The Cold Case Equation: Why 99 in 100 Investigations Never End in Conviction

The Cold Case Equation: Why 99 in 100 Investigations Never End in Conviction

There is a number that sits quietly at the center of every cold case discussion, rarely stated plainly: one percent. According to statistical estimates of cold case investigation outcomes, only one in every 100 reopened investigations results in a conviction. One in 20 produces even an arrest. These are not the metrics of a functioning system. They are the metrics of a system operating at the outer edges of its capacity — a system where the evidence, the funding, and the institutional memory have all begun to decay at roughly the same rate as the cases themselves.

Understanding why cold cases fail — and why, occasionally, they succeed — requires moving past the narrative drama that surrounds them and into the structural, forensic, and psychological mechanics that determine their fate.

---

What Makes a Case Go Cold

A cold case is formally defined as a criminal investigation in which all initial leads have been exhausted without resolution. In practice, the transition from active investigation to cold case is rarely a formal declaration. It is a gradual depletion — of witnesses willing to talk, of physical evidence, of departmental bandwidth, and eventually of investigator attention.

Between 1997 and 2017, the accumulated total of cold cases in the United States surpassed 100,000. That figure reflects not a single catastrophic failure but a slow, compounding arithmetic: each year, the number of new uncleared homicides outpaces the number of cold cases that get resolved. The backlog grows not because cases are forgotten, but because the infrastructure to address them was never adequately built.

The data on institutional readiness is stark. Only 20 percent of law enforcement agencies have a formal protocol for initiating cold case investigations. Only 7 percent maintain a dedicated cold case unit. The remaining agencies operate without structured frameworks, meaning that whether a cold case gets revisited often depends on factors as arbitrary as personnel turnover, media attention, or a family member filing a persistent inquiry.

---

The Funding Architecture of Failure

Beyond protocol, there is the question of money — and here the structure becomes even more precarious. Only 20 percent of cold case work is funded through standard budget line items. The majority relies on grants or supplemental funding streams that are, by definition, temporary and competitive.

This creates a fundamental operational problem. Investigations that depend on grant cycles cannot sustain the long-arc work that cold cases require. When a grant expires, so does the investigator's time, the laboratory access, and often the momentum of a case that may have taken years to build. The institutional knowledge accumulated during a funded period does not automatically transfer. Files get archived. Leads go unfollowed.

The financial instability also shapes which cases get selected for review. Research indicates that cases reopened solely due to family requests or routine administrative review have significantly lower odds of clearance compared to cases reopened because of new, actionable leads — a DNA match, a deathbed confession, a newly surfaced witness. This creates an uncomfortable triage logic: the cases most likely to generate political pressure for review are not necessarily the cases most likely to be solved.

---

The Forensic Genealogy Inflection Point

If there is a genuine inflection point in cold case methodology, it arrived with the widespread adoption of forensic genealogy and advanced DNA profiling. These tools have fundamentally altered the solvability calculus for cases where biological evidence was collected and preserved — even imperfectly — decades ago.

Forensic genealogy works by constructing familial genetic profiles from trace DNA evidence, then cross-referencing those profiles against consumer genealogy databases to identify possible relatives of an unknown suspect. From those relatives, investigators build family trees forward and backward through generations until they can triangulate a plausible suspect. The process is labor-intensive, but it has produced results that would have been considered impossible under prior investigative frameworks.

The 1979 murder of 17-year-old Esther Gonzalez in Riverside County, California, illustrates both the power and the particular character of these resolutions. In 2024 — 45 years after the crime — forensic genealogy identified the killer as Lewis Randy Williamson. Williamson had died in 2014, ten years before his identification. The case was closed without an arrest, without a trial, and without the possibility of prosecution. For the Gonzalez family, the identification represented a form of resolution. For the justice system's statistical record, it registered as a solved case — though the legal process never ran its course.

Similar dynamics appeared in the resolution of the 1974 murder of hitchhiker Mary K. Schlais, solved decades later through modern DNA testing, and in the Oklahoma City Police Department's closure of a long-unsolved homicide by matching DNA evidence to a suspect named Charles Droke. Each case demonstrates that forensic technology can recover identities across time. What it cannot always recover is accountability in any conventional legal sense.

---

The Psychology of Degraded Evidence

Farther from the laboratory, cold case investigators face a set of challenges that no technological advance has fully addressed: the psychology of human memory and the compounding unreliability of witness testimony across decades.

Research on witness recall consistently demonstrates that memory is reconstructive rather than reproductive — it does not play back like a recording but rebuilds itself each time it is accessed, incorporating subsequent experiences, media exposure, and suggestion. In cold cases, witnesses who were interviewed once in 1979 and again in 2019 are not simply older; their memories have been reorganized, overwritten, and contaminated by 40 years of intervening cognition. The reliability of that testimony in an adversarial legal proceeding is, at best, contested.

Compounding this is the phenomenon investigators and researchers call tunnel vision — the cognitive bias through which early investigative assumptions narrow the frame of inquiry in ways that persist long after those assumptions should have been revisited. In cold cases, tunnel vision from the original investigation can be embedded in the case file itself, shaping how a new investigator reads the evidence without realizing they are inheriting a conclusion rather than a record.

Addressing these psychological factors has become an increasingly formalized area of cold case methodology, with some units now employing structured case review protocols designed to force investigators to treat the file as if no prior conclusions had been drawn.

---

What the Data Actually Tells Us

The evidence, taken together, describes a system with a narrow and specific capacity for success. Cold cases are most likely to move toward resolution when three conditions align: preserved biological evidence exists and has been tested with current technology; new actionable intelligence — forensic, testimonial, or informational — enters the investigation; and institutional resources are stable enough to sustain the work to its conclusion.

When those conditions are not met, the statistical baseline reasserts itself: one arrest in 20, one conviction in 100. The accumulation of more than 100,000 unresolved cases is not the product of investigator failure alone. It is the product of a funding architecture that treats cold case work as supplemental, an institutional landscape where 93 percent of agencies lack dedicated units, and an evidentiary reality in which time is almost always working against resolution.

Forensic genealogy has opened a door. But the data suggests that door opens onto a corridor that most cases will never reach — not because the science is insufficient, but because the systems surrounding the science were never designed to carry this weight.