Skip to main content

Making a Murderer

Police Work and Forensic Evidence in Making a Murderer

Compiled by Michael Creim, Napakamol Havanond, Angela Kim, and Rachel Rasp

I.   Introduction

One simple way to avoid any allegations that Manitowoc County Sheriff’s officers framed Steven Avery was to keep them out of the Halbach investigation. Why would Manitowoc County Sheriff’s officers have incentive to frame Steven Avery? Because there was a pending $36 million wrongful conviction lawsuit Avery had filed against the county over the 1985 mistaken rape conviction. Although authorities announced that Manitowoc County would not be involved in the Halbach investigation, Detective Lenk and Colborn played an active role in the investigation. In fact, Detective Lenk discovered a RAV 4 Toyota key, which was a key piece of evidence presented during the trial. In other words, Manitowoc County was deeply involved with the Halbach investigation and had ample opportunity to let their conflict of interest and bias taint the investigation.

Moreover, even experts assembling, interpreting, and presenting forensic evidence at trial are also subject to all-too-human bias. In the case of Steven Avery, the use of certain pattern evidence, such as tire tracks, bite marks, fingerprints, and ballistic markers has come under fire for general inaccuracy.1 The use of DNA evidence, a generally very accurate tool, has also posed problems, including fraud and the inability of juries and judges to properly weigh the methodology of testing. When looking at DNA exoneration cases, 63% of wrongful convictions surveyed were the result of forensic science testing errors, with false and/or misleading forensic expert testimony being the fifth most common error.2 Considering that experts who knew that Avery was the suspect in the case developed forensic evidence, the probative value of forensic evidence in the Avery case seems very questionable and, thus, deserves discussion.

II.   Distinguishing Conflict of Interest from Bias

Conflict of Interest is Not Bias

In discussing bias, it is important, first, to distinguish conflict of interest from bias.3 A conflict of interest is different from bias—conflict of interest is a tendency toward bias.4 While the individual’s special interest in a matter gives a rise to conflict of interest, conflict of interest does not necessarily result in partial decision because one may exercise impartial judgment, despite the special interest.5 It is also important to note that conflict of interest “arise only when a person is required to exercise judgment on behalf of another.”6 Thus, if an individual is required to exercise judgment only on one’s behalf, without regard to another individual or institution for which one may work, conflict of interest poses no threat to the situation.

Obviously, however, conflict of interest is an unavoidable problem in police work. Conflict of interest in police work is especially noteworthy because “[u]nlike other decision-makers in the criminal justice system (e.g., prosecutors, judges), police officers make legal decisions in a context of low visibility [, with] . . . wide discretionary power over who will be subject to legal intervention and control.”7 Although we ideally expect police to enforce the law uniformly, many scholars agree that uniform enforcement is neither possible nor desirable in police work because of situational factors that influence police work.8

Studies: Impact of Conflict of Interest on Police Arrest Decisions

Given that situational factors influence police work, the relevant question in the context of the Avery case will be whether Avery’s lawsuit against the county and, thus, the officers’ incentive to prove to the public that they were right to convict Avery in the first place because he is the bad guy, led the officers to exercise partial judgment against Steven Avery. Studies seem to support such concern of conflict of interest playing a role in the Avery case.

One research examines police arrest practices by studying public drunkenness encounters.9 During the 15-months period, a quantitative participant-as-observer study was undertaken, with the observers riding on a randomly selected patrol car for a full shift. In this study, Lundman observes that the probability of arrest increased, as public drunkenness offenders are disrespectful in their interaction with the police.10 That is, when making the arrest decisions, the police are sensitive to the offenders’ demeanors and more frequently arrest those who show disrespect for the police: “Drunkenness offenders who evidence disrespect by being impolite in their interaction with the police or by failing to comply with police orders are more likely to be arrested; offenders who evidence respect are arrested significantly less frequently.”11 Noting that police are abusing the authoritative privilege by exercising non-negotiable coercive force, Lundman emphasizes that it is not illegal or criminal to be disrespectful of police.

In another research, Smith and Visher examine variations in police arrest practice by looking at factors such as arrest location, presence of bystanders, suspect race, and suspect antagonism.12 The data used in this study were collected as part of a larger evaluation of police services, with trained civilians riding on a patrol car as observers. In this study, Smith and Visher observe that in making arrest decisions, police do consider the legal seriousness of the offenses but also consider situational factors—that is, non-legal factors (see Figure 1 below).13 Specifically, they find that police respond with a higher incidence of arrests in encounters with antagonistic suspects, those who offer a direct challenge to police authority. Smith and Visher note that the inequality of power and authority between police and the public allows police to effectively control people and that police act in manners that maintain such disparity.


Both studies commonly find that behaviors that challenged police authority resulted in increased rate of police arrest. That is, non-legal factors influence police arrest decisions. Going back to the Avery case, Manitowoc County Sheriff’s officers probably saw the lawsuit for wrongful conviction as a challenge to police authority, putting police to public shame. Although the above studies examine the impact of non-legal factors on arrest decisions at the scene of offenses, it is reasonable to extrapolate that the studied non-legal factors would also influence police investigation in relation to those suspects exhibiting behaviors that challenge police authority. Considering that police officers are often “duty-bound to infringe or curtail people’s rights,”14 conflict of interest is a noteworthy element of police work. While appropriate exercise of impartial judgment will justify curtailments of rights, partial judgment due to conflict of interest will taint investigations and potentially harm innocent individuals.

III.   Pervasiveness of Bias in Police Work and Forensic Evidence

Confirmation Bias & Adversary Bias: Two Sides of the Same Coin

Confirmation bias is a broadly known psychological phenomenon that refers to “the tendency to bolster a hypothesis by seeking consistent evidence while minimizing inconsistent evidence.”15 Put in simple words, confirmation bias describes individuals’ tendency to search for or interpret evidence in a way that confirms their preexisting beliefs.16 In criminal investigations and forensic work, confirmation bias can negatively influence the accuracy of investigation and forensic work.17

In the video below, Dr. Itiel Dror addresses the role of confirmation bias on the objectivity of expert testimony in the context of the adversarial legal system:

Dr. Itiel Dror Keynote Address on The Psychology and Impartiality of
Forensic Expert Decision Making

A related concept, adversary bias refers to “witness bias that arises because a party to an adversarial proceeding retains experts to advance its cause.”19 In other words, adversary bias describes the bias that arises when experts work for a partisan team in an adversarial setting and, therefore, tend to reach an opinion in favor of that team. Adversary bias is particularly important in American legal system because of its adversarial adjudicatory system, in which each side present its arguments along with witnesses and evidence while a judge presides as the neutral party. Such system reinforces the use of expert witnesses by both parties to bolster their claims. Nonetheless, there have been questions and skepticisms towards the objectivity of expert testimony.20 Reflecting the concern for objectivity of expert testimony, the word “hired gun” has been used to describe partial expert witnesses who testify in favor of the side that hired them, without giving regards to the accuracy of the testimony that they give. More problematically, even with expert witnesses who are not consciously biased in favor of those who hire them, research shows that those experts might be subject to unintended adversary bias.

All-too-Human Bias in Interpreting Forensic Evidence

It is an overstatement that forensic evidence is free of subjectivity and bias. In 2009, the National Academy of Sciences, referring to commonly used pattern evidence like fingerprints and tracks, stated that “[n]o forensic method has been rigorously shown to have the capacity to consistently, and with a high degree of certainty, demonstrate a connection between evidence and a specific individual or source.”21 This is alarming, given the extent to which these types of evidence are admitted and used in court proceedings. One of the biggest fears with forensic evidence is that juries are not able to properly understand the scientific methods used to analyze the evidence and their veracity, or able to understand the difference between a probabilistic statement about evidence and an actual identity claim—i.e., “There is a 1 in 1,000 chance that this fingerprint does not belong to this suspect,” versus “This fingerprint definitely belongs to the suspect.” Fear that juries have inflated or unrealistic expectations of forensic evidence due to television and media persist as well.

While there are reasons to be concerned about judges’ and juries’ understanding of the nuances of forensic evidence, luckily the situation is not quite as dire as some might fear. The much-touched CSI effect—wherein jurors will improperly give far too much weight to forensic evidence due to its portrayal in media, especially crime drama television like the show CSI—seems to be mostly anecdotal on the part of judges and lawyers, with almost no empirical evidence to support claims of its existence.22 While the CSI Effect made headlines for years,23 actual surveys of jurors suggest that there really is not much of a difference in how viewers of crime dramas understand forensic evidence versus non-viewers.24 Furthermore, Judge Donald Shelton, one of the conductors of large juror surveys on the impact of media on juror understanding of forensic evidence, posits that jurors are actually reacting to rapid developments of technology, rather than what they see on television.25

The CSI: Effect26

In fact, the more problematic issues of interpreting forensic evidence surface with experts. First introduced in the mid-1980s, DNA evidence was accepted fairly rapidly by courts.27 What resistance there was due to conflicting opinions in the scientific community had greatly diminished by the mid-1990s.28 DNA testing methods have greatly improved in the years since, but instances of fraud, improper testing methods, and skewed public perception regarding the weight DNA evidence had have renewed some of the controversy surrounding reliance on DNA evidence.29 The general recognition of the usefulness of DNA evidence has led to the enactment of several federal and state statutes requiring certain individuals, including individuals convicted of crimes to contribute DNA samples to large databanks.30 Only a few cells are required to obtain DNA evidence, and those cells can come from blood, sweat, saliva, skin, or any other bodily fluids on any number of objects—even if the biological sample is invisible to the naked eye.31 DNA evidence can be used to a greater degree of accuracy than forensic pattern evidence; for example, DNA evidence of another person’s saliva found in a bite mark on a victim’s skin can be far more reliable than attempting to match the bite pattern to a set of teeth.32

However, DNA evidence is very easily compromised, and, thus, there are specific procedures and protocols law enforcement and labs doing forensic DNA testing are supposed to observe.33 There are multiple types of DNA testing, but the methodology begins with collecting the biological material, extracting the DNA, and evaluating the quantity of DNA that is recovered from the source.34 After the DNA has been isolated and extracted, a technique called polymerase chain reaction is used to produce millions of copies of the DNA segments that are of interest, allowing even tiny amounts of DNA to be tested.35 Using the appropriate techniques, the lab then, according to the National Institute of Justice, tests to provide “another opportunity for the result of ‘exclusion’ if the known individual being used for comparison is not the source of the DNA from an evidence sample of unknown origin . . . . When a sufficient number of tests have been performed in which an individual cannot be excluded as the source of the DNA by any of the tests, a point is reached at which the tests have excluded virtually the world's population and the unique identification of that individual as the source of the DNA has been achieved.”36

Studies suggest that jurors and judges weigh DNA evidence very heavily, finding it to be “the most accurate and persuasive type of evidence compared to other types of forensic evidence,” even across people with varied levels of exposure to media portrayals of DNA evidence.37 In fact, studies show that expert testimony on DNA has strong effects on the decisions jurors reach even after cross-examination is conducted.38 This, of course, may not be all bad—but the problem is, DNA evidence can be improperly tested, not following the guidelines set by the National Institute of Justice.  Especially with advanced technology, the issue is not whether or not DNA evidence can be accurately matched, but whether or not the methodology was sound enough as to make the DNA test reliable. Ultimately, the experts’ exercise of discretion in forensic work—that is, the potentially biased interpretation of forensic evidence—leads judges and juries to struggle with evaluating scientific methodology.39

Studies: Impact of Confirmation Bias on the Exercise of Judgment

Studies show the impact of confirmation bias on police investigation. In one study by Ask and Granhag, 49 experienced investigators were asked to evaluate witnesses’ testimony in a homicide case, in which there was a hypothesis that there is a high probability of the suspect being guilty.40 Participants were provided with witness statements, one of which confirmed the hypothesis, incriminating the suspect, and the other of which led to exoneration—and, thus, disconfirming the hypothesis. In this study, 42 out of 49 participants decided that the suspect was guilty, and, thus, the evaluation of witness quality (i.e., credibility and reliability) favoring the witness statement that incriminates the suspect (see Chart below).41


In another study by O’Brien, 108 college students were randomly assigned to two groups—experimental group and control group—to assess the culpability of a criminal based on the provided information.42 The first half of the information described weak evidence on a targeted suspect and no evidence on other suspects. After assessing the given first half of the information, only the experimental group was asked to identify the potential culprit. Then both groups were presented with further evidence—some of which disconfirmed the evidence presented in the first half, while others somewhat confirming the culpability of the targeted suspect. 75.9% of participants in the experimental group identifying the target suspect as guilty, while 61.1% of participants in the control group identifying the target suspect as guilty.

Studies also show the impact of confirmation bias on development of forensic evidence. Specifically, studies show that pre-existing beliefs influence interrogators, eyewitnesses, as well as professional forensic experts.43 In the study by Dror et al., five experienced fingerprint experts were asked to assess fingerprint images that they have already assessed as a match before—but without knowing that fact.44 They were also purposefully told that these fingerprints were from an erroneous identification case. Only one of the experts correctly concluded the pair of fingerprints as a match. Langenburg et al.’s study confirms such observation.45 In this study, 43 fingerprint experts were divided into three groups: control group and two experimental groups (low bias group and high bias group). Langenburg et al. find that contextual information provided in the two bias groups influence the judgment of the experts. Furthermore, Langenburg et al. note that other domains of forensic evidence involving visual similarity judgment, such as bloodstain pattern, handwriting and tire tracks, are expected to be susceptible to such confirmation bias as well.

Studies: Impact of Adversary Bias on the Exercise of Judgment

Many social science studies display adversary bias among expert witnesses. For example, Murrie and Boccacini conducted an experiment by recruiting more than 100 forensic psychologists and psychiatrists holding a PhD or equivalent and assigning them randomly to either the prosecution team or a defense team.46 Each team had a chance to meet their attorney and was provided with identical case materials and records. Also, the participants were paid $400 a day to score the standard risk measurements in forensic evaluation in four cases. In this study Murrie and Boccacini find that the participants become allied with the team that retained them, interpreting data in a way that favors the side that retained them.47

Photo cropped from Daniel C. Murrie and Marcus T. Boccaccini,
      Annual Review of Law and Social Science, Vol. 11: 48 (Nov. 2015).

While Murrie and Boccacini’s study was looking at the impact of adversary bias on mental health experts, Murrie and Boccacini suggest that the same study finding would be observed from forensic experts as well.48 Although there has not been any research done specifically on adversary bias of forensic experts, existing research on subjectivity and bias of forensic experts show parallel results.49 For example, Dror and Hampikian’s research examines the experts’ DNA evidence interpretation and finds that when experts are put in a context-free environment, they interpret the same evidence different from they would working with either the prosecution or the defense in litigations.50 Similarly, Dror and Cole’s study find that fingerprint experts’ visual judgments are “subjective and susceptible to influences,” depending on “extraneous information” such as “emotional context, expectation, and motivation . . . .”51 Such findings are not surprising considering that forensic experts work alongside with the law enforcement officials—and that in many cases, forensic labs are affiliated with the police or prosecution’s office. Such affiliation seems to reinforce adversary bias of forensic experts by giving them a sense that they work together with the prosecution as a team—which ultimately undermines the objective and neutral scientific procedure of forensic work.52

Exercise of Judgment in the Avery Case

A number of events in the Avery case involve the exercise of judgment and, thus, reflect bias in the course of police investigation and forensic work. In the Halbach case, Steven Avery became an immediate suspect soon after Halbach went missing. Although it is a norm to investigate individuals who are close to the victim in such cases (in this case, those individuals would include Halbachs brother and ex-boyfriend), when Halbach’s car was found on the Avery property, Detective Jacobs immediately called dispatch, asking whether Avery is already in custody. Clearly, at the time he made that call, Detective Jacobs did not have any other supporting evidence to put Avery at the scene of the crime.

 Halbach’s RAV 4: Found in the Avery Property

Furthermore, after Avery’s attorney raised allegation of evidence tampering—that is, Avery’s blood sample stored in Manitowoc County Sheriff Office was opened and, thus, there is a possibility that the police accessed the blood sample and planted blood drops in Halbach’s car— the prosecution asked FBI to run an EDTA test to prove that the police did not plant the blood drops found on Teresa Halbach’s RAV 4. That is, the prosecution actually communicated such intent to the lab technicians conducting the test: the prosecution’s message requesting the EDTA test stated, “The purpose of this request is to establish the presence of EDTA in the vial of blood, thereby eliminating the allegation that this vial was used to plant evidence.” One could easily argue that this extraneous information is likely to influence the lab technicians to conclude that the blood evidence was not planted, if not to think that they should conclude that the police did not plant the blood evidence. In fact, Investigator Fassbender messaged lab technician Culhane, asking her to “try to put Halbach in Avery’s property.” The close relationship between the prosecution and the lab (and thus its technicians) inevitably poses a great risk of adversary bias by giving the lab technicians an idea that they work with the prosecution as one team against the defense. Moreover, considering the easily manipulable nature of forensic evidence, one can easily call into question the probative value of the DNA evidence that ultimately imprisoned Avery, even to this day.

IV.   Conclusion

Criminal investigations and forensic work involve the exercise of judgment at every stage. Police investigations involve officers who come in close contact with suspects, and development of forensic evidence involves lab technicians who potentially add “human bias” in interpreting forensic evidence. Adding to such potential for all-too-human bias a close relationship between the prosecution and lab technicians, one cannot help but to question the probative value of forensic evidence. Interestingly, one scholar points out that introduction of such tainted evidence in the courtroom is due to the use of Daubert Principle—that the Daubert reliability test fails to filter out evidence tainted by bias.53 It may be true that the Daubert factors should be modified to filter out tainted evidence,54 but more importantly, the Avery case demonstrates that police work and forensic evidence in and of themselves should not necessarily legitimize incrimination of an individual.

1. See Kelly Servick, Reversing a Legacy of Junk Science in the Courtroom, Science Magazine (Mar. 7, 2016).

2. See M.J. Saks & J.J. Koehler, The Coming Paradigm Shift in Forensic Identification Science, Science, 309, 892-895 (2005).

3. See infra Part III for discussion of adversary bias.

4. Stephen Coleman, Conflict of Interest and Police: An Unavoidable Problem, 24 Crim. Just. Ethics 2, 3 (2005).

5. Id.

6. Id. at 4.

7. Douglas A. Smith & Christy A. Visher, Street-level Justice: Situational Determinants of Police Arrest Decisions, 29 Soc. Probs. 167, 167 (1981).

8. See Herman Goldstein, Policy discretion: The ideal versus the real, 23 Pub. Admin. Rev. 140 (1963); Peter K. Manning, Police Work: The Social Organization of Policing (Cambridge: MIT Press 1977); Richard Quinney, The Social Reality of Crime (Boston: Little, Brown & Co. 1970); James Q. Wilson, Varieties of Police Behavior (Cambridge: Harvard University Press 1968).

9. See Richard J. Lundman, Routine Police Arrest Practices:  Commonweal Perspective, 22 Soc. Probs. 127 (1974).

10. In studying individuals’ disrespectfulness, Lundman considered factors such as raised voice above a normal conversational level and threatening verbal statement.

11. Id. at 135.

12. See Smith & Visher, supra note 7.

13. Id. at 175.

14. Coleman, supra note 4, at 10.

15. Barbara O’Brien, Prime Suspect: An Examination of Factors that Aggravate and Counteract Confirmation Bias in Criminal Investigations, Psychology, Public Policy, and Law, Vol. 15(4) 315 (2009).

16. Raymond S. Nickerson, Confirmation Bias: A Ubiquitous Phenomenon in Many Guises, Review of General Psychology 2 (2): 175 (June 1998).

17. See supra note 1; Dror, Kassin & Kukucka, The Forensic Confirmation Bias: Problems, Perspectives, and Proposed Solutions, Journal of Applied Research in Memory and Cognition 2 (2013) 42–52.

18. Itiel Dror, Keynote address on The Psychology and Impartiality of Forensic Expert Decision Making (2015).

19. See David Bernstein, Expert Witnesses, Adversarial Bias, and the (Partial) Failure of the Daubert Revolution, 93 Iowa Law Review 451 (2008); George Mason Law & Economics Research Paper No. 07-11, at 103, available at

20. See Daniel C. Murrie & Marcus T. Boccaccini, 11 Annual Rev. L. & Social Science 38 (2015).

21. Saks & Koehler, supra note 2.

22. Article, Should Judges Worry About the “CSI Effect?”, 47 Court Review 20, 25.

23. See e.g., Kathianne Boniello, ‘CSI’ has ruined the America justice system, New York Post, Sept. 27, 2015 (discussing the CSI Effect).

24. See supra note 22.

25. Donald E. Shelton, Twenty-First Century Forensic Science Challenges for Trial Judges in Criminal Cases: Where the "Polybutadiene" Meets the "Bitumen," 18 WIDENER L.J. 309 (2009); Donald E. Shelton et al., A Study of Juror Expectations and Demands Concerning Scientific Evidence: Does the "CSI Effect" Exist, 9 VAND. J. ENT. & TECH. L. 331 (2006); Donald E. Shelton et al., An Indirect-Effects Model of Mediated Adjudication: The CSI Myth, the Tech Effect, and Metropolitan Jurors' Expectations for Scientific Evidence, 12 VAND. J. ENT. & TECH. L. 1 (2009).

26. The CSI: Effect (2015).

27. 2-18 Scientific Evidence § 18.05 (2015).

28. Joel D. Lieberman et al., Gold versus Platinum: Do Jurors Recognize the Superiority and Limitations of DNA Evidence Compared to Other Types of Forensic Evidence?, 14 Psych. Pub. Pol. & L. 27.

29. Id.

30. Paul C. Giannelli, Edward J. Imwinkelried, Andrea Roth & Jane Campbell Moriarty, Scientific Evidence (5th Ed.).

31. National Institute of Justice, Guide to DNA Evidence.

32. Id.

33. Id.

34. Id.

35. Id.

36. Id.

37. Lieberman et al., supra note 28.

38. Id.

39. See M. Kovera, & B. McAuliff, The effects of peer review and evidence quality on judge evaluations of psychological science: Are judges effective gatekeepers? 85 Journal of Applied Psychology 574-75 (2000).

40. Ask & Granhag, Motivational Bias in Criminal Investigators' Judgments of Witness Reliability, 37 Journal of Applied Social Psychology 561-91.

41. Id. at 575.

42. O’Brien, supra note 15, at 319-324.

43. See Dror, Kassin & Kukucka, supra note 17; Langenburg et al., Testing for Potential Contextual Bias Effects During the Verification Stage of the ACE-V Methodology when Conducting Fingerprint Comparisons, 54 J. Forensic Sci. 571-582 (2009).

44. Dror, Kassin & Kukucka, supra note 17.

45. See Langenburg et al., supra note 43.

46. Murrie & Boccaccini, supra note 20.

47. One advantage of this experiment is that the selection effect is controlled, as opposed to previous field studies where attorneys in real cases might select the experts according to the likelihood of their favorable opinion. See id. at 39-40. Moreover, while previous experiments were vignette-based experiments, this study was conducted with participants believing the situation to be real. See id. at 41. Although the real-world experts are likely to be paid more than what was paid in this study and have more interactions with lawyers, such discrepancy between the study and the real-world practice only indicates that there likely is higher probability of adversarial allegiance with the real-world experts. See id. at 46.

48. See id. at 49-50.

49. Id.

50See I.E. Dror & G. Hampikian, Subjectivity and bias in forensic DNA mixture interpretation, 51 Sci. Justice 204-08 (2011).

51. I.E. Dror & S.A. Cole, The visit in “blind” justice: Expert perception, judgment, and visual cognition in forensic pattern recognition, 17 Psychon. Bull. & Rev. 165–66 (2010).

52. In its report, the National Research Council points out that potential bias among forensic experts results from the lack of organizational and structural independence. See National Research Council, Strengthening Forensic Science in the United States: A Path Forward (Nat’l. Acad. Press 2009).

53. See Bernstein, supra note 19, at 101-37.

54. See id.