The brutal beating of Brandy Marie Arreola - allegedly by a man released from jail just days before the attack - has become a key example for critics of a law shifting responsibility for thousands of convicted felons from the state to the counties in October of last year. But others, including some noted criminal justice experts, say that understanding the effects of the new law will take careful analysis beyond looking at a few headline-grabbing cases.

Parolee Raoul Leyva allegedly beat Arreola, then 20, into a coma in April of this year in San Joaquin County. Shortly before the attack, Leyva had been sentenced to jail for 100 days for violating his parole. He was let out of jail after two days because of overcrowding. Otherwise, he still would have been behind bars the day he allegedly attacked Arreola. Before the passage last year of the criminal justice reform law - AB 109 - he would have been subject to prison time for a violation.

Critics of prison realignment, as AB 109 is commonly known, say that crime rates are surging in certain cities because fewer people like Leyva are going to prison and some may be getting out of jail early because of overcrowding. Leyva's last prison term was for motor vehicle theft, a non-violent offense. Crimes classified as non-violent are now met with jail or community supervision instead of prison. Violations of parole for non-violent offenders are also met with jail time rather than prison time.

The law followed a court order to reduce the state's prison population that was upheld by the Supreme Court. The prisons were at double their capacity at the time of the order.

Now, some counties are shifting people who once would have gone to prison to jails - some of which also are overcrowded, like those in San Joaquin County. Others counties are using a combination of jail, electronic monitoring and community supervision.

"It's diminishing public safety," said Lynne Brown, director of Advocates for Public Safety, a group that represents law enforcement officers who want to repeal AB 109.

Republican legislators agree, pointing out in a press release that certain crimes have increased in cities including Stockton, Sacramento, Oakland and Los Angeles, according to preliminary data from police departments - an increase they say was caused by AB 109.

Police data offers a less conclusive statement on crime in California.

Part I crimes, those that are reported to the FBI and eventually become the uniform crime rate for the city, are up by 8.1 percent overall year-to-date in Sacramento compared to the same period last year. Homicide, however, decreased by 18.5 percent, according to Sacramento police department data.

Violent crime is currently down in Los Angeles by 7 percent and property crime is the same year-to-date when compared to the same period last year. Violent crime is up slightly in two categories in Los Angeles year-to-date as compared to the same period in 2011: rape (up 6 percent) and homicide (up 1 percent). Crimes that decreased include robbery (down 10 percent) and aggravated assault (down 5 percent), according to Los Angeles department data.

In Oakland, Part I crimes have increased by 20 percent overall year-to-date compared to the same period last year, according to Oakland Police Department data. Some increases - like those in rape (up 21 percent) and robbery (up 20 percent), are striking. Part II crimes, however, have decreased by 10 percent overall.

In Stockton, the number of homicides year-to-date is 51 - six more than in the same period last year, according to Stockton Police department spokesman Detective Joseph Silva.

"Clearly, what's happened with [AB 109] is that criminals learn there are no consequences," said Assemblyman Bill Berryhill, a Republican whose San Joaquin County district include Stockton and Modesto, of the increases. Berryhill pointed to the Raoul Leyva case and other crimes committed by people who would have been in prison before realignment and an increase in crime in general, such as a rash of gold chain snatchings in Stockton. Jail, he said, is less of a deterrent than prison.

Brown also says that the increases in certain crimes are due to realignment. "What else can we attribute it to?" she said. "We have been in a poor economy for several years."

But determining the effect of a single policy on crime rates is actually very difficult, said Joan Petersilia, professor of law at Stanford and co-director of the Stanford Criminal Justice Center. Petersilia also served as a special advisor to Gov. Arnold Schwarzenegger on prison reform.

"If we could, we would be up for the Nobel Prize," Petersilia said of knowing whether or not realignment has affected crime rates. "That is one of the hardest questions to answer in crime."

Factors that influence crime rates range from the economy and the unemployment rate to family life, Petersilia said. Diminishing police forces in cities hard hit by the budget crisis might also have an effect on crime, noted Barry Krisberg, Director of Research and Policy at the Chief Justice Earl Warren Institute on Law and Social Policy at UC Berkeley.

Looking at individual cases doesn't tell the full story either, Krisberg said. "When you release people from prison, there are going to be some bad stories."

There are, however, factors besides the crime rate that demonstrate the effects of realignment, Petersilia said. Some are more readily apparent, including changes in arrest rates, prosecution rates, shifts in the jail population, whether or not victims think the new system is working, the impact of having more offenders in the community and the impact on community resources including drug treatment programs and hospitals.

"Whether or not realignment works depends on where you are looking," Petersilia said.

Counties also had discretion in how to deal with their new responsibilities. They have responded with great variety, from sending people back into the community with ankle bracelets to putting people who once would have gone to prison in jail. Realignment in Los Angeles, which is increasing its jail population, is different from realignment in San Francisco, where the focus is on rehabilitation and reducing the jail population.

"Realignment isn't one thing," Krisberg said, "it's 58 things."

"Some counties are shifting people to jail," said David Muhammad, the former probation chief of Alameda County who oversaw the county's transition to realignment. "That's not the spirit of the law, and not the best thing to do." Muhammad is currently advising counties on realignment.

Even in counties that are more focused on rehabilitation, realignment could do more to help offenders, Muhammad said. He is advocating for creating more employment opportunities through programs like Job Corps on the county level or incentives for private employers to hire people being supervised under realignment.

More than 80 percent of the 628 people on probation in Alameda County because of realignment have had no new violations, according to data from the probation department. Eighty-nine have been arrested for new crimes, including drug offenses, property offenses and offenses against persons and weapons offenses.

Counties that have accepted technical assistance from the state are required to report on their realigned population, Muhammad said, but there are no set standards for what specific data counties must report.

The lack of funding to study realignment is striking, both Krisberg and Petersilia said. A method for assessing the effectiveness of realignment was not part of the law. Researchers, including Petersilia, are working on assessments of different aspects of realignment funded by foundations.

"The state is not collecting data on this," Krisberg said. "I think it is scandalous."

Nuanced analysis is essential to understanding the effects of realignment, Petersilia said. "We do a great disservice when we ask if it is working and only look at one measure."

Berryhill is also advocating for more consideration of realignment. He is calling for a special session of the Legislature later this year to reexamine solutions to prison overcrowding. Both Berryhill and Brown say they want a different approach, in particular one that considers re-opening closed prison facilities until more jails can be built.

In the rush to comply with the court's order, realignment failed to create a system that truly distinguished between violent and non-violent offenders, Brown said.

The first 9 months of realignment saw a 39 percent overall reduction in new prison admissions, according to an analysis by the Center on Juvenile and Criminal Justice of California Department of Corrections and Rehabilitation data. The overall prison population dropped by 26,480 between October 8, 2011 and August 8, 2012.