Posts Tagged ‘csi’

As Technology Improves, Solving Murders Gets Harder (fractal weirdness)

Thursday, January 8th, 2009

Homicide Clearance Rates

In 1963, the first year of comparable recordkeeping, 91% of murders were solved. In 2007, the number was only 61%.

At the same time, the technological ability to solve murders increased dramatically. Scientific crime scene investigation significantly increases the amount of useful evidence that can be found. Digital crime labs and computerized analysis make it easier to interpret that evidence. And of course, modern DNA techniques enable police to make unbelievably accurate identifications from the smallest particle of hair or fluid. Today’s reality would have been a science fiction fantasy twenty years ago.

So what gives?

For one thing, the kinds of murders have changed. In previous generations, murder was almost always a personal matter. The victim and the killer knew each other, had a relationship. Husbands killed wives. Friends killed friends. Rivals killed each other. To begin a successful investigation, a detective would paint a bull’s-eye on the victim. The closer a suspect was to that bull’s-eye, the more likely they were to be the killer. Cases were solved not so much by technology and physical evidence, as by getting people to talk or confess. Acquaintance homicides were, and still are, often solved because the killer contacted the police or surrendered himself.

But now, a significant number of murders are committed by gang members. Gang members and drug dealers get killed by their own groups, who aren’t likely to talk lest they be killed themselves. They get killed by members of rival gangs, and may not even know their killers. Killers may even kill completely unrelated, innocent people, through mistaken identity or reckless “drive-by” shootings. Witnesses are intimidated by the threat of being killed themselves if they come forward. So relying on people to talk or confess is not as likely to solve these crimes.

For another thing, technology only gets you so far. DNA only identifies someone if you have a sample of their DNA to compare. Gunshot residue only helps if you have the suspect’s fingers in the first place. Fingerprints are harder to find than people think, and even then can only be compared to known fingerprints. In other words, technology helps you confirm that you have the right suspect, but first you have to get that suspect. And getting the suspect in the first place often means an old-fashioned investment of shoe leather — hitting the streets, talking to possible witnesses, and conducting skilled interrogations.

Because of the advances in technology, acquaintance homicides are truly being solved at a greater rate than they were in previous decades. The suspects are known, or easily found, so the DNA and other scientific tests make identifying the killer much more certain. The scientific identification also helps get confessions.

But stranger-to-stranger homicides have increased dramatically. And despite the technological advances, these continue to have a high probability of never being solved. Motive is hard to figure out. The killings are often part of a planned crime, so that less evidence will be left behind for law enforcement to find. And any connection between killer and victim is going to be hard or impossible to identify.


So what can be done?

Studies find no correlation between the number of available police officers, or the amount of their budget, and the ability to clear homicide cases. So shoving more officers on the street, or shoveling more money at the problem, is not a solution.

Studies do show, however, that cases get cleared when detectives are ambitious and they are held accountable for the success or failure of their investigations. Cases get cleared more often when the detectives have the necessary time to devote to the investigation, and when they are part of a specialized unit where everybody is focusing on the same kind of crime.

How do you get ambitious detectives? Study after study shows this to be a huge factor. Media attention can help, when there is a lot of pressure to solve a high-profile case. But in urban areas the media is often antagonistic, media praise of police rare, and so is an underdeveloped tool. Better P.R. by the police could improve ambition. Increased internal attention, status and reward for greater clearance rates would help, as well.

Solving stranger or gang-related murders requires witnesses to come forward. They fear retribution, or being punished themselves for their own crimes. Most murders, even stranger murders, are witnessed. So a critical need is to overcome witness fears.

Studies have found that most witnesses were actually involved with the crime. They either took part in some way, they brought the killer and victim together, or they tried to stop the murder from taking place. “Innocent bystanders” only make up 9% of witnesses.

Civic pride is not likely to cause the majority of witnesses to come forward. Gang culture, and the culture of the communities where such gangs flourish, teaches witnesses to do the opposite. Cash rewards sometimes help, but the amounts commonly offered are simply too small to justify the risks a witness would run if he came forward.

Ensured anonymity is a must. But in a judicial system that properly allows the accused to see and confront his accusers, anonymity cannot be ensures. Witnesses know this. Only a real and system-wide practice of concealing the appearance and identity of witnesses to violent crimes is likely to inspire the necessary confidence. And in our legal culture, we as Americans simply value the confrontation rights of the accused more than we value the evidence we might gain by limiting those rights. That’s just the way it is.


Reducing gangs themselves, and changing the culture in which they flourish, is the long-term solution.

Gangs arise within subcultures where there is little other societal bonding and community for young males, where those young males lack (or do not see) the ability to gain status and women otherwise, and where there is a general lack of control over one’s life. Entertainment media have a huge impact on perceptions of the world. These factors create perverse incentives, so that gang membership and codes of behavior can seem to be the right choice to make.

Common factors of such communities are a lack of value placed on education, a reliance on government or others, a lack of ownership, and a xenophobic relationship with the larger community. Undervalued education minimizes earnings and options in adulthood, as the lack of parental involvement kills schools and a thou-shalt-not-do-better-than-us attitude among peers kills student ambition. Reliance on welfare, the police, programs and others to take care of life’s needs leads to an endemic lack of personal responsibility, which kills family ties and any bond to a larger civic society. Illiteracy, immersion in the skewed reality of television and musical entertainments, and a perception that the rest of society is foreign and irrelevant, further impact perceptions of how the world works.

These problems have often been many generations in the making, and are not susceptible to overnight changes. Policy changes would be required that strengthen the family bond, rather than giving incentives to father children from multiple mothers without requiring any long-term ties and responsibilities. Policy changes would be required that lead community members to see themselves as part of the larger society, and not separate from it, subject to separate rules. Policy changes would be required that create incentives for parental involvement in schools, and pave the way for cultural views of education as the means to success.

Thought Police?

Monday, October 20th, 2008

brain scan

Guilt or innocence, one might say, is all in the mind. After all, there are very few crimes that can be committed without the requisite mens rea, or mental state. If we’re going to punish someone, their acts cannot have been mere accident. We want to know that they had some knowledge that their actions could cause harm, and we want that awareness to be sufficiently high as to require punishment.

The standard criminal levels of mens rea are “negligence” (you ought to have known bad things could happen), “recklessness” (you had good reason to believe that bad things would probably happen), “knowledge” (bad things were probably going to happen), and “intent” or “purpose” (you wanted bad things to happen). If your foot kicks someone in the ribs while you’re falling downstairs, you’re not a criminal. But if you kick someone in the ribs because you don’t like them, then society probably wants to punish you.

We cannot know what anyone was thinking when they did something, however. So we rely on jurors to use their common sense to figure out what an accused must have been thinking at the time.

In recent years, however, there have been enormous advances in neuroscience. Brain scans, the software that processes the data, and good science have approached levels that would have been considered science fiction as recently as the Clinton years. Experts in the field can see not only how the brain is put together, but also what an individual brain is doing in real time. Experimental data show which parts of the brain are active when people are thinking certain things, with good detail.

Functional magnetic resonance imaging (fMRI), in particular, can act as a super lie-detector. Instead of measuring someone’s perspiration and heart rate while they answer questions during a polygraph exam, fMRI looks at actual real-time brain activity in areas having to do with logic, making decisions, perhaps even lying. Experimental data of large groups is pretty good at identifying what parts of the brain are associated with different kinds of thinking.

Every brain is slightly different, of course. Brain surgeons have to learn the individual brain they’re operating on before they start cutting. So general group data don’t translate to an individual person 100%. So any lie-detector use for fMRI would have to require some baseline analysis before proceeding to the important questions.

The issue is whether it will be admissible in court. Polygraph tests generally aren’t admissible, because they’re more an art than a science. But fMRI is all science. In addition, brain scans are already widely admissible for the purpose of reducing a sentence because the defendant had damage to his brain. As forensic neuroscience expert Daniell Martell told the New York Times in 2007, brain scans are now de rigeur in capital cases. In Roper v. Simmons, the Supreme Court, ruling that adolescents cannot be executed, allowed brain scan evidence for the purpose of showing adolescent brains really are different.

Outside the United States, brain scans have already begun to be used by the prosecution to show guilt. In India, a woman was recently convicted of murdering her ex-boyfriend with the admission of brainwave-based lie detection. There was other evidence of guilt as well, including the fact that she admitted buying the poison that killed him. But the brainwave analysis was admitted.

There are deeper policy issues here. Is reading someone’s brain activity more like taking a blood sample, or more like taking a statement? The Miranda rule is there, at heart, because we do not want the government to override people’s free will, and force them to incriminate themselves out of their own mouths against their will. That’s why the fruits of a custodial interrogation are presumed inadmissible, unless the defendant first knowingly waives his rights against self-incrimination. And because the DNA in your blood isn’t something you make of your own free will, by taking a blood sample against your will the government has not forced you to incriminate yourself against your will.

So is a brain scan more like a blood sample? Is it simply taking evidence of what is there, without you consciously providing testimony against yourself? Or will it require the knowing waiver of your Fifth and Sixth Amendment rights before it can be applied?

We’re interested in your thoughts. Feel free to comment.