Scientific Evidence, Politics, and Convictions – The Deeper Civil Rights Implications of Academic Freedom

Science Doesn’t Lie

I recently saw a sign from the “March for Science” saying, “Science doesn’t lie.” Is that true? Is it even scientific? It is an important question because every day attorneys ask jurors to evaluate scientific evidence in cases from DWIs to rape and murder.

Under the rules of evidence, an expert may testify about scientific evidence “if the expert’s scientific…knowledge will help the trier of fact to understand the evidence or to determine a fact in issue.”¹ The testimony must be “based on sufficient facts or data;…the product of reliable principles and methods;” and the expert must have applied the methods and principles reliably.²

Generally, to be admissible, scientific testimony must pass the Daubert/Robinson test. In Daubert the Supreme Court provided guidance for judges and juries to evaluate the reliability of scientific evidence.³ The Daubert case involved a drug that parents believed caused their childrens birth defects. The court upheld the dismissal because there were no peer reviewed studies to support the claim. In Robinson the Texas Supreme Court outlined some factors that could be used to evaluate scientific evidence, including:

• whether the methodology has been subject to peer review,
• whether the methodology or opinions are generally accepted within the scientific community,
• the extent to which the theory can be tested,
• and the error rate⁴

Or Does it?

But what happens when the scientific community is dominated with monolithic political views? What good does it do to subject your theory to peer review when all of your peers agree with you? In 2015, Behavioral and Brain Sciences published a paper arguing that academic psychology has lost nearly all of its political diversity in the last fifty years, “most likely due to a combination of self-selection, hostile climate, and discrimination.”⁵

The paper argues that bias resulting from the lack of diversity (studies suggest only 5%-8% of social science professors in the United States identify as conservatives) reduces the quality and can “undermine the validity of social psychological science.”⁶ Researchers can imbed their values into their theories and methods, negative attitudes can cause researchers to mischaracterize conservative values, and confirmation bias can cause them to look only at evidence that supports their assumptions. The paper cites examples of this. For instance, if you disagree that “we will soon experience a major environmental catastrophe,”then you are charactarized as in “denial of environmental realities.”⁷ Or, if you agree that hard work gets results, then you are engaging in “rationalization of inequality.”⁸

So what if it does?

This has profound implications in law under Daubert. Take for example a child custody evaluation. Child custody evaluations are a wonderful tool and I use them. But let’s see an example of how the political environment in higher education can have implications that people don’t think about.

A psychologist evaluates a child and parents and may come up with a numerical score to make a recommendation on who should be making moral, religious, and educational decisions concerning the child. It all sounds so objective and scientific if numbers don’t lie and science doesn’t lie. Take a moment to consider how insidious this could be.

The psychologist  may have no bias whatsoever. The psychologist is most likely trying to be as fair and objective as possible. The bias is imbedded in the number. The number came from a methodology which is based on research embedded with values that half the country does not share. The methodology is accepted because everyone has the same biases. Because of this, the bias is built into the rules of evidence under United States and Texas Supreme Court decisions. This probably has not happened yet, but the bias on college campuses has only been getting worse over the years. How will things look 20 years from now? 30 years?

Criminal law has more profound implications. Mental health professionals are increasingly used in criminal cases. Imagine you are the Defendant in a self-defense case. The alleged victim had a gun and you claim you were being robbed. The prosecution says no, Defendant shot the alleged victim because Defendant is a racist. He puts on an expert who testifies that you have “white supremacist” tendencies. The expert’s conclusion is based on, among other things, the fact that the Defendant thinks taxes should be lower and likes to wear red hats — two things a hypothetical future study says are associated with White Supremacists. Would you want the guy holding the “Science doesn’t lie” sign to be on the jury?

So what if it doesn’t?

Confirmation bias affects everyone. “People tend to search for evidence that will confirm their existing beliefs while also ignoring or downplaying disconfirming evidence.”⁹ There are two recent cases where political confirmation bias may have influenced prosecutors. This is different than a political prosecution. When a prosecutor prosecutes someone because of his political beliefs or other invidious political reasons, then the prosecutor is engaging in political prosecution. On the other hand, when an investigator reaches a conclusion prematurely due to politics so that all of the subsequent evidence the investigator sees confirms it, then it is political confirmation bias. In other words it is unintentional.

The first case is the infamous Freddie Gray case. I won’t get into whether the officers were guilty or not guilty of anything, but there was a lot of conflicting evidence. Witnesses testified about the officers breaking bones that weren’t broken. His neck on video appeared to be broken prior to transport. The medical investigation found that Gray had sustained the injuries when he was standing in the van due to an abrupt change in direction. The prosecutor said the knife he was carrying was legal. The knife was in fact illegal under city code. The officers’ acquittals and subsequent lawsuits seem to indicate that the officers were over-charged because the prosecutor focused on evidence confirming her beliefs and ignored or minimized other evidence. The over-charging was likely due to political confirmation bias in the wake of violent protests. It may be that the officers would have been found guilty of something if they had been charged based on the evidence viewed objectively.

The second case is the case against infamous police officer, Daniel Holtzclaw. Crime Watch Daily is doing a three part series on this case and you are welcome to reach your own conclusions. For purposes of this post, I am going to assume he is not guilty. As presented by the Hortzclaw family and CRTV, Officer Holtzclaw was accused of sexual assault by a woman he stopped. Detectives in this case immediately assumed that he was guilty. During the investigation, detectives went out in search of women who would say they were victims and found some. Among the conflicting statements the alleged victims made to investigators was that the white six foot one inch former linebacker was a short black man. Detectives minimized and even failed to look for evidence favorable to Hortzclaw, including tests to exclude innocent DNA transfer. Hortzclaw was tried for 32 counts of related offenses of which he was found guilty of 18.

Unfortunately for Holtzclaw the lab found DNA on his pants. That sentence sounds pretty damning. DNA is science; there was science on his pants! Immediately your brain relaxes because it doesn’t have to make a decision. Someone else already showed he is guilty with DNA. We love DNA! The problem is, DNA is probably the biggest liar in criminal law¹⁰. It convicts where it shouldn’t and exonerates people who are guilty of horrible crimes. The truth is, if you leave the house for a normal day, there is no way you are going to make it home without getting someone’s DNA on your pants. That could be you behind bars because you shook the wrong sweaty hand one day.

These are two cases where confirmation bias likely lead to criminal charges. In the first, science may have lead to unjust acquittals. In the second, science may have lead to an unjust conviction.

What are the implications?

“Science doesn’t lie” is a declaration that betrays the political irrationality of the declarant. What subject lies? History? Architecture? Badminton? As citizens, we need to be informed not compliant. What “Science doesn’t lie” means is that criticism of my beliefs is not allowed. That is not science. It corrupts science.

The implications are that attacks on free speech, the presumption of innocence, and the right to a fair trial cut deeply into our constitutional protections. There are consequences to firing a professor who asks a political incorrect question, and there are consequences to rioting to prevent students from hearing criticism of your ideas. These things prevent scientific discovery and advancement and lead to bad science. Bad science leads to bad evidence. Bad evidence leads to injustice.

Moreover, reverence for science is misplaced. When we are growing up our parents have all the answers. Once we are on our own it is tempting to believe that science has all the answers, so we treat it practically like a religion. We even say, “I believe in science.” But science isn’t something to believe in. It is a tool-one of many we need to build a just society¹¹.

¹ Fed. R. Evid. 702.
² Id.
³ Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579 (1993)
⁴ E.I. du Pont de Nemours & Co. v. Robinson, 923 S.W.2d 549, 557 (Tex.1995).
⁵ Duarte, J.L., Crawford, J.T., Stern, C., Haidt, J., Jussim, L. and Tetlock, P.E. (2015) ‘Political diversity will improve social psychological science’, Behavioral and Brain Sciences, 38. doi: 10.1017/S0140525X14000430
⁶ Id.
⁷ Id. at 4.
⁸ Id. at 5.
⁹ Id. at 7.
¹⁰ This may be hyperbole, but it is a hypothesis worth testing.
¹¹ For more tools we need to build a just society, see The Dreyfuss Civics Initiative.