Case Study 13.2 — Dueling Entomologists: People v. Westerfield and the Limits of the Insect Clock

Sourcing and tone. This case study draws on the public record of a U.S. trial (San Diego County, California, 2002) and is used to teach a single methodological point: that two qualified forensic entomologists, examining the same evidence in good faith, can reach different time-since-death estimates — and why that is a feature of the method's honest uncertainty, not proof that one of them is a fraud. We treat a child's death soberly and stay within documented, public facts. The defendant was convicted; the entomological dispute is presented to illustrate the science's error bars, not to relitigate the verdict.

Background

In February 2002, seven-year-old Danielle van Dam disappeared from her home in San Diego, California. Weeks later her body was recovered in a remote location off a rural road. A neighbor, David Westerfield, was charged with and ultimately convicted of her kidnapping and murder.

The case is studied in forensic-science courses for one reason above all: the forensic entomology became a genuine battleground. The defense and the prosecution each called qualified entomologists, and their estimates of the postmortem interval — specifically, how long the body had been at the recovery site, as indexed by insect colonization — did not agree. The disagreement mattered enormously to the legal theory of the case, because the timing interacted with the window during which the defendant had access to the victim before he was under law-enforcement attention.

The forensic dispute

The competing entomological estimates turned on exactly the variables this chapter identifies as the method's sources of uncertainty (§13.3–13.4). Without adjudicating who was right, the structure of the disagreement is the lesson:

• Temperature reconstruction. Insect development depends on accumulated temperature (§13.3, accumulated degree days). The recovery site was outdoors and remote, and reconstructing the temperatures the colonizing insects actually experienced over the unobserved interval — from regional weather data, corrected to a microclimate — involves assumptions. Reasonable experts can differ on which temperature history best models the scene, and small differences in assumed temperature produce meaningfully different interval estimates (recall §13.3's worked example, where the same required ADD took 10 days at one temperature and 22 at a cooler one).

• When colonization began. The estimate is a minimum time since colonization (§13.1), and whether colonization was prompt or delayed — by weather, by the body's condition, by access to the recovery site — bears directly on how the insect age maps onto the time the body had been present. Experts weighing the likelihood and length of any colonization delay differently will reach different windows.

• Species, development data, and specimen interpretation. Which species were present, what published developmental data apply, and how to read the particular specimens recovered are all points where expert judgment enters. The defense and prosecution experts brought different readings to these questions.

The result was two scientifically defensible estimates that pointed toward somewhat different timelines — one more consistent with the defense's theory of the relevant window, the other with the prosecution's.

What the dispute did — and didn't — establish

It is tempting to conclude from "the experts disagreed" that entomology is junk science. That conclusion is wrong, and the chapter explains why. The disagreement here was not between a valid method and an invalid one; it was within a method whose honest output is a range built on reconstructed inputs. The experts disagreed about the inputs (temperature history, colonization delay, specimen interpretation) — and because the method propagates those inputs into the answer, reasonable disagreement about the inputs yields different ranges. This is precisely the uncertainty §13.3 warns is built into an accumulated-degree-day estimate, not a sign that the underlying physiology is unsound.

Contrast this with a single-source DNA match (Chapters 7–9), where the central comparison yields a quantity with a tight, well-characterized error structure. Two competent DNA analysts examining a clean single-source profile will not produce wildly different random match probabilities. Entomology's inputs are noisier and more reconstructed, so its honest error bars are wider — and when two careful experts each report their range candidly, those ranges can diverge. The Westerfield entomology is, in this sense, the validity spectrum made visible: a real method, honestly bounded, sitting well below DNA precisely because its inputs cannot be pinned down as tightly.

The jury, for its part, convicted — relying on the totality of the evidence (which included other forensic links not at issue here), and evidently not finding the entomological dispute sufficient to create reasonable doubt about the case as a whole. That outcome is consistent with the chapter's framing: entomology is rarely the load-bearing wall of a case; it is one thread, and a thread with visible error bars, to be weighed alongside the rest.

The lesson

Three lessons, all central to this chapter:

1. Honest uncertainty is not fraud. When two qualified entomologists disagree about a time-since-death window, the most likely explanation is not that one is a charlatan but that the method's inputs are genuinely uncertain and the experts weighed them differently. A method that reports wide, input-dependent ranges is being honest about its limits — exactly what §13.3–13.4 demand. The reader should distrust a single confident decimal-point answer (§13.3 Junk-Science Alert) far more than two candid, overlapping-or-diverging ranges.

2. The error lives in the inputs, so attack the inputs. A competent cross-examination (Chapter 30) of entomological testimony does not sneer at "bug science"; it probes the temperature reconstruction, the assumed colonization delay, the species identification, and the developmental data — because that is where the estimate is actually made and where reasonable people can differ. Understanding the method tells you exactly where to push.

3. Entomology constrains; it rarely decides. As in the cold case (where the fire makes entomology corroborating rather than leading, §13.4) and as in the Casey Anthony case (Case Study 13.1, where old skeletal remains pushed the insects to the edge of what they can say), entomology here was one constrained thread. The case turned on the totality. This is exclusion over proof (Theme 1) and the validity spectrum (Theme 2) in a single example: a real method, useful, honestly bounded, and not to be asked to carry more certainty than its reconstructed inputs can bear.

Discussion questions

1. The two experts disagreed largely about temperature reconstruction and whether colonization was delayed. Using §13.3, explain mechanically how a difference in assumed average temperature translates into a difference in the estimated minimum PMI. Why does even a few degrees matter?

2. A commentator says, "Since two entomologists got different answers, entomology is junk science like bite marks." Refute this using the distinction between a method with an unsound foundation (e.g., bite marks, Chapter 16) and a method with sound foundations but noisy, reconstructed inputs. Where does each sit on the validity spectrum, and why?

3. You are cross-examining an entomologist (Chapter 30). List four specific input assumptions you would probe, and for each, state what concession would widen the estimate's uncertainty.

4. Compare the role of entomology in (a) this case, (b) Casey Anthony (Case Study 13.1), and (c) the Mill Creek cold case. In which is entomology closest to "leading," and in which is it most clearly "corroborating"? What feature of each case sets that role?

5. Why would it be less defensible for two competent analysts to reach sharply different conclusions about a clean single-source DNA profile than about an entomological PMI? Use this contrast to explain what "wider error bars" really means about a method's inputs.

6. Ethics tie-in (Chapter 31, previewed). Suppose each side's entomologist was told, before analysis, which timeline their retaining attorney needed. Explain how contextual bias could push two honest experts farther apart, and what safeguard (blind analysis, sequential unmasking) would reduce that effect.