Journalism: The City

Part II is the first of the book’s four modules. Each module pairs one professional lineage with one level of government, follows the same order (pressure, lineage, value, genre), and ends in one portfolio piece. This module pairs journalism with the city. Its pressure is enclosure, the narrowing of who can observe the systems that govern public life. Its value is openness, understood not as a file at a URL but as the durable legibility of evidence. Its genre is the op-ed. In the course it covers weeks 3 through 5.

Why the city

City government is where most residents meet the state: housing, zoning, permits, police budgets, transit, and the elections that decide them. It is also where the public record is richest and least usable. In practice, as Keegan (2026) argues in the case of mining municipal archives, city records are often subject to soft enclosure: posted as scanned PDFs, scattered across vendor platforms, and hard to follow from an agenda item to a vote to a contract to an outcome. You can see enclosure up close at the city level, and do something about it with tools a laptop can run.

The local cases are the City of Boulder and the City and County of Denver: a Boulder dataset that is open in every formal sense and still hard to read, and Denver’s council records, whose metadata is published through an API while its substance sits in PDFs.

The lineage

Journalism treats access to information as a democratic requirement. Its contribution to data science is a set of working routines (sourcing, corroboration, right of reply, and publication with documentation) that keep systems observable and contestable when powerful actors would prefer otherwise.

The cases

National anchor: ProPublica’s “Machine Bias.” The 2016 investigation of the COMPAS risk score published its data and notebook, which is why you can reproduce it and why a rejoinder and a research literature could follow.
Non-US counter-case: the Pegasus Project. Forbidden Stories and Amnesty International’s Security Lab coordinated reporters across countries to investigate spyware, and published their forensic method as a tool others could run. It tests what journalism’s routines look like when the subject is a state and the risk is physical.
Enclosure cases. Reddit’s API pricing and the end of Pushshift (US) show enclosure at platform scale. Article 40 of the EU’s Digital Services Act (non-US) shows a legal counter-move: researcher access required by law rather than granted by a platform.
Local case: Boulder and Denver city records. Open data tables and council records, read through Keegan’s Case 1.

The portfolio piece

Piece 1 has four parts:

Technical artifact. Reproduce one published finding (from “Machine Bias” or a local data story) and extend it with city data that you scraped or downloaded responsibly, with a documented provenance file.
Public text. A 750-word op-ed for a named local outlet, plus a one-paragraph pitch.
Installed-base note (about 300 words): the pressure you met, the installed-base elements your artifact built and those still missing, with provenance, license, and an AI-use disclosure.
Graduate methods memo (INFO 5871 only; 750 to 1,000 words): the piece situated in at least five scholarly sources, defending one methodological choice.

Each chapter in this part has an exercise that builds a component of the piece, and the last exercise of 7 Op-Eds is the piece itself.

Roadmap

4 Enclosure: Who Can Observe? (week 3) names the pressure. It traces platform enclosure, introduces soft enclosure at city hall, and builds a responsible scraper that writes a provenance log. Its Piece 1 exercise is an access memo and the start of your provenance file.
5 Journalism (week 3) presents the lineage. You reproduce “Machine Bias,” compare your numbers to the published ones, read the Flores, Bechtel, and Lowenkamp rejoinder, and set the Pegasus Project beside it. Its Piece 1 exercise is a reproduction plan.
6 Openness (week 4) builds the value. You ingest, clean, document, and publish Boulder’s election contributions data, and trace a decision through city meeting records to see where posting stops and publishing begins. Its Piece 1 exercise packages your city data as a publication.
7 Op-Eds (week 5) is the genre. It covers the local outlet ecosystem, who gets published, the anatomy of a 750-word argument, and pitching, with a worked example on Boulder data. It ends with Piece 1.

One piece of logistics links this module to the next. The week 4 lab ends with a preview of Part III: you browse the Public Records Archive in Boulder County’s Open Records Center, where the county posts Colorado Open Records Act requests that others have already filed and the county has already answered, and you shortlist released requests to analyze. You do not file new requests for Part III. 9 Law covers the legal details. Part III moves up the ladder to the county, with a new lineage, pressure, and case.

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

# Journalism: The City Part II is the first of the book's four modules. Each module pairs one professional lineage with one level of government, follows the same order (pressure, lineage, value, genre), and ends in one portfolio piece. This module pairs **journalism** with the **city**. Its pressure is **enclosure**, the narrowing of who can observe the systems that govern public life. Its value is **openness**, understood not as a file at a URL but as the durable legibility of evidence. Its genre is the **op-ed**. In the course it covers weeks 3 through 5. ## Why the city City government is where most residents meet the state: housing, zoning, permits, police budgets, transit, and the elections that decide them. It is also where the public record is richest and least usable. In practice, as Keegan [-@keegan2026] argues in the case of *mining municipal archives*, city records are often subject to soft enclosure: posted as scanned PDFs, scattered across vendor platforms, and hard to follow from an agenda item to a vote to a contract to an outcome. You can see enclosure up close at the city level, and do something about it with tools a laptop can run. The local cases are the City of Boulder and the City and County of Denver: a Boulder dataset that is open in every formal sense and still hard to read, and Denver's council records, whose metadata is published through an API while its substance sits in PDFs. ## The lineage Journalism treats access to information as a democratic requirement. Its contribution to data science is a set of working routines (sourcing, corroboration, right of reply, and publication with documentation) that keep systems observable and contestable when powerful actors would prefer otherwise. ## The cases - **National anchor: ProPublica's "Machine Bias."** The 2016 investigation of the COMPAS risk score published its data and notebook, which is why you can reproduce it and why a rejoinder and a research literature could follow. - **Non-US counter-case: the Pegasus Project.** Forbidden Stories and Amnesty International's Security Lab coordinated reporters across countries to investigate spyware, and published their forensic method as a tool others could run. It tests what journalism's routines look like when the subject is a state and the risk is physical. - **Enclosure cases.** Reddit's API pricing and the end of Pushshift (US) show enclosure at platform scale. Article 40 of the EU's Digital Services Act (non-US) shows a legal counter-move: researcher access required by law rather than granted by a platform. - **Local case: Boulder and Denver city records.** Open data tables and council records, read through Keegan's Case 1. ## The portfolio piece Piece 1 has four parts: 1. **Technical artifact.** Reproduce one published finding (from "Machine Bias" or a local data story) and extend it with city data that you scraped or downloaded responsibly, with a documented provenance file. 2. **Public text.** A 750-word op-ed for a named local outlet, plus a one-paragraph pitch. 3. **Installed-base note** (about 300 words): the pressure you met, the installed-base elements your artifact built and those still missing, with provenance, license, and an AI-use disclosure. 4. **Graduate methods memo** (INFO 5871 only; 750 to 1,000 words): the piece situated in at least five scholarly sources, defending one methodological choice. Each chapter in this part has an exercise that builds a component of the piece, and the last exercise of @sec-ch-07-op-eds is the piece itself. ## Roadmap - **@sec-ch-04-enclosure** (week 3) names the pressure. It traces platform enclosure, introduces soft enclosure at city hall, and builds a responsible scraper that writes a provenance log. Its Piece 1 exercise is an access memo and the start of your provenance file. - **@sec-ch-05-journalism** (week 3) presents the lineage. You reproduce "Machine Bias," compare your numbers to the published ones, read the Flores, Bechtel, and Lowenkamp rejoinder, and set the Pegasus Project beside it. Its Piece 1 exercise is a reproduction plan. - **@sec-ch-06-openness** (week 4) builds the value. You ingest, clean, document, and publish Boulder's election contributions data, and trace a decision through city meeting records to see where posting stops and publishing begins. Its Piece 1 exercise packages your city data as a publication. - **@sec-ch-07-op-eds** (week 5) is the genre. It covers the local outlet ecosystem, who gets published, the anatomy of a 750-word argument, and pitching, with a worked example on Boulder data. It ends with Piece 1. One piece of logistics links this module to the next. The week 4 lab ends with a preview of Part III: you browse the **Public Records Archive** in Boulder County's Open Records Center, where the county posts Colorado Open Records Act requests that others have already filed and the county has already answered, and you shortlist released requests to analyze. You do not file new requests for Part III. @sec-ch-09-law covers the legal details. Part III moves up the ladder to the county, with a new lineage, pressure, and case.