Caleb Johnston

Washington, D.C.

Institutions are experienced at the front desk.

I am finishing an M.A. in American Government at Georgetown this August, and I work as a Reentry Assistant at the university’s Prisons and Justice Initiative. Most of what I understand about how government actually reaches people, I have learned from the side of the desk where they come to ask it for something.

My work keeps landing in the same place: the distance between what an institution promises on paper and what a person meets when they walk in. I am looking for research, advocacy, campaign, or nonprofit work on democratic accountability and criminal-legal reform. I use AI tooling as a standing part of my practice to move faster through research and outreach.

Portrait of Caleb Johnston.
M.A., American Government, Georgetown University, conferred August 2026. B.A., Economics and Government, the University of Texas at Austin.
01Reentry

The front desk turned out to be a vantage point.

At Georgetown’s Prisons and Justice Initiative I am often the first person someone meets — students, staff, community members, partner organizations, and people calling to find out whether we can help at all. I support PJI’s six programs across prison education, reentry services, and wrongful-conviction work — PIVOT, the Paralegal Program, the Scholars Program, the Bachelor of Liberal Arts program, Making an Exoneree, and the Family First Initiative. In practice that means keeping participant records and program documentation in order, running outreach, fulfilling research requests for students, and sitting with people as they work through housing, identity documents, phone service, storage, and online portals that were plainly not designed with them in mind.

I support and observe intake interviews and reentry assessments. I do not conduct them independently. Separately, I research and build profiles of reentry service providers across D.C. and Maryland — housing, employment, and other resources — so that the organization’s referrals rest on something current rather than something remembered.

The part that changed how I think is smaller than any of that. Reentry is rarely settled in a hearing. It is settled in whether the person who picks up the phone knows which form you need, and whether they treat the asking as a nuisance or as the job. Georgetown wrote about that shift in a profile of my work at PJI.

02Capstone

So I spent seven months studying the organization I work for.

My M.A. capstone asks whether the frontline reentry practices of a nongovernmental organization interrupt, soften, or reproduce punitive governance — and what the answer reveals about reentry as a governance problem. In the District, custody and supervision belong to federal agencies and identity documents to District agencies, while the connective work of interpreting rules and moving people between institutions falls to intermediaries the formal system neither commands nor replaces.

It is a single-case qualitative study conducted from inside the organization between January and July 2026: nine recorded interviews with staff, faculty, and program participants; fifteen weekly field journals; participant observation; and a documentary record assembled to check the organization’s account of itself against sources it does not control.

The finding is that all three verbs hold at once, and that their distribution tracks control. The paper develops a framework of three regimes of discretion for analyzing nongovernmental intermediaries as governing actors, argues that their most consequential outcomes are structurally unobservable, and sets out implementation recommendations. Its conclusion locates the constraint in the design of the reentry field rather than in any single organization’s performance.

Not posted here

“Our Sacred Space”: Governance of Reentry at Georgetown’s Prisons and Justice Initiative

M.A. capstone · Georgetown University · 59 pages · July 2026

The paper draws on confidential interview material, so I am not publishing it. I am glad to send it to anyone considering me for a role.

Full paper available on request — email me

03Writing

Work you can open and read.

Three pieces are posted here in full. Each one is my own writing, and each says plainly what it can and cannot support.

Policy memo

A Two-Tier Strategy for Public Safety around Banner Lane and Sursum Corda

Written for the D.C. Council Committee on the Judiciary and Public Safety · 11 pages · 2025

Banner Lane and the surrounding Sursum Corda neighborhood in Ward 6 have absorbed decades of violence and disinvestment, and a 2015 settlement that added lighting and cameras did not resolve it. The memo recommends pairing two tiers. The first is Crime Prevention Through Environmental Design — sightlines, lighting, lot maintenance, and remediation of vacant land — which is cheap and fast. The second is social investment: youth employment, addiction treatment and drug courts, and community violence intervention. It weighs the alternatives against a do-nothing baseline, addresses the displacement objection directly, and closes with phased implementation, funding, and accountability measures.

I started reading this literature after a mugging in the neighborhood the memo covers.

Read the memo PDF · 11 pp

Quantitative research

Recreational Marijuana Legalization and Weekly Hours Worked: A State-Level Econometric Analysis

Undergraduate capstone · The University of Texas at Austin · 20 pages · 2024

I built an OLS model in Stata on 2021 cross-sectional data covering all fifty states, drawing labor, wage, health, education, demographic, and cannabis-policy measures from the Bureau of Labor Statistics, the CDC, FRED, and the Census Bureau. I ran diagnostics for multicollinearity and heteroskedasticity and reported a logged alternative specification alongside the main model.

It finds a modest negative association between legalization and weekly hours worked, significant only at the ten percent level. With fifty observations and a single cross-section of aggregate data, the design cannot support a causal claim, and I would not defend it as more than exploratory. I am posting it because the limitations are the honest part of the exercise.

Read the paper PDF · 20 pp

Research assistantship

Winners and Losers: Electoral Rhetoric and Democratic Norms

Research assistant to Prof. Bethany Albertson · UT Austin, Department of Government · Summer 2024

This was faculty-led research, not my own scholarship. I built candidate-level records for presidential, Senate, House, and gubernatorial races from 2000 to 2024 — party, incumbency, vote share, margin, winner and loser status, seat flips — and located and verified the victory and concession statements attached to them. I then applied coding indicators for democratic-norm language: unity appeals, references to democracy, voter-fraud claims, recount and litigation context, and whether candidates spoke to each other.

The document below is my own sanitized summary of what I contributed. It deliberately excludes the underlying dataset and response material, which belong to the project rather than to me.

Read the brief PDF · 2 pp

04Field work

The unglamorous half, which is most of it.

Carmichael for Ward 5 Council

Deputy campaign manager · February – June 2026

I ran field and operations for Bernita Carmichael’s campaign for the Ward 5 seat on the D.C. Council. The campaign lost. I coordinated a field team of ten volunteers across canvassing, phone banking, and community events; supported ballot-access verification by helping with technical review and compliance documentation against Board of Elections thresholds; organized two campaign events and represented the candidate at roughly ten more, including a D.C. teacher summit; and handled the logistics, including late-night runs to get materials where they needed to be by morning. I drafted volunteer briefings and outreach materials.

Teaching

Graduate teaching assistant, Department of Government, Georgetown · August 2025 – May 2026

I led weekly discussion sections for about forty undergraduates across two semesters — one course in the fall, then two sections of a different course in the spring — in The U.S. Political System and Introduction to International Relations. Teaching is where I learned to explain institutions to people meeting them for the first time.

Pappasito’s Cantina

Waiter, headwaiter, bartender, to-go · 2021 – 2025

I worked at Pappasito’s for four years while I was in college, at high-volume locations in Houston and Austin, and I was still doing it the year I started applying to graduate school.

05Contact

Get in touch.

I am actively looking for full-time work beginning after my degree is conferred in August 2026, in campaigns, advocacy organizations, think tanks, or nonprofits.

Email

cj691 [at] georgetown [dot] edu

Elsewhere

LinkedIn

Focus

Democratic accountability and criminal-legal reform.