In the first half of 2013, as the thermometer began to rise, so did the anxiety of those living in the Chicago Police Department’s 19th District. The district’s neighborhoods—Lake View, Boystown, Wrigleyville, and North Center, plus portions of Uptown and Lincoln Square and the north part of Lincoln Park—are among the most desirable in the city. But residents were increasingly sharing horror stories about robberies, beatings, drug deals, and bloodstains on previously safe sidewalks. “I moved here in 1981, and I have never felt as unsafe as I do now,” Lake View resident Michael Smith, 56, an art director at a marketing firm, told Chicago last fall.

In fact, in the months of May, June, and July, one of the police beats within the 19th District—a small area bordered by Belmont Avenue, Addison Street, Halsted Street, and Southport Avenue—notched more robberies than any other beat in Chicago, according to the police’s own statistics. The beat also ranked among the 10 worst, citywide, for violent crime. “Does it compare to what’s happening on the South and West Sides of the city? No,” says Craig Nolden, a 45-year-old marketing manager who lives with his wife and two children in the beat. “But it was out of control.”

Last summer, Nolden says, he called 911 four times: to report someone breaking in to his wife’s car, a couple brawling in a park, a fight outside his home, and someone dealing drugs nearby. In every case, by the time police arrived, the bad guys had departed. In the case of the drug deal, 20 minutes passed before cops showed up, Nolden says. When he asked them what took so long, the officers said they were answering another call, for an attempted apartment burglary. “I said, ‘With all due respect, are we in a take-a-number situation?’ And the officer said, ‘It’s such a colossally bad issue, I can’t give you an explanation.’ ”

But even as 19th District residents remained on high alert for a roving group of thugs who were threatening victims with a hammer, police leaders were assuring them that all was well. In fact, according to reports posted by the Chicago Police Department on its website each week, crime was consistently down nearly 20 percent in the district compared with the same weeks in the previous summer.

Frightened and frustrated residents started packing formerly quiet community policing sessions. “You would go to these meetings with the [police] commander and alderman, and there was a complete denial that anything was happening,” Smith says incredulously. “They would tell us it was all our perception and point to statistics. . . . You have no idea how furious that makes people.”

By the time 2013 ended, according to the police department’s count, the number of “index crimes”—the key crimes that virtually all cities report to the Federal Bureau of Investigation—had dropped 17 percent for the year. “You can say your statistics are down,” says Sarah Gottesman, 36, a food-company brand manager who moved to Lake View three years ago, after which one friend got her wallet snatched, another had a bike stolen, and a neighbor’s apartment was burglarized. “But that doesn’t mean the crime didn’t happen.”

North Siders aren’t the only ones worked up about the disconnect between what Mayor Rahm Emanuel and Police Superintendent Garry McCarthy have claimed about falling crime and what they believe has actually been happening in their neighborhoods. Carrie Austin, the alderman of the 34th Ward on the Far South Side—which includes Roseland, one of the city’s highest-crime neighborhoods told a Sun-Times reporter in January: “Don’t tell me about no statistics of McCarthy’s. You say, ‘Well, statistically, we’re down.’ That means crap to me when I know that someone else has been shot.”

Murders grab the headlines. But they make up less than 1 percent of the total number of crimes committed. So when Emanuel and McCarthy talk about the huge drop in crime since they took over in May 2011, they’re referring mostly to reductions in the number of break-ins, car thefts, muggings, sexual assaults, and the like—the kinds of crimes much more likely to befall the typical Chicagoan.

“You can have a 100 percent reduction in murders, and as sad as this may sound, it won’t have anywhere near the effect [on the overall statistics] of a 25 or 30 percent drop in burglaries,” says Jody Weis, Chicago’s police chief from 2008 to early 2011. “If you’re looking at driving down crime, property crimes are the ones that are going to make a big difference.”

And as far as public perception is concerned, no property crimes are more important than those tracked by the Federal Bureau of Investigation for its Uniform Crime Report. Of the hundreds of crimes on the books, the FBI compiles data from virtually every U.S. city on just four property crimes (burglaries, thefts, motor vehicle thefts, and incidents of arson) and four violent crimes (homicides, criminal sexual assaults, aggravated batteries/assaults, and robberies).

The level of index crime in a city is widely viewed as a gauge of its safety—essentially, its crime report card. “[Low crime] figures serve a political end,” says Eli Silverman, a professor emeritus at John Jay College of Criminal Justice in New York City. “It brings in tourism; it’s good for business.”

As mentioned in part 1 of this report, which focused on homicides, from 1993 through 2010 the average annual decline in the number of index crimes in Chicago was less than 4 percent. If you go to the police department’s website and compare the index crime totals for 2010 (found in that year’s annual report) with the totals in the 2013 year-end CompStat report (more on that later), you’ll see a drop of 56 percent, or 19 percent per year on average. It’s akin to a chronically mediocre student all of a sudden earning straight As.

The plunge hasn’t happened in just a few parts of the city, either. Index crimes have dropped sharply in all 22 police districts. In 20 of them, in fact, the total number committed in 2013 was a mere one-half to one-third the number committed in 2010, according to police figures.

Of all index crimes, motor vehicle thefts have plunged most. Over the past three calendar years, they’re down 35 percent, again according to the department’s own statistics. (They fell 23 percent last year alone.) Over that same three-year period, burglaries fell 33 percent; aggravated batteries, 20 percent; robberies, 16 percent.

Current and former officers and several criminologists say they can’t understand how a cash-strapped and undermanned department—one that by its own admission has been focusing most of its attention and resources on combating shootings and murders and protecting schoolchildren in a few very violent neighborhoods—could achieve such astounding results. “God Almighty! It’s just not possible,” opines a retired high-ranking officer who reviewed the department’s statistics.

To get to the bottom of the numbers, Chicago studied police reports and court documents, examined the department’s internal and publicly available crime data, and interviewed more than 70 crime victims, neighborhood activists, criminologists, and former and current police sources. (Officers agreed to speak only if their names were withheld, some citing fears of retaliation.) We also reviewed a recently released audit by the city’s office of the inspector general that found the police department failed to report nearly a quarter of aggravated assaults and aggravated batteries in 2012, based on the cases surveyed.

Together, this information shows what Smith, Gottesman, and countless Chicagoans have been saying all along: The city’s crime numbers seem too good to be true. One former lieutenant has a name for the system: the washing machine. “They wash and rinse the numbers,” the lieutenant says.

Documents and interviews reveal how this may be happening. First, on McCarthy’s watch, the department quietly changed several bedrock crime reporting and scoring policies. For example, in the statistics it compiles and shares with the public on its website each week, it stopped including certain crimes that had been counted in the past.

Second, many police sources say they have been pressured by superiors—explicitly and implicitly—to underreport crime. There are, according to an expert source on the department’s statistics, potentially “a million tiny ways to do it”—including misclassifying and downgrading offenses, counting multiple incidents as single events, and making it more difficult for people to report crimes or actively discouraging them from doing so.

Finally, some of the drop is simply a byproduct of reduced manpower. Many officers say that their ranks have become so depleted that they can’t respond to all 911 calls. It’s like the proverbial tree falling in the empty forest: no victim, no report, no crime.

Since Chicago published part 1 of this special report, many current and former police officers have contacted us to share “washing machine”-like stories. Crime victims, too, have come forward to describe Kafkaesque dealings with police. And two aldermen—Scott Waguespack, from the North Side’s 32nd Ward, and Willie Cochran, from the South Side’s 20th Ward—have introduced resolutions in the City Council calling for hearings on the reliability of the department’s crime counting and reporting practices. At presstime, more than a third of Chicago’s 50 aldermen had signed.

“The allegations [in part 1 of the special report] are false, and built on information that is factually incorrect, misleading, and unsubstantiated,” police department spokesperson Adam Collins said in a statement. (See “On Accuracy” for more on this.) “We take the tracking, compiling, and reporting of crime data extremely seriously … and [it] is shared with the public to provide an accurate understanding of crime conditions.” (While Superintendent McCarthy did not agree to an interview, Collins sent McCarthy’s answers to a few of our questions via e-mail on May 2, just before presstime for this article. Mayor Emanuel did not respond to repeated requests for comment.)

Such allegations are far from new. Police departments around the country didn’t start collecting and reporting crime statistics until the 1920s. And it didn’t take long for accusations to arise that the numbers were being manipulated. A 1926 Chicago Daily Tribune investigation revealed that the police department was “covering up on the real crime situation” by “doctoring the books.” The paper found that only 60 percent of the burglaries, thefts, and “stickups” made it into the department’s public report.

So the scandals went, flaring up and then dying down every two or three decades. By 1958, the Chicago Crime Commission was accusing police of “minimizing” the city’s crime numbers. Timothy Sheehan, a former Republican congressman who ran for mayor in 1959 against Richard J. Daley, described the numbers as a “farce”: “What man with brains can believe [them]?”

In 1982, Channel 2 investigative reporter Pam Zekman unearthed evidence that the department had been “killing crime” for years by dismissing legitimate crime reports as unfounded. She reported that it undercounted robberies and burglaries by one-third or more and rapes by nearly one-half. The ensuing scandal led to an FBI review, which concluded that Chicago police were 14 times more likely to deem crime reports “unfounded” than were police in at least 30 other big cities.

Admitted Richard Brzeczek, Chicago’s police chief at the time: “There were, in fact, problems in the integrity of our reporting system.” Lo and behold, when the department released its new figures for the first four months of 1983, the numbers shot skyward: burglaries and robberies up 23 percent, thefts up 30 percent, rapes up nearly 50 percent.

Inaccuracies are hardly a Chicago-specific phenomenon. In the recent past, police departments in many other big cities—Atlanta, Dallas, Milwaukee, New York, Philadelphia, Phoenix, and Washington, D.C., to name a few—have admitted to fudging the numbers. Each of those cases can be traced to the pressures of various versions of one system: CompStat.

A comprehensive method of mapping and recording crime patterns and data, CompStat was rolled out in 1994 in New York City by its police commissioner, William Bratton. The process involves weekly meetings in which top brass grill field commanders about what they’re doing to solve crime problems in their areas. Garry McCarthy, a Bronx-born detective’s son, would eventually run the NYPD’s CompStat meetings as the deputy commissioner of operations, holding commanders’ feet to the fire at any sign that crime in their districts wasn’t heading downward.

After CompStat’s launch, New York’s crime numbers plummeted. Copycat systems quickly sprang up across the country. Police Superintendent Terry Hillard introduced a CompStat-like system in Chicago in early 2000; his successors Phil Cline and Jody Weis put in place their own variations. Nearly two decades after CompStat was born, about 80 percent of the nation’s major police departments had adopted some variation of the system, according to a survey by the Police Executive Research Forum, a Washington-based think tank.

But New York’s CompStat program was dogged by talk that its numbers weren’t all that they seemed. In May 2010—four years after McCarthy left the NYPD to become the top cop in Newark, New Jersey—The Village Voice published excerpts of transcripts from hundreds of audio recordings of station house roll calls from 2008 and 2009 made by an officer in the 81st Precinct, in Brooklyn’s tough Bedford-Stuyvesant neighborhood. NYPD bosses were caught on tape pressuring street cops to downgrade crimes and discourage victims from reporting offenses, among other things, to bolster the precinct’s stats. An outside report ordered by former police commissioner Raymond Kelly would confirm that manipulation went well beyond one precinct and “may have an appreciable effect on certain reported crime rates.”

In their 2012 book The Crime Numbers Game: Management by Manipulation, Silverman and John Eterno, a former NYPD captain who is now a professor of criminal justice at Molloy College in Rockville Centre, New York, describe how over time data-driven policing turned the NYPD to the dark side. “Initially, they started off with all the good intentions in the world, but then the political pressures and the pressures from the top became so great that they changed,” explains Eterno. “When the pressure comes from the political powers, it is very difficult for honest working police officers to battle that.”

At first, Eterno and Silverman say, producing steep declines is easy. Silverman compares the process to squeezing an orange. “It’s very easy to extract the juice when the orange is new,” he says. “But there is a point where it’s more and more difficult. So you have to be more creative.”

In early 2011, Mayor-elect Emanuel went hunting for a new top cop who would help him fulfill his campaign promise of reducing crime. He selected McCarthy, extolling his numbers-driven approach to policing. McCarthy brought to Chicago both CompStat and his friend and former NYPD colleague Robert Tracy to run it.

Before McCarthy’s arrival, the Chicago Police Department classified and reported incidents in just one way. The numbers that it shared with the public—for example, in clear, detailed annual reports—were the same ones it gave to the state of Illinois and the FBI every month for their Uniform Crime Reporting programs. (Very slight variances between those numbers can exist if the department reclassified cases after it released the annual report but before it submitted the year’s numbers for the UCR.)

Now the department essentially keeps two sets of books. One follows FBI rules for what crimes to count and how to count them. The other—in CompStat—doesn’t always. “CompStat is created by the administration for their own purposes, to do their own analysis,” explains the expert source on the department’s statistics. “They collect whatever data they want, they use it however they want to use it, and they don’t have to tell anybody [what data they’re collecting or how they’re doing the counting].”

For example, McCarthy’s administration doesn’t include arson in CompStat reports. Spokesman Collins did not directly answer repeated requests as to why, but did tell us that the department supplies arson figures for the Uniform Crime Report. The FBI’s 2011 and 2012 UCR reports, however, have blank spaces where Chicago’s arson figures should be. “Chicago arson counts were incomplete for those years,” explains an FBI spokesperson.

Now take the assault-related crime category. It includes 34 classifications, from aggravated domestic batteries to nonfatal shootings, based on the severity of the injury (or threat of injury), the kind of weapon, and the location. The UCR programs for Illinois and the FBI require all types to be recorded and counted. But the police department doesn’t count all types in CompStat.

Look at theft, too. For decades, the department counted a theft as a theft. Stealing $1 was the same as stealing $1 million. That practice follows Uniform Crime Report guidelines, which clearly state: “Agencies must report all larceny offenses regardless of the value of the property stolen.”

But CompStat counts only felony thefts—that is, thefts of more than $500. The difference between the number of felony and nonfelony thefts is huge. In 2010, the last full year before McCarthy took over, the department reported 74,764 thefts for the UCR. In 2011, under the CompStat system, the department’s year-end report listed just 15,665 thefts.

That same year, the department gave the FBI a theft number of 72,373. That works out to a theft reduction of a little over 3 percent from 2010 to 2011, right around the long-term annual average reduction before McCarthy arrived. But look at the CompStat numbers and you’d see almost 60,000 crimes effectively subtracted. Poof!

You don’t need to be a rocket scientist to realize that not counting nearly 60,000 index crimes can do wonders for a city’s overall crime rate. This accounting change helped drive the total number of index crimes from 152,031 in 2010 (according to the department’s annual report) to 86,174 in 2011 (according to CompStat).

Collins told Chicago that CompStat “is an internal tactical analysis and crime reduction strategy” and “is not a reporting mechanism in the way that UCR is.” However, McCarthy and Emanuel have used CompStat as the source of official statistics they give the press and the public, even elected officials.

For example, on October 31, when the superintendent appeared before the City Council for an annual budget hearing, Cochran asked him for crime statistics by ward. On November 22, McCarthy sent a memo to budget committee chairman Carrie Austin stating, “Crime statistics are not available by ward. They are available by district.” In an attachment, he included one week’s worth of CompStat reports for the city’s 22 districts and three areas, plus a citywide report. That’s right: He didn’t send UCR statistics. He sent CompStat statistics.

Emanuel and McCarthy have also compared CompStat crime data with pre-2011 crime data, a practice that Eterno says is “comparing apples and oranges.” Consider the department’s January 1 press release declaring, “Chicago Ends 2013 at Historic Lows in Crime and Violence.” It states, among other things, that “felony thefts were at the lowest level since 1972” and “both criminal sexual assualts and aggravated battery/assault [sic] were at the lowest level since 1982.” The year-over-year drops cited in the press release—down 23 percent in motor vehicle theft, 22 percent in burglary, and 18 percent in murder, for example—exactly mirror those in the 2013 year-end CompStat report. (The FBI has not yet made final UCR data available for 2013.)

By the way, the year-end CompStat report is one of the few that remains on the department’s website longer than seven days. Each week, the department typically replaces that week’s CompStat report with a new one, removing the old one from public view for good.

And what about those easy-to-understand annual reports that the department had been issuing since 1965? McCarthy discontinued them. Collins says there’s no longer a need for them, given that “we’ve increased data available to the public and put it right on our website.”

Well, not exactly. On the department’s home page (click on “Crime Statistics” in the News tab and what you’ll see are CompStat data. The numbers are helpfully crunched to show percentage changes in the frequency of seven crimes (representing all of the index crimes except for arson) over the past week, four weeks, one year, two years, and three years.

There is a more detailed public source of crime stats—not on the department’s website, but on the city’s. This data portal contains raw crime numbers in enormous spreadsheets—not exactly user-friendly. What’s more, nowhere will you find numbers of arrests, numbers of murders solved, or data on offender and victim gender, race, and age, for example, all of which had been included in the annual reports.

Finally, another significant discrepancy between CompStat data and FBI data is due to how the department counts cases in which there are multiple victims. Imagine that a group of thugs armed with baseball bats beats up three people walking along the street. Three victims, three incidents, according to state and federal guidelines. The inspector general’s audit found that the Chicago Police Department was counting such cases as only one incident. The result: The department had underreported assault cases by 24 percent in 2012, based on the sample cases reviewed.

Collins seemed to imply that McCarthy didn’t realize there was any problem: “The issue was brought to this administration’s attention during the course of [the] audit.” Noting that the previous administration failed to follow the guidelines, the department pledged to review all of the aggravated assault and battery cases from 2012 and 2013 to correct them.

Incidentally, McCarthy’s administration records data for shootings­—a category he created in Chicago’s CompStat system—exactly the same way: by incident rather than by victim. For example, when 13 people got shot in last September’s mass shooting at a pickup basketball game on the South Side, the department counted the shootings only once.

No wonder many in the police rank and file don’t trust the CompStat numbers. Just read some of the grousing on the widely read blog Second City Cop about “Con-stat,” “Compost,” and “Comcrap.”

Last winter, as the polar vortex hunkered down for a bone-chilling stay, an elderly woman heard a knock on the front door of her house on the Northwest Side. (She requested that her name be withheld, saying that she was afraid of angering police.) The woman peered out and saw a young man “who looked like Justin Bieber,” she says.

He seemed harmless enough, so she opened the door. “He asked me if I had seen his dog.” The woman told the Bieber doppelgänger she hadn’t seen the lost animal. But he kept standing there. “He wouldn’t go away,” she says. So she closed and locked the door and went upstairs to do some chores.

Soon after, she heard a loud noise coming from the first floor. She went downstairs to investigate and saw three men wearing black hoodies, their faces covered by scarves, creeping down her hallway. When the intruders saw her, they fled. “I went upstairs and called 911,” she says.

A few minutes later, the shaken woman heard two cops. They had come in through the back door, which the men had broken open. (The front door was still locked.)

The officers handed her a police report and left. Relieved that the whole thing was over, the homeowner tried her best to fix the back door. The men had ripped the frame from the wall, knocking off a metal plate. Wood chips littered the floor. “I got a mallet and a hammer and some wood putty, and it held pretty good,” she says.

On the report, police had recorded the incident as a case of criminal trespassing (which state law defines as entering private or public property after being told not to, either verbally or in a written notice). But a family friend pointed out that “they didn’t just walk in; they broke in.” That’s burglary: a forced or unlawful entry into a home, business, or garage with the intent to steal. It’s both a felony and an index crime. Trespassing isn’t either one.

Days later, when she called police to follow up, the woman reminded the officer that the three masked men had broken in through her locked back door. “I told them my report wasn’t complete, that there was nothing about the broken door,” she recalls. “He told me that it didn’t make a difference and just kept telling me to make sure my doors were locked.”

Sure enough, a second report mailed days after their conversation still read criminal trespassing, not burglary. At press time, that had not changed.

Veteran detectives who reviewed this case for Chicago were not surprised. “It’s become standard operating procedure,” says one sergeant. “Cops are being told by their supervisors to change things and change classifications. We have a lot of cases where a neighborhood is experiencing a rash of burglaries and the cases are basically being eliminated.”

Because the current Chicago Police Department culture is built on a wink-and-nod axiom that crime must only go down—and because careers are made or broken by good or bad “performance”—pressure to reduce crime numbers in this way has become acute at all levels. Patrol officers have to please sergeants; sergeants have to please lieutenants; lieutenants have to please captains; they all have to please district commanders; district commanders have to please the chiefs with the gold stars; and, of course, the superintendent has to keep his boss, the mayor, happy.

Much of the daily pressure lands on the beat cops on the bottom and the sergeants and lieutenants supervising them. Officers say the punishments for reporting numbers their supervisors don’t like can include denying time due, splitting up longtime partners, and changing shift times so that officers can’t make it to, say, their children’s Little League games.

Multiple officers have the power to downgrade a crime at just about any stage of the process. “It happens all the time,” says one detective. “It’s so easy.” First, the responding officer can intentionally misclassify a case or alter the narrative to record a lesser charge. A house break-in becomes “trespassing”; a garage break-in becomes “criminal damage to property”; a theft becomes “lost property.” The former lieutenant mentioned earlier calls this the “speed wash cycle”: Major crimes are immediately rinsed out of the stat books.

If the officer’s supervisor won’t accept the initial classification, the officer almost always has to change it. “What are you going to do?” says one patrol officer. Meaning: That’s how top-down command works.

Victims usually don’t find out about the reclassifications until later. Like the man we spoke to whose garage was set on fire but who later discovered that his case had been classified not as arson but as simple criminal damage to property—which is not an index crime. Or the college student who was knocked to the sidewalk and robbed of his phone and laptop, only to be told afterward by his insurance company that police reported his case as lost property. (The result: His insurer refused to cover it.)

Or the professor whose wallet was snatched out of her purse last winter when she was leaving a concert at Symphony Center with her husband, headed to their car in the underground garage at Grant Park. Her 911 call was routed to someone who “said he could not take a report until we first alerted the credit card company,” her husband says. “It sounded bizarre and, to us, irrelevant.”

But the couple did as they were told, then again called police. Weeks later, the department sent them a report categorizing the incident as lost property. “That struck me as very strange,” he continues. “It was clearly a criminal case.”

Later, their bank statement revealed that someone had somehow withdrawn $2,000 from their account. Now the original case report was becoming a serious issue. So the husband headed to the police district office and asked a desk officer to correct the original report, which the officer did. A copy provided to Chicago leaves little room for misunderstanding: “This was a theft, not a loss, and should be classified as such.”

Four months later, though, nothing has changed. At presstime, the city’s data portal shows no such theft case.

While jogging near Montrose Beach on a beautiful day last fall, a 20-something woman took a path to the bird sanctuary, a secluded area shrouded by trees, bushes, and brush. (Because of the nature of the crime, we are withholding her name.) Once she was out of sight of other parkgoers, a man appeared. Clutching his crotch, he lunged toward her and shoved his hand between her legs, grabbing her genitals.

Adrenaline surging, the woman punched him and then chased him as he ran away. “Stop him!” she yelled. But no one did. So she kept after the man until he reached the street, jumped into his car, and sped off. She called 911 on her cell phone and told the dispatcher everything, including the man’s license plate number. And she waited.

Half an hour went by, she says, before a 19th District police officer arrived. “I told him my story,” she says. “And the officer says, ‘I want to catch this guy.’ And I said, ‘It’s been 30 minutes already. He’s gone.’ ”

That same fall, during a heated community meeting in which residents blasted the police department for failing to respond to crimes quickly enough, the 19th District’s commander, Elias Voulgaris, gave an unusually candid speech that lasted nearly half an hour. Yes, Voulgaris said, he was short on cops. The district now had 69 fewer beat cops than when he took over in August 2012. “I don’t think there is any district commander in this city who doesn’t want more officers,” he said.

Yes, on some days the district got so overwhelmed by 911 calls that his officers couldn’t get to crime scenes in a reasonable amount of time. “It’s a major concern,” he said. “It’s a major concern citywide and, yes, part of the issue is manpower.”

It was such an issue, Voulgaris told the group, that he had to prioritize. Given the shortage of officers, he didn’t want his cops making arrests for minor crimes that would take them off the streets. “I’m being totally frank and honest with you, and I hope you appreciate that,” he said, adding, “I’ll probably be fired tomorrow.”

The group responded with a nervous laugh. The shortage of patrol officers was no secret—according to an analysis by the Tribune, the city had 779 fewer beat cops in December 2013 than it did in the fall of 2011, a decline of 10 percent—but it was rare to hear a senior official be so specific about what sorts of problems that shortage was causing.

Officers say that in busy districts, it can take hours for anyone to show up in response to a 911 call. “It sucks, but it’s true,” says a patrol officer on the South Side. “There may be only one car freed up for the entire district that can answer jobs.”

According to a 2012 federal study, only 44 percent of violent crime victims (and 34 percent of property crime victims) report the incidents anyway. The longer victims wait for police to arrive, the greater the chance they will leave the scene. No victim, no report. “If people are waiting around for a squad car to show up because you don’t have enough manpower, your reported crime is going down,” a detective says.

The problem is especially acute in minority neighborhoods, says Harvey Grossman, legal director of the American Civil Liberties Union of Illinois. The ACLU has a pending case against the city, on behalf of a West Side community group, alleging that the city failed to deploy police equitably across communities. The case is currently in the discovery phase.

On top of that, if you call 911 these days, you may wind up getting shunted off to the city’s nonemergency 311 call center. As of February 3 last year—days after the headline-grabbing murder of teenager Hadiya Pendleton—citizens who call 911 to report garage burglaries, vehicle thefts, thefts, simple assaults, and similar crimes in which the suspect is no longer on the scene and no one is in danger are directed to 311. (Under Weis, only 911 calls about lost property and a few other minor issues were rerouted.) McCarthy called the change “the best use of our resources,” a move that would free up 44 officers a day and absorb 137,000 hours of case reporting that was once handled by patrol officers.

Sounds reasonable in theory. But cops say the pushing of calls to 311—the line Chicagoans use to, say, report potholes or rats in their alley—has worsened the underreporting of crime. That’s because the typical wait for someone to pick up a 311 call is much longer than for a 911 call, according to both victims and officers interviewed. And when a police officer does answer, the report is almost always taken by phone. Says one cop: “Victims get frustrated because they are talking to someone on the phone [instead of in person], and they just say, ‘To hell with this.’ ”

None of this is to say that crime in Chicago hasn’t gone down since Emanuel and McCarthy arrived. It almost certainly has. After all, the city saw a steady drop in index crimes over the preceding 17 years.

That kind of modest decline was the norm in 2013 for the 10 biggest U.S. cities. Preliminary FBI statistics show that index property crimes (burglaries, thefts, auto thefts, and arsons) declined by an average of less than 3 percent last year in those cities. Of those, burglaries and motor vehicle thefts fell the most: by just under 7 percent each, on average. In contrast, Chicago recorded seismic drops of 22 and 23 percent, respectively, for those two crimes in 2013.

If Chicago has indeed succeeded in slicing property crime by such huge amounts—especially when most of its manpower has been devoted to preventing shootings—then you’d think Emanuel and McCarthy would have been touting their winning strategies. The two have never been bashful about their accomplishments. But when was the last time you heard them describe their strategies to prevent burglaries? Or robberies? Or auto thefts? (When Chicago asked Collins for those details, he gave a four-point response that basically boils down to: The decreases shown in CompStat are due to CompStat.)

It would be easy enough for an independent authority to make sure that crime statistics are accurate. But none do. The Illinois Uniform Crime Reporting program—a one-person office within the state police department—merely submits the department’s numbers to the FBI after checking for data-entry errors, not for accuracy, according to a state police spokeswoman.

And the FBI? Its Uniform Crime Reporting program leaves the counting and reporting to the police. (The departments can refuse to participate, by the way.) Which essentially means the departments police themselves. Audits are rare: From 2009 to 2013, the FBI audited just 451 of the nation’s nearly 18,000 law enforcement agencies that report crime data to the bureau. The last time the FBI audited the Chicago Police Department, the FBI spokesman said, was April 2006.

Here’s the capper: Even if the FBI finds problems, it cannot impose any penalties. “It’s an absolute joke,” says Eterno.

Even the recent audit by the office of Inspector General Joseph Ferguson was quite limited. It examined only 383 of the 83,480 assault-related incidents reported by the department in 2012. And it didn’t check to make sure the facts in the reports matched the victims’ original complaints.

When we asked Collins what safeguards the department has to prevent or catch the manipulation of crime reports, he pointed to the department’s Quality Assurance Unit. It randomly checks case reports for accuracy, Collins said, and reviews every report that has been reclassified. In 2012, the department’s Robert Tracy told Chicago: “I have quality assurance . . . ensuring that the numbers we put out publicly are accurate. Nothing worse can happen than not having the right numbers.”

Tracy, as you’ll recall, also runs the CompStat program. It’s the old fox and hen house dilemma.

Changes to how crimes get recorded and counted. Allegations by cops of rampant downgrading and underreporting. Lack of transparency. All this has a familiar ring to Silverman, the criminologist, who has studied police data practices in numerous U.S. cities for years. His opinion: “It’s like they wrote the script here [in New York], and you guys [in Chicago] took it.”

Since part 1 of this special report was published on April 7, strange things have happened to the department’s crime data. On April 11, RedEye reported that eight homicide cases were missing from the 2014 year-to-date murder total on the city’s public data portal. Collins called the problem a “temporary technology glitch.” An online vendor, he added, would “re-sync” the data that same day.

But three weeks later, the problem hadn’t been fixed. Also, when we checked the portal a week after part 1 of this story ran, 364 murders were listed for 2013, though the department’s widely reported official year-end tally was 414. That error persisted until at least May 1. Also on May 1, the department announced 95 murders for the first quarter of 2014, but the portal showed just 72 for that period. (It has since been corrected.) The data for 2011 and 2012 are consistent with previously reported totals.

At the end of April, another odd thing happened. A blog called Crime in Wrigleyville and Boystown found that the data portal was missing at least 22 percent of all reported crimes in Lake View between March 15 and April 14. When we asked Collins about it, he said “the issue should be completely correct by the end of the day [May 5].”

Just as puzzling, a week before part 1 of our story was published, the department stopped its usual practice of posting CompStat reports on its website each week. It did not resume posting until May 2, at which time it put up a report covering the week ending April 27. What about data for the weeks ending March 30, April 6, April 13, and April 20? Never posted. “The person who updates the CompStat data on our website was on vacation for several weeks,” Collins responded, calling the lapse an “administrative oversight.”

These issues, and others described in this story, aren’t just academic. Reliable data are integral to virtually every police-budgeting and crime-fighting decision, from how many officers to hire to where to deploy them. Undercounting allows crimes to go unsolved and criminals to get lighter penalties. And it leaves victims—the real people who become the statistics—feeling betrayed, distrustful, and vulnerable.

Consider the unfortunate jogger mentioned earlier. According to veteran police sources, her attacker’s behavior constitutes criminal sexual abuse, defined as sexual conduct by the use of force or threat of force. And that’s exactly how the police report, filed a few days later, classified the incident.

Shortly afterward, however, a detective told the woman that the crime was not sexual abuse after all, but battery. The woman recalls the detective saying that “the man who attacked me would’ve had to have said something sexual or touched himself” for the crime to be sexual abuse. So she reiterated: “He grabbed his crotch and grabbed mine.”

She says the detective told her, “To be honest, I don’t think the state’s attorney is going to do anything with this. . . . It’s just not a big deal compared to all the other crime that’s going on.”

Maybe so. But according to the FBI, crimes are supposed to be classified and counted based solely on the police’s investigation, not on what officers think prosecutors might do later in court.

Weeks later, the woman learned, to her dismay, that the police report had been changed anyway. It no longer contained any reference to the sexual nature of the attack, which was now classified a simple battery—the kind of charge given to a person who, say, slugs someone.

“It was a sexual attack,” the woman insists. “He could do it again, and he could do it to someone who couldn’t fight back, who couldn’t get away. . . . It’s not fair that the sexual nature of this wouldn’t be on his record.” (Because the victim captured the man’s license plate number, the police caught a suspect; the case is working its way through the courts.)

“There’s no reason to reclassify [this case] to a battery,” agrees a veteran detective who reviewed the case at Chicago’s request. “Unless someone just wanted to make the lakefront seem a lot safer.”

Before Commander Voulgaris and his officers closed last fall’s meeting in Wrigleyville, the 19th District’s community policing sergeant, Jason Clark, urged the people gathered in the room to be aware of their surroundings. He reminded them to call police when they spot anyone suspicious. He added that the department had grant money coming in for “personal alarms” that they could attach to their key chains. The crowd gave him a blank stare.

Moments later, everyone stood up to leave. It was already dark. Toward the back of the room, an older woman gathered her things, including the packets of monthly police statistics distributed when the meeting began. She stopped for a minute and turned to a neighbor.

“Could I get someone to walk me home?” she asked.