The millions of dollars poured into the 911 emergency telephone system have shaved all of 10 seconds off the time it takes a police car to get to the scene of a crime.
That's the conclusion of a newly released study of how citizens report crime. While the last 10 years have seen a slow, but steady, acceptance of dialing 911 in any emergency -- a crime, a fire, a medical emergency -- the Police Executive Research Forum study questions whether the expensive 911 network is worth the expense.
The study, ''Calling the Police: Citizen Reporting of Serious Crime,'' covers more than 3,000 crimes that took place in 1979-80 in Peoria, Ill.; Jacksonville, Fla.; Rochester, N.Y.; and San Diego, Calif. Peoria was the only city with a 911 system covered by the study. A $530,000 federal grant paid for the three-year study, which included interviews with some 4,000 crime victims, official witnesses, and onlookers. The study conducted by the forum, a national organization of big-city police chiefs and sheriffs, did not include citizen responses to fires and medical emergencies, which are also covered by a 911 system.
According to Gary Hayes, forum executive director, half of the time that ticks away between the time a crime is committed and the time a squad car arrives at the scene results from delays in reporting crimes to police.
''We found that people delay an average of 4 1/2 minutes before they call [ the police],'' says Mr. Hayes. In that context the struggle to shave mere seconds off police-response time ''becomes unimportant,'' he says.
''We're spending millions of dollars to cut seconds off the police-response time when we should be focusing on the minutes that it takes citizens to respond ,'' he adds.
The cost to taxpayers of a 911 system varies. In Atlantic City, N.J., it cost $25,000 to install the system -- and $31,000 a year to operate it. In St. Louis, installation was $900,000; annual costs are more than $1.4 million.
Roger W. Reinke of the US Commerce Department's National Telecommunications and Information Administration (NTIA) says the forum study confirms the fact that people do a lot of other things at the scene of a crime before they call the police. NTIA assists the municipalities and 13 states installing 911 systems.
Many people instinctively call the operator when an emergency occurs. ''But dialing 'O' today is little more than an exercise in frustration,'' says Mr. Reinke.
As people learn more about 911, Reinke says, response time will lessen. ''It's a lifesaver in many cases. . . . Dialing 'O' is not the fastest way to get help.''
A quick police response leads to on-the-scene arrests in only 29 out of 1,000 cases, according to the report. The 911 system raises the number of arrests by only 0.3 percent -- to 32 on-the-scene arrests per 1,000 cases. According to Mr. Hayes, this is because three-quarters of all serious crimes are called in only after the criminal has long since left the scene of the crime. So a quick response is important in only a quarter of the crimes, when there's a face-to-face confrontation with a criminal.
The study shows that ''people delay calling the police for completely rational reasons,'' Mr. Hayes says. Reasons for delay mentioned in the study include pausing to mull over whether the crime was important enough to report to the police. Some people don't want to get embroiled in the criminal justice system; some are afraid they could be blamed for the crime.
''People want to get their emotions collected first,'' Mr. Hayes says. Solutions recommended in the forum study include:
* Coaxing police departments to screen telephone calls for help. For example, a squad car may be put to better use checking out reports of a gang of kids roaming a neighborhood than driving to a burglarized home to write up a report on what was taken, Mr. Hayes says.
* Touting neighborhood-watch programs. A main reason for citizens to hesitate before calling the police is the haziness of not knowing for sure if there's something fishy going on. In a neighborhood-watch program, you know what's going on.