Criminal Justice Process

Investigation of crime usually involves three elements. First, public officials, usually the police, must learn that an offense may have been, or is to be, committed. Second, law enforcement agents must identify the likely offender or offenders. Finally, they must collect and preserve evidence that the courts will accept as proving the suspect's guilt beyond reasonable doubt.

Police may learn of crimes in two basic ways. Police officers may themselves observe the offense or evidence of it, or they may receive a report or complaint from someone else. It might seem that relatively few crimes would be directly observed by the police, but a surprisingly large number of offenses are in fact discovered in just this way. Police on patrol may observe suspicious behavior, as when a vehicle circling slowly in a commercial area after hours has a license plate registered to a vehicle recently reported stolen. Moreover, a great many crimes, prominently including prostitution, illegal weapons, and drug offenses, do not involve anyone inclined to complain to the police. Although voluntary informants often come forward in these cases, effective investigation largely depends on undercover police agents.

Citizen reports provide the other major source of information about crimes. Not all crimes that occur are reported to the police, and not all crimes that are reported in fact took place or took place as the initial informant described. False reports, motivated by revenge or insurance fraud, are relatively rare. The failure to report crime is far more frequent.

Victims or other witnesses of crime may not come forward for a variety of reasons. They may perceive the chance of apprehending the offender as too remote to justify the time of reporting and testifying. They may be afraid of vengeance by the offender or by those acting on his behalf. They may be related to, or on friendly or intimate terms with, the offender. When researchers estimate the crime rate by surveying sample populations and asking how often the respondents have been victimized (victimization surveys) the rate of actual crime appears to exceed the rate of reported crime by a very wide margin. It is generally agreed that homicide and auto theft are most frequently reported. It is also generally agreed that sexual assault and domestic violence are disproportionately underreported.

Once the police have determined that an offense has taken place, they must determine the likely perpetrator (or perpetrators). They must also gather evidence of guilt that will stand up in court. Although these two processes are closely related they are not identical, because some of the evidence police routinely use to identify the likely offender is not admissible in court. For example, police investigations often rely heavily on representations by informants that are based on what the informants have heard rather than what they have witnessed personally. Even if such an informant were willing to testify (and they typically are not), the hearsay rule usually would prevent the informant from testifying in court about what the informant has heard others say about the crime. Another important example is the criminal record of individuals previously arrested or convicted for crimes similar to the one under investigation. The police routinely consider the records of potential suspects, while the courts typically exclude such evidence under the character-evidence rules.

Even when the information before the police is admissible in court, the police must weigh its probative force in selecting potential suspects. Eyewitnesses are notoriously inaccurate and may give police completely incorrect descriptions of the offender. In rare cases individuals may confess to crimes they did not commit. Far more commonly they may attribute crimes in which they were involved to other persons who were not involved or involved to a lesser degree. Witnesses may shield the guilty with false alibis and the like. Physical evidence is not subject to the risk of deception but may mislead in other ways, as when illegal drugs are discovered in an automobile containing several passengers who accuse each other of sole possession.

The police must select potential suspects in the face of these challenges under severe time and resource constraints. As police departments often evaluate the work of their officers based on the clearance rate (the percentage of reported crimes that result in an arrest), the police may have an incentive to focus on the most likely suspect however unlikely his guilt is relative to that of persons unknown. On the other side of the equation the police may often have very strong suspicions about the identity of the suspect but be unable to prove his guilt beyond reasonable doubt by evidence admissible in court.

The U.S. Supreme Court has construed the Constitution to regulate many phases of police investigation. With many important qualifications, the police may not detain people on the street for investigation without some objective evidence of criminal activity; they may not search homes without a judicial warrant based on facts showing probable cause; they may search automobiles based on a determination of probable cause, without first obtaining a judicial warrant; and they may not arrest individual without probable cause to believe the individual has committed an offense. A suspected who is arrested may not be questioned without first receiving Miranda warnings and waiving his rights to silence and counsel, but the police may interview persons not yet arrested without following the Miranda rules (Miranda v. Arizona, 384 U.S. 436 (1966)).

These regulations are enforced almost exclusively by the judicial exclusion of evidence obtained in violation of the applicable constitutional rules. Police more interested in seizing a cargo of drugs than in prosecuting the courier may have little incentive to obtain a warrant before searching for the drugs. Police who are willing to lie about how they obtained their evidence likewise may feel little incentive to comply with the rules, so long as their testimony is likely to be accepted in court.

The police investigation thus follows a dialectical or give-and-take process, in which the investigators formulate a hypothesis about the identity of the offender and collect evidence tending to confirm that hypothesis until some new item of information surfaces to exonerate the suspect. It is hard to imagine the process operating any other way, because without some sense of the potential perpetrators' identities there would be no way to distinguish evidence from completely irrelevant facts. The police do not have the resources to conduct every investigation by a process of elimination in which they begin by establishing the alibi of every person in town. Yet the choice of suspects influences the evidence collected; if the police focus on Smith as the perpetrator, Jones will not be made to stand in a lineup.

"Realistic" fiction of the police-procedural variety, whether in print or on film, is not far wide of the mark in capturing the basic tenor of this process. But police fiction grossly overstates the epistemic power of the investigatory process. The fictional police always get their man. In real life, only about 20 percent of all reported crimes are cleared by an arrest.

Observers have always known that not all arrested persons are in fact guilty, but experience with DNA testing indicates that the investigatory process is more prone to false positives than many believed. Nationwide about a quarter of the conclusive DNA tests run at the request of the police exonerate the suspect. That is good news for the innocent (and bad news for the guilty) in cases where physical evidence permits a test. But it is not an encouraging sign about the frequency with which the police investigation identifies the wrong individual as the offender in the large majority of cases in which physical evidence does not permit scientific tests of identity.

In a world in which most crimes are not reported and only about one-fifth of the reported crimes result in an arrest, it is obvious that law enforcement officials must make difficult decisions about how to allocate their scarce resources. Patrol officers must be assigned to neighborhoods. Detectives and undercover operatives must be assigned to certain types of offenses. Should undercover officers be devoted to enforcing the drug laws or the prostitution laws? How heavily should the police concentrate patrol efforts in a high-crime neighborhood? Given too few officers, the residents (generally poor and often disproportionately racial minorities) may be unfairly denied police protection. Given too many police, there may be either the perception or the reality of discriminatory over-enforcement.

The model of a police-dominated investigation followed by adjudication in court must be modified to include those situations in which the investigation is run primarily by prosecutors. Prosecutorial investigations usually involve either "white collar" crime of an economic or political nature, or organized crime of the narcotics, gambling, and loan-sharking variety. In the white-collar context prosecutors will not need much assistance from the armed police and will ordinarily interview witnesses and then take sworn testimony before an investigatory grand jury prior to filing charges. In the organized-crime case, prosecutors need to work closely with the police or federal agents. The prosecutors support the officers' applications for warrants for electronic surveillance and negotiate immunity for informants, while the officers recruit or plant informants and execute the searches and arrests.