayback recovery, datasheet) plus either a stewardship plan or a refusal specification, as a reproducible notebook or small repository a classmate can run. The **public text** is a substantive public comment on a live federal rulemaking, ready to submit on regulations.gov. The **installed-base note** (about 300 words) names the pressure you met and the installed-base elements your artifact built and left missing, with provenance, license, and an AI-use disclosure. Graduate students in INFO 5871 add a **methods memo** (750 to 1,000 words) that situates the piece in at least five scholarly sources and defends one methodological choice. **Course weeks.** This part covers **Weeks 13 through 15**. Week 13 pairs erosion with planning and ends in a lab that audits a federal dataset. Week 14 pairs ownership with critical data studies and refusal and ends in a lab on stewardship plans and refusal specifications. Week 15 teaches the public comment and ends in a studio for Piece 4. **Roadmap.** - @sec-ch-16-erosion names the pressure. You audit data.gov records for link rot, recover losses from the Wayback Machine, and write a datasheet for a federal dataset you did not make. - @sec-ch-17-planning introduces the lineage through the Central 70 highway project and its federal environmental review: a stakeholder map, a "who was missing" analysis, and a `geopandas` map, then the same tools applied to your dataset. - @sec-ch-18-ownership develops the value: tiered access to a federal campaign finance mirror, a threat model, and a stewardship plan, set against Indigenous data sovereignty. - @sec-ch-19-critical-data-studies-refusal turns the framework on itself with a category audit of a federally defined homelessness data system and a refusal specification. - @sec-ch-20-public-comments teaches the genre with the regulations.gov API and ends in Piece 4.