In 1963, Ernesto Miranda, 23, who had dropped out of school in the ninth grade and had a prior arrest record, was picked up by Phoenix police as a suspect in the kidnapping and rape of an 18-year-old girl. After two hours of questioning, Miranda confessed orally to the crime. He then wrote out and signed a brief statement admitting and describing the rape. It contained a typed paragraph stating that his confession was made voluntarily without threats or promises of immunity and that he had full knowledge of his rights and understood that the statement could be used against him. At Miranda's trial, the confession was admitted despite his lawyer's objections, and Miranda was convicted and sentenced to 20 years in prison.

Miranda's appeal eventually reached the U.S. Supreme Court. In 1966, in its landmark decision in Miranda v. Arizona, the Court established procedural requirements that law enforcement officials must follow before questioning suspects in custody, and overturned Miranda's conviction because police had not followed the new rules. The Court's 5-4 ruling specified four warnings that police must deliver to criminal suspects about to be questioned. Unless the warnings were read, nothing an arrested suspect might say afterwards during questioning, even in the anguish of conscience, could be used against him in court.

Miranda was retried and again convicted. The confession could not be used, but a former girlfriend testified that he had told her about the kidnapping and rape. After Miranda was paroled in 1972, he was in and out of prison before he was stabbed fatally in a bar at the age of 34. Ironically, a suspect in the stabbing was unsuccessfully questioned and was released, and no one was ever charged in the death.

The changes wrought by Miranda can be best understood by comparing the new rules to those in place before the decision. Before June 13, 1966, police questioning of suspects in custody was covered by the "voluntariness" doctrine. Under the Fifth and Fourteenth Amendments to the Constitution, courts admitted a defendant's confession into evidence if it was voluntary, but they excluded any involuntary confession. In making the voluntariness determination, courts considered a host of factors. For example, if police officers or prosecution investigators used physical force or the threat of force, courts deemed the resulting confession involuntary. Courts also considered such factors as length of interrogation and types of questions asked in making the voluntariness determination. ¹

Miranda v. Arizona² radically changed these rules, adding a stringent warning-and-waiver requirement. Under this approach, a confession police obtained from a suspect in custody would not be admissible in court unless that suspect had been read his or her rights. The rights specified are familiar to anyone who has ever watched a police show on television:

You have the right to remain silent.

Anything you say can be used against you in a court of law.

You have the right to talk to a lawyer and have him present with you while you are being questioned.

If you cannot afford to hire a lawyer, one will be appointed to represent you before you answer any questions.

While the Miranda "warnings" are the most famous part of the decision, perhaps even more important are additional requirements that the Court imposed. After reading a suspect his rights, an officer must ask whether the suspect agrees to "waive" those rights. If the suspect refuses to waive - that is, declines to give his permission to be questioned - the police must stop questioning. At any time during an interrogation, a suspect can halt the process by retracting his waiver or asking for a lawyer. From that point on, the police cannot even suggest that the suspect reconsider.

At the time Miranda was handed down, dissenting Justice John M. Harlan clearly warned that the decision would "entail harmful consequences for the country at large. How serious those consequences may prove to be only time can tell." ³ Other critics of the decision predicted that it would "handcuff the cops." This question of Miranda's practical effect has far more than academic significance. Since 1966, the Supreme Court has repeatedly held that Miranda is a realistic preventive measure -"a carefully crafted balance designed to fully protect both the defendants' and society's interests." ⁴ If the costs of Miranda are greater than is generally acknowledged, the Court would presumably need to rethink the current doctrine. What, then, are the costs?

Immediately after Miranda, a handful of researchers attempted to measure the effects of the decision. The studies generally suggested significant reductions in the number of suspects giving confessions under the new rules. For a recent article in the Northwestern Law Review, I exhaustively canvassed the empirical evidence on Miranda's social costs in terms of lost criminal cases. ⁵ Examining direct information - before-and-after studies of confession rates in the wake of the decision - I concluded that Miranda significantly depressed the confession rate. ⁶ For example, in 1967:

• Research revealed that confession rates in Pittsburgh fell from 48 percent of suspects questioned by detectives before the decision to 29 percent after. ⁷

• New York County District Attorney Frank Hogan testified before the Senate Judiciary Committee that confessions fell even more sharply in his jurisdiction, from 49 percent before Miranda to 14 percent after. ⁸

The reliable data from the before-and-after studies¹⁰ show that confession rates fell by about 16 percentage points after Miranda. In other words, if the confession rate was 60 percent before Miranda, it was 44 percent after - meaning that in about one of every six criminal cases Miranda resulted in a lost confession. The reliable studies also indicate that confessions are needed in about 24 percent of all cases to obtain a conviction. Combining these two figures produces the result that about 3.8 percent (16% x 24%) of all criminal cases in this country are lost because of the restrictions imposed by Miranda.¹¹ Extrapolating across the country, each year there are 28,000 fewer convictions for violent crimes, 79,000 fewer for property crimes and 500,000 fewer for crimes outside the FBI crime index.

Studying Long-Term Effects. These estimates of Miranda's harmful effects come solely from before-and-after studies that rely on data from the months immediately preceding and following Miranda. The studies accordingly fail to capture Miranda's long-term effects, effects that would reflect criminal suspects' full understanding of the protection Miranda offers them. To gain a better view of Miranda's historic effects, we need some solid statistical indicator that extends beyond 1967 and, indeed, into the 1990s.

In theory, the ideal study would review confession rates since 1967 to see whether, despite initial declines after the decision, the rates have since "rebounded" - in other words, a before-and-after study of confession rates over several decades rather than several months. Unfortunately, no such statistics exist. The only figures that do exist were gathered by individual researchers for particular cities on a one-time basis. Although broad generalizations are hazardous, confession rates before Miranda were probably 55 percent to 60 percent. ¹² After Miranda, the few studies available reveal lower confession rates. The most recent empirical study, in 1994 in Salt Lake County, Utah, found an overall confession rate of only 33 percent. ¹³ [See Figure II.]

Richard Leo's 1993 study from Berkeley, Calif., found an in-custody questioning success rate by detectives of 64 percent. If we adjust this figure for comparability with earlier studies, it translates into an overall confession rate of about 39 percent. ¹⁴ A 1979 National Institute of Justice study of Jacksonville, Fla., and San Diego, Calif., reported confession rates of 33 percent and 20 percent, respectively. When statements admitting presence at a crime scene are added, the overall rates for incriminating statements rise to 51 percent and 37 percent, respectively. ¹⁵ A 1977 study of six cities reported a confession rate of 40 percent. ¹⁶

Taken together, these studies suggest that confession rates have been lower since Miranda. But this conclusion, too, could be attacked on the grounds that studies from individual cities may not be applicable across the country. Because no national data exist, we must search for an alternative measure.