Thirteen years ago, in a fit of idealism, I left my comfortable career as a scientist and joined the Los Angeles Police Department.
Like most of the academics, I believed that the police were corrupt. I wanted to learn the reality of policing and to serve as a model for positive change. When I started, major scandals in the LAPD were breaking out. Officers were allegedly involved in the murder of rapper Notorious B.I.G., providing protection to criminal rap artists, and the Rampart scandal had tarnished the image of the world-famous police agency once again.
The LAPD was in transformation. Chief Bernard Parks had terrified the entire LAPD with his heavy-handed disciplinary system; cops were leaving the department in droves.
Though suspicious, I found that the majority of officers were professional and honorable. True criminal cops, like Rafael Perez and David Mack, were the rarity. More prevalent was misconduct that was based on laziness and cutting corners, not with a criminal intent, but with a "the ends justify the means" mentality.
The major shortcoming in policing was something far more dangerous and something that has not been addressed seriously by our criminal justice system. Major harm could result from our reliance on two very fallible tools: eyewitnesses and shoddy forensic science.
Consider a recent example: On Feb. 17, 2004, Texas executed Cameron Todd Willingham for the arson deaths of his three daughters. In September, an investigative article in The New Yorker revealed that Mr. Willingham was innocent. It sparked a series of investigations that found he was the victim of shoddy crime scene investigation and outdated theories.
The tragic miscarriage of justice in Willingham's case is not isolated. Many harrowing tales exist: Frank Lee Smith spent 14 years on Florida's death row and was exonerated as a result of DNA testing 10 months after he died awaiting execution.
The number of postconviction DNA exonerations in the US is more than 200. Since 1973, the US has released 135 people from death row due to DNA exonerations. Some were convicted due to mistaken eyewitnesses, others due to police and prosecutorial zeal and misconduct.
People, especially in law enforcement, have a hard time comprehending such stories. Don't we live in a free society where criminals have too many rights and the police's hands are tied up by too many regulations? Every cop knows at least a few criminals who walk around free because we can't charge or convict them.
Yet innocent people are railroaded through our "justice" system, all the way to lethal injections and the electric chair. Why?
The answer lies both in our human and institutional natures. As humans, we tend to believe whatever fits our self-interest, discarding facts that tend to challenge our hypotheses. The errors of deduction can therefore multiply in an investigation when shoddy science is applied or where we rely solely on eyewitnesses. As Willingham's case demonstrates, the combination can be fatal.
Early in my career, my training officer and I responded to an attempted murder investigation. The victim was shot from close range. Thankfully the shooter missed the intended target's vitals and only managed to hit the victim's biceps.
Within moments, my partner had stopped three kids who he believed had shot the victim. Our victim emphatically stated that these suspects were not the shooters. The victim's cousin, who was present during the shooting, insisted that the three were the shooters. Whom could we believe? Absent a video or other clinching evidence, how could we be certain that we were making the right decision?
Another example: A victim of an assault with a deadly weapon in my division identified the criminal. When my officers and sergeants went to arrest the suspect, he, in common with almost everyone, proclaimed his innocence. My officers spent considerable time checking the alibi of the suspect and finally determined that a video at his work established that he was indeed innocent. What if there had been no video to confirm his alibi?
Our criminal justice system would fail without eyewitness assistance to try and convict criminals. Yet numerous scientific studies have shown that even the most well-intentioned eyewitnesses can identify the wrong person or fail to identify the perpetrator of a crime.
Aside from the fallibility of eyewitnesses, our political model of control over the police can lead to inadvertent mistakes. Municipal police departments dance to the tune of their political masters who thrive on the constant drumbeat of "tough on crime" rhetoric. The only evidence that police can measure to tout our tough-on-crime rhetoric is increasing the number of arrests and reducing crime rates.
This pressure to be productive, the lack of personnel and time, and the desire to wrap up an incident, combined with the unreliability of eyewitnesses, increases the odds that an innocent person may be arrested, or worse, convicted. Add to this our system that rates prosecutorial performance on conviction rates and we are on a slippery slope. And when corrupt prosecutors who present false evidence, even in death penalty cases, such as the now disbarred Arizona prosecutor Kenneth Peasley, enter the mix, only God can save the innocent.
As the Texas father's case proves, sometimes what amounts to voodoo passes for scientific analysis of crime scenes. That the agencies which take part in prosecution and conviction (police, fire, district attorneys) consider anything less than scientific analysis should be considered malpractice.
Police departments across the country need to establish standardized tough, scientifically accurate procedures for evidence collection and analysis. Without that, innocent individuals will continue to be wrongly jailed because of an unwitting and erroneous eyewitness or an overzealous but scientifically illiterate investigator. As error-prone humans, we can't guarantee a fail-safe criminal justice system; therefore, to be truly just we must abolish capital punishment.
We need to move away from number-based evaluation of police departments' productivity and increasing reliance on crime statistics to measure police success.
It should also be mandated that all police contacts be recorded on video. And when a case for prosecution is built solely on eyewitnesses or informants, it should go through enhanced scrutiny by the judicial branch. The benchmark should be a nationwide standardization of policies on evidence collection and analysis.
The success of police agencies should be evaluated based upon satisfaction of the communities they serve. Such satisfaction surveys should be conducted by independent external entities. As procedures for police training and evidence collection and analysis already exist and only need to be streamlined and improved, none of the proposed solutions entail major financial expenditure or rely on political process, ensuring a quicker implementation. This will help create a truly just criminal justice system in our country.
Sunil Dutta, PhD, is a lieutenant in the Los Angeles Police Department. The views expressed here are his own.