top of page

When a Record Closes One Door but Leaves Another Open

Oct 2
2 min read

A felony conviction can coexist with access to the presidency. For ordinary job applicants, a criminal record can close doors—and racial discrimination can make that barrier more severe.

In Devah Pager's Milwaukee hiring experiment, published in 2003, white applicants presenting a criminal record received callbacks at 17%, compared with 5% for Black applicants presenting one. Black applicants without a record received callbacks at 14%; white applicants without a record, at 34%. These are study-specific callback rates, not current national employment rates, job offers, or measurements of job quality.

New York court records document Donald Trump's May 30, 2024 conviction on 34 counts of first-degree falsifying business records. Reuters reported an unconditional discharge on January 10, 2025, with no jail time or fine, and an unsuccessful attempt to move the case to federal court on August 28, 2026. The current White House identifies him as president.

Presidential elections and ordinary hiring operate under different rules. Their comparison cannot by itself isolate race from wealth, political connections, voter preferences, and other differences. The matched hiring experiment, however, directly demonstrates unequal employer treatment.

Our conclusion: Black applicants can face compounded barriers from racial discrimination and criminal-record stigma. Claims that no Black person with a felony can obtain a good job, or that every white person receives favorable treatment, exceed the evidence.

For teaching: What does a callback measure? Why is a matched experiment stronger evidence than a comparison of unrelated individuals? What new data would establish whether a hiring reform improves outcomes?

Sources: Devah Pager, The Mark of a Criminal Record (2003): https://doi.org/10.1086/374403

Callback figures and research context: https://www.hks.harvard.edu/fierce-commitment-and-inestimable-warmth-harvard-kennedy-school-mourns-loss-devah-pager

Court record: https://www.nycourts.gov/LegacyPDFs/press/PDFs/CPL330.30Dec.pdf

Reuters, August 28, 2026: https://www.reuters.com/world/trump-loses-bid-move-already-decided-criminal-hush-money-case-us-court-2026-08-28/

Presidential identification: https://www.whitehouse.gov/administration/donald-j-trump/

Colorized Community Friends educational case-study supplement. Prepared October 2, 2026.

 
 
 

Recent Posts

See All
Racism in the record: how law becomes a system

Racism becomes institutional when discriminatory practices are authorized, enforced, or protected by organizations. A historical study should identify those mechanisms and the people with decision aut

 
 
 
How to reach a conclusion others can check

A credible conclusion shows its path from evidence to claim. Our educational project uses four labels: verified document fact, attributed allegation, interpretation, and unresolved question. Start wit

 
 
 

Comments


bottom of page
Guardian Enterprise Studio OS
Render orchestration · GPU routing · SLA · cost control
Queue
P1 / P2 / P3
Routing
CPU + GPU
Retries
3 max
Leases
Heartbeat
Worker orchestrationREADY
GPU-aware schedulingREADY
Artifact deliveryREADY
Prometheus / GrafanaREADY
Finance exportREADY
P1GPU / HERO / FX
P2Standard production
P3Background / batch
Live job data is sourced from the Enterprise Studio collections and the render-worker API when that execution endpoint is connected.
CPU rate$0.40/hr
GPU rate$2.50/hr
Cost attributionENABLED
RenderSeconds and worker type are retained for job-level cost attribution.
Queue latencyMONITORED
Failure rateMONITORED
Worker heartbeatMONITORED
SLA alerts are designed to feed Slack/email escalation from the existing render-farm control plane.