Los Angeles' red-light traffic cameras are stirring new financial concerns at City Hall, just as a police study has concluded the program helps reduce accidents.

With officials scouring expenditures to close a $200-million budget gap, updated estimates from the city's budget office are painting a sobering picture of the cash generated by the city's 32 camera-equipped intersections.

The photo enforcement program, which catches tens of thousands of violators annually, appears to be generating about $3.8 million a year in traffic ticket revenue, said Matt Crawford, senior administrative analyst with the budget office. That is millions less than previous Police Department estimates and roughly what the program costs, mostly for fees to a private contractor that supplies and operates the camera systems.

Because some red-light ticket income is dedicated to other traffic safety programs, such as school crossing guards and roadside signs, the city's cash-strapped general fund has been paying an extra $1.6 million a year to keep the camera program going, Crawford said.

That's a concern when the city is considering laying off up to 4,000 employees, said Councilman Dennis Zine, a member of the Public Safety Committee.

"I don't think we can afford to take any money out of the general fund" for the cameras, Zine said. But he noted that redirecting all camera-generated money to support the program would require cutting other services.

"No one's going to stand up and say, 'We're getting rid of crossing guards,' " he said.

The city's red-light camera program, one of the largest in the nation, has drawn praise from supporters who say it helps efficiently police dangerous intersections, discourages red-light running and frees up patrol officers for other duties.

Critics contend that the safety benefits are mixed at best and that the cameras mainly are revenue-producing tools for private vendors and state and local governments.

Mayor Antonio Villaraigosa's office is reviewing the costs and benefits of the program. "We're looking at our options," said David Beltran, the mayor's press secretary. "I don't think we're looking at eliminating the program."

Recent Los Angeles Police Department estimates indicated the cameras produced several million dollars in net revenue in recent years.

But those figures were based on the number of citations issued as a result of the cameras, multiplied by the city's potential share of penalties, officials say.

Further analysis has shown that the actual revenue the city collects is greatly reduced by, among other things, motorists' failing to pay tickets, fines being reduced by judges and a growing number of drivers doing community service in lieu of paying ticket fees that can top $500.

The downsized cash flow estimates come when the city is considering doubling the number of intersections covered by cameras and putting the program up for a new round of bidding by contractors. How the latest revenue and cost estimates may affect that effort is not clear.

Zine, a former LAPD traffic cop, said the cameras cannot add to the city's budget woes. And, he said, officials need to carefully examine whether the systems are producing significant safety benefits.

A new Police Department report argues that is the case. Serious injury accidents and potentially dangerous crashes involving red-light running declined at intersections where cameras were activated, the report says. And although five deaths were attributed to red-light violations at the intersections from 2004 through 2006, no such fatalities have been reported since the cameras were activated, the report says. Most of the cameras were activated in 2007.

Photo enforcement "is basically doing what it's supposed to," said Lt. Ron Katona.

However, the study's data present a complex picture. Comparing the six months before and after camera-equipped intersections were activated, accidents increased 5%. That figure is misleading, the report says, because many incidents were caused by pedestrians, occurred on private property or midblock or were otherwise not relevant to the photo enforcement program.

What were deemed to be red-light-related accidents declined 9% across the studied intersections, the report says. Yet, at more than a third of the crossings, those accidents increased.

Studies elsewhere have found that rear-end collisions, which tend to be less serious than broadside collisions caused by red-light violations, increase with photo enforcement because drivers make panic stops to avoid getting tickets.

In Los Angeles, red-light-related rear-end crashes remained flat at the intersections, the study found, although total rear-end accidents rose about 40%.

Zine said his committee would delve deeper into the accident data. But even using the LAPD's criteria, a 9% reduction in accidents is disappointing, he said. "It doesn't seem that significant. . . . You should be in the double digits" at 20% to 40% accident reductions, he said.