Is the LAPD cooking the books to make the city’s crime rate look lower than it is?

That’s the allegation from Captain Lillian Carranza of Van Nuys Division, who has filed a claim alleging that aggravated assaults have been intentionally underreported in more than one division of the LAPD.

Carranza contended at a news conference that there has been a “highly complex and elaborate coverup.” Police Chief Charlie Beck responded angrily that the accusations were “untrue,” “outrageous” and “damn lies.”

Beck noted that violent crime in L.A. is up 4 percent in 2017. “If I’m cooking the books, I’m not doing a good job,” he said.

But you don’t have to be a math major or a chef to wonder if 4 percent is perhaps the best they can make it look. Beck might have been more convincing if he had simply promised a full investigation into Carranza’s charges.

Crime statistics, cooked or raw, are not a perfect indicator of a city’s safety. Many factors are not measured, including, obviously, crimes that go unreported for any number of reasons.

That’s why federal officials use two measurements to get a more accurate picture of crime rates: the FBI’s Uniform Crime Report, and the National Crime Victimization Survey.

The Uniform Crime Report is made up of statistics reported to the FBI on various major offenses from nearly every law-enforcement agency in the United States — local, county, tribal and federal. The UCR Program dates back to 1929 and collects information on murder, rape, robbery, aggravated assault, burglary and approximately 20 other categories of crimes.

The National Crime Victimization Survey, run by the Bureau of Justice Statistics, began in 1973 and was overhauled in 1993 to improve the questions and broaden the scope. Twice a year, U.S. Census Bureau personnel interview a nationwide representative sample of about 75,000 people in 43,000 households. The same households stay in the sample for three years. New households are rotated in as the earlier participants are rotated out.

On the basis of these 150,000 or so interviews every year, the National Crime Victimization Survey estimates the proportion of each type of crime that is reported or not reported. It summarizes the reasons for not reporting crimes, as well as the demographic breakdown of crime victims and offenders. It offers a window into methods and frequency of self-defense by victims, possible substance abuse by offenders, and the experience of victims with the criminal justice system.

Carranza’s complaint relates to the type of crime reports that are sent to the FBI. She alleges that aggravated assaults in 2016 were understated by 10 percent in the Pacific and Central divisions because the crimes were misclassified as less serious offenses. Something similar was confirmed in a 2015 audit by the LAPD’s inspector general, who found that more than 25,000 aggravated assaults in the city were misclassified during the previous six years.

The LAPD established a Data Integrity Unit made up of detectives and data analysts to try to improve the department’s accuracy in categorizing crimes. Carranza says she has been telling supervisors about the data errors since 2014 but was warned against “meddling” and denied a promotion to commander.

Beck complained about Carranza’s complaints, calling her “very litigious,” but the Los Angeles Police Protective League sided with Carranza. “It’s time for transparency and honesty,” the union said in a statement, “not cooking the books to fool our elected officials and the public.”

It’s possible that the books are not cooked intentionally, but still give a misleading picture of crime in Los Angeles. Errors and omissions in reporting can leave dangerous gaps in law enforcement databases.

That’s what happened in the case of the shooter who massacred worshippers at a church in Sutherland Springs, Texas. Because the Air Force failed to inform the FBI that this individual was court-martialed in 2014 for domestic assault, the shooter’s gun purchase was not stopped, as it should have been, by a background check.

The Pentagon inspector general is now investigating the failure to report military criminal history information to the FBI, an ongoing problem that the Defense Department has known about for 20 years.

Crime reports are far from perfect. So if you have the sense that crime is getting worse in your neighborhood, even though statistics may not show it, you’re probably right.