In 2022, a thirteen-year-old Guatemalan boy, Ever Reyes Mejia, was released by the Office of Refugee Resettlement (ORR) to a sponsor in Alabama. He ended up working twelve-hour overnight shifts at a Hyundai parts supplier, operating industrial equipment without adult supervision.
He was not alone. Investigators and journalists later documented migrant children working night shifts in auto plants, slaughterhouses, and construction sites across a dozen states — children who had been processed, released, and forgotten by the federal apparatus that accepted legal custodial responsibility for them.
This failure is the record of specific organizational decisions made by specific people in specific agencies that received specific federal dollars to perform a specific task. In this case, the task was caring for children.
In December 2025, three academic investigators—I was one of them—submitted a peer-reviewed manuscript to an academic journal that specializes in issues involving children and youth.
The paper examined U.S. policy and service responses to unaccompanied migrant children — a population for which the federal government had accepted custodial responsibility, disbursed approximately $17 billion in taxpayer funds, and subsequently lost track of an estimated 330,000 individuals, most of whom are believed to have been trafficked for labor or sex.
In May 2026, the journal rejected the manuscript.
The reviewers’ objections were instructive — not as a legitimate methodological critique, but as a record of how academic gatekeeping protects preferred narratives from investigative accountability. If read carefully, those objections reveal more about the reviewers’ ideological commitments than about the quality of the research.
They deserve a public response.
Reviewer One raised two concerns grounded in sound methodology.
The manuscript lacked a formally declared methodology statement — a standard journal requirement — and one significant statistical claim about trafficking rates was cited from a secondary source rather than traced to its primary source.
Both criticisms can be fixed. Neither goes to the substance of the investigation.
Reviewer Two objected to the term “surge” as “militarized language that is inappropriate when referring to children.” He suggested that the authors stop describing children who missed immigration court hearings as having “evaded” proceedings because the term “invokes criminality and fugitivity.” And he recommended that the authors “move away from reductive push-pull explanations” and toward frameworks that treat migration as “a form of care and protection rather than evidence of parental failure.”
These criticisms have nothing to do with methodology or substance.
They are ideological directives. The reviewer is not asking the authors to produce better research. The reviewer was asking us to adopt a specific political outlook — one that treats any scrutiny of the system responsible for 330,000 missing children as inherently suspect.
Most revealing was Reviewer Two’s central objection: that the paper’s argument would be stronger if it “more directly addressed the contradiction at the center of the system that ORR is positioned as a child welfare actor while operating within a broader immigration enforcement regime.”
This reframing neither sharpens nor clarifies the analysis. It defangs the argument. It transforms a direct accountability investigation — who is responsible for these children and where did the money go — into an abstract structural critique in which no specific actor bears responsibility.
Systems fail. Bureaucratic tensions produce suboptimal outcomes. Nobody goes to prison.
Reviewer One’s most substantive methodological demand was that the authors follow the Arksey and O’Malley scoping review framework — a structured academic protocol for systematizing literature synthesis. The demand is revealing because of what it assumes: that the research subjects cooperated, that the data were available, and that the investigators’ primary challenge was organizing what they found rather than fighting to find anything at all.
The federal agencies stonewalled. The NGOs refused to cooperate. The organizations receiving federal reimbursements averaging $42,500 per child in FY2022 and $41,600 per child in FY2023 — figures derived directly from the totals in the HHS Administration for Children and Families’ own documentation, divided against the children actually served — declined to provide information to us, a group of independent investigators.
Organizations that processed hundreds of thousands of children, three-quarters of whom subsequently vanished, declined to answer questions about what happened to those children.
There is no Arksey and O’Malley framework for that situation. Academic methodology is designed for cooperative institutional environments, whereas investigative methodology is designed for hostile ones.
The peer review process has no rubric for evaluating research conducted amid active institutional resistance — and so it evaluates that research as if the resistance didn’t occur, finding it methodologically deficient for failing to do what the subjects of the investigation prevented it from doing.
The non-cooperation was, in itself, a finding.
It was arguably the most important finding. An organization that takes billions in federal funds to protect children and then refuses to answer questions about what happened to those children has told you something significant through that refusal. The peer review process treated this as a methodological weakness. In fact, it is evidence.
Reviewer One criticized the manuscript for treating “peer-reviewed studies, government reports, congressional hearing transcripts, and investigative journalism with roughly equal evidentiary weight.” This criticism warrants examination because it rests on the assumption that peer-reviewed literature represents the highest evidentiary standard — an assumption that is itself a significant analytical error.
Consider what peer-reviewed research on NGO effectiveness for unaccompanied migrant children actually entails. The NGOs under study control researchers’ access. They determine which staff are interviewed, which facilities are visited, which documents are provided, and which outcomes are measured.
Researchers receive what the organization wants them to see, process the information using an academic methodology, and produce a paper that appears to offer independent verification while resting on a foundation of managed disclosure.
A congressional hearing transcript, by contrast, is a sworn record produced under oath during a sometimes hostile examination. An Inspector General report is produced by an agency that is legally independent and has the power to issue subpoenas. Investigative journalism produced without NGO cooperation relies on what the public record shows when no one manages access. These are not methodologically inferior to peer-reviewed studies that relied on the cooperation of the organizations under scrutiny.
In many cases, they are superior.
The evidentiary hierarchy that academic peer review enforces — privileging credentialed institutional sources over independent investigation — is precisely the hierarchy that shields powerful institutions from accountability.
It is not a neutral methodological standard. It is a bias built into the review process that systematically advantages the subjects of investigation over the investigators.
The peer review community in the social sciences lacks political diversity. Published research from leading American universities found that faculty registered as Democrats outnumbered Republicans in psychology departments by a 17.4-to-1 ratio — and in history departments by a 33.5-to-1 ratio.
In social work, the journal’s disciplinary home of those who rejected this manuscript, the imbalance is almost certainly greater.
Jonathan Haidt and Philip Tetlock formally documented the political homogeneity problem in peer review and found that political commitments precede and shape methodological evaluations — reviewers evaluate manuscripts differently depending on whether findings align with their political worldview.
Conservative social psychologists reported hostile work climates for their political beliefs. A significant percentage of liberal academics explicitly endorsed discriminating against conservative colleagues in hiring and publishing decisions. Reviewer Two’s politically infused instructions to the authors of this manuscript are consistent with that literature.
Reviewer Two recommended that the authors cite Galli, Heidbrink, Ruehs-Navarro, and other scholars in the literature on unaccompanied children, suggesting that this engagement would “free the authors to focus on their argument.” The recommendation is worth examining for what it is: a directive to frame the research from the perspective of the academic community whose theoretical frameworks and political commitments are known, accepted, and enforced through the review process. In other words, we must adopt the accepted narrative if we want to be published.
This is not what peer review is supposed to be; here, it functions as gatekeeping for a storyline. An investigator who declines to frame findings through a particular academic community’s preferred theoretical lens does not demonstrate a weakness in methodology.
He is declining to subordinate his findings to a community whose position is fixed and settled and dare not be challenged.
The history of scholarship is full of foundational studies that passed peer review and shaped decades of subsequent research — only to collapse under independent scrutiny.
Alfred Kinsey’s research on sexual behavior, for example, cited as the foundation for thousands of subsequent studies, was built on a methodology so structurally corrupt that its findings have never survived independent replication: convenience and criminal populations presented as representative samples, statistical procedures that concealed rather than revealed their assumptions, and conclusions that served an ideological program shared by the reviewer community.
John Money’s gender identity research, which became the clinical foundation for decades of pediatric treatment protocols, was built on a Kinsey foundation and on a case study — the Reimer case — that Money fraudulently reported, suppressing outcomes that directly contradicted his published conclusions.
For both Kinsey and Money, peer review failed because reviewers shared the researchers’ ideological commitments and evaluated the methodology through a lens that rendered the foundational errors invisible.
When the foundation is corrupt, every study built on it carries that corruption. Citation volume is not corroboration. It is propagation. A literature built on managed NGO cooperation, cooperative institutional access, and theoretical frameworks that treat accountability scrutiny as inherently suspect does not become more reliable by accumulating citations.
It becomes more dangerous.
Researchers have documented that replication success rates in psychology and related social science disciplines often fall below 50%—meaning published peer-reviewed findings fail to hold up under independent testing more often than not.
The publish-or-perish incentive structure rewards novel, statistically significant findings over rigorous, replicable ones. The result is a literature that appears authoritative yet, in aggregate, is substantially unreliable. Directing investigators to anchor their findings in that literature is not a methodological improvement. It is an invitation to inherit the errors.
More than 400,000 unaccompanied children crossed the southern border during the Biden administration. The federal government accepted custodial responsibility for each of them. According to the Department of Health and Human Services Administration for Children and Families’ own Justification of Estimates for the Appropriations Committees — a budget document submitted to Congress — ORR spent $5.4 billion in FY2022 and approximately $4.9 billion in FY2023 to manage the pipeline of unaccompanied children.
Divided among the children actually served, that amounts to $42,500 per child in FY2022 and $41,600 in FY2023. Congress itself, in the findings section of H.R. 2644 (118th Congress), placed the average daily cost of ORR custody at $500 per child.
The organizations receiving those funds were not charities, as they presented themselves, that were also receiving government support. They were government contractors sustained almost entirely by federal payments. Southwest Key Programs — the single largest operator, sheltering 43,054 children in FY2023 — reported $919.5 million in government grants against total revenue of $921.5 million: 99.8 percent federal dependence.
Family Endeavors reported $562.8 million in contributions, representing 98.8 percent of total revenue of $569.6 million.
Global Refuge, formerly Lutheran Immigration and Refugee Service, reported that contributions accounted for 98.7 percent of total revenue.
The U.S. Committee for Refugees and Immigrants reported a 99 percent rate. These figures come from the organizations’ own IRS Form 990 filings, available in the public record.
The OpenTheBooks audit of federal disbursement data found that between 2012 and 2022, ORR disbursed $12.8 billion in grants to this contractor network, with Southwest Key Programs receiving $3.7 billion and BCFS Health and Human Services receiving $3.1 billion. In FY2023 alone, Southwest Key’s federally funded unaccompanied children program accounted for $911 million in federal expenditures — as confirmed by the organization’s audited financial statements filed with the Federal Audit Clearinghouse.
Southwest Key alone has received nearly $6 billion from the federal government since 2007 — a figure cited in the Department of Justice’s civil lawsuit against the organization, filed in July 2024, alleging sexual harassment of minors in its care.
The federal government paid approximately $43,000 per child. Three out of every four were lost.
When the Trump administration took office in January 2025, it found that ORR had accumulated a backlog of more than 65,000 unaddressed Notifications of Concern — internal reports flagging individual children with allegations ranging from administrative issues to active trafficking and sexual and physical abuse.
Sixty-five thousand reports of children potentially in danger. Left sitting. Unaddressed.
The phones rang, but nobody answered, and the children stayed wherever they were.
That is not a bureaucratic tension between child welfare and immigration enforcement mandates. That is an agency that received specific reports of children in danger and chose not to act. The academic literature on contradictions in federal child welfare administration offers those children nothing useful about that decision.
No peer-reviewed methodology is adequate to the moral weight of that finding. No citation to Galli or Heidbrink changes what happened to 330,000 children. No reframing of ORR’s structural contradictions accounts for the $17 billion disbursed to organizations that lost three-quarters of their protected population.
The academic literature on trauma-informed care is irrelevant to children who never received any care before disappearing into labor trafficking networks and commercial sexual exploitation — a pattern the Senate Permanent Subcommittee on Investigations subsequently confirmed.
The reviewers acted as if they were evaluating a parking ticket rather than the mass-casualty event the investigators documented.
Academic peer review in the social sciences is designed to evaluate incremental contributions to the existing literature within established theoretical frameworks. It was not designed to evaluate — and actively suppresses — investigations that challenge an institution’s preferred narratives that those frameworks were built to protect.
What happened to this manuscript is not an isolated editorial decision. It is a specimen of a systematic process in which accountability investigations are converted into policy reform discussions, abstract critiques shield specific actors, unfavorable evidence is reframed as a methodological weakness, and investigators who decline to adopt the preferred political perspective are rejected on procedural grounds.
The system works as follows. An academic investigator produces research based on public records, government oversight reports, congressional testimony, and documented refusals to cooperate by the agencies and NGOs under investigation. A peer reviewer evaluates the manuscript and finds it wanting: insufficient participation by NGOs and agencies, inappropriate weight given to non-academic sources, and language that implies accountability rather than structural critique.
The manuscript is rejected. The investigative findings do not enter the academic record. The NGOs continue to receive federal funding. The children remain missing.
This is an emergent property of a peer review system staffed by researchers whose careers depend on institutional access, whose theoretical commitments drive their evaluations, and whose professional incentives run counter to the adversarial, independent investigation that discovering truth and providing accountability requires.
The investigation into what happened to 330,000 unaccompanied migrant children does not need peer-review validation. It needs an HHS Inspector General complaint documenting the financial irregularities and the systematic refusal to cooperate with independent investigators. It needs congressional oversight with subpoena power directed at every NGO that received $43,000 per child and cannot account for the children.
It needs Department of Labor enforcement records cross-referenced with the geographic distribution of unaccompanied minor placements to identify the labor trafficking system that received them. It needs named executives, named organizations, named federal officials, and documented evidence of who knew what and when.
The peer reviewers who rejected this manuscript on methodological grounds performed exactly the function they were intended to perform: they kept an accountability investigation out of the academic record, protected the institutional network under scrutiny from independent examination, and recast a documented pattern of organizational failure and stonewalling as a research design problem.
Three-quarters of the children are missing and believed to have been trafficked, despite billions spent to care for them.
The academic literature on this may be methodologically sound and appropriately cited, but it’s not useful to those children. And it does not make the system more effective or prevent it from squandering many billions of dollars annually.
Stop accepting this. Take action and do something about it.
Contact the HHS Office of Inspector General to demand a formal investigation into ORR’s handling of the 65,000 unaddressed Notifications of Concern and into every NGO that received federal funds. File a complaint online at oig.hhs.gov/fraud/report-fraud/ or call 1-800-HHS-TIPS.
Contact your member of Congress and U.S. Senators to demand oversight hearings targeting the NGO executives who took $43,000 per child from federal sources and cannot account for what happened to three-quarters of those children.
Share this article. The children cannot speak for themselves, and the academic literature offers no help. Someone needs to be their voice.
Why couldn’t you be the one to speak up?
