We’ve been talking about crime trends a lot over the last few weeks, so I thought it’d be a good time to step back and take a look at the nuts and bolts of crime data this week. Back in July I wrote a post about a change in reporting that the FBI was allowing to enhance the estimative power of its national crime estimates.
Thanks, Jeff - This is an interesting and valuable way to look at things. I think the quality of NIBRS reporting is at least equally if not more important than quality. Increased reporting of poor quality data does not get us anywhere. At a local level, the analysis of quality NIBRS data, including location types, property, victim-offender relationships, etc., provides critical support toward the identification of patterns. What I find, however, is that many agencies are not even aware of these fields and make use of them. Many do not even get that far. Most often the majority of thefts are classified as 23H, all other larceny, but if analysis is done on location types and property types, a good percentage of these should be shoplifting. This does nothing to support strategic operations. There are further issues with burglaries and larcenies, simple and aggravated assaults, robberies, etc. Average-sized, local agencies need to be given much more incentive to report NIBRS accurately and perhaps vendors need to be called to task to better serve agencies. I could go on and on, but you get my point. Thanks for the thoughtful commentary and I will soon do what I can to support your efforts. Thanks. Deb Piehl
Is NIBRs data the same data that some departments upload to cityprotect.com?
I've been using that data for some analysis, but it has challenges. Aside from the spotty coverage, there are some holes in the data. For example, the Santa Clara County Sheriffs department provides block-level addresses for each incident, but not the city. If you want to get incidents by city, you have to do a reverse geocoding lookup, which is prone to error.
Is there a better way to get the data straight from the FBI?
The fix is twofold . Congress should prohibit any federal funding for jurisdictions that do not participate in the NIBRS program. And report writing software should automatically upload to NIBRS
Then, there's the conservative sherriffs and police chiefs in the rural/small town places (where I've spent my life) who aren't all that excited about working with the Big Bad Feds, anyway (till they need them), so they might slow-walk things like NIBRS.
Maybe it doesn't matter all that much if we capture every single crime. The Biggies - Homicide, for example - mostly get reported/investigated. Surely, many smaller crimes slip through the cracks one way or another, so there's some built-in slop. Logically, the slop can be assumed to be consistent over time, thus in relative terms, the data are valuable.
THIS year's, or THIS month's or THIS day's crime data doesn't seem all that valuable to me, anyway (altho' the media just looooves 'em). The real point is Crime Trends Over Time where the bigger datasets tend to cancel some of the short-term statistical noise.
Thanks, Jeff - This is an interesting and valuable way to look at things. I think the quality of NIBRS reporting is at least equally if not more important than quality. Increased reporting of poor quality data does not get us anywhere. At a local level, the analysis of quality NIBRS data, including location types, property, victim-offender relationships, etc., provides critical support toward the identification of patterns. What I find, however, is that many agencies are not even aware of these fields and make use of them. Many do not even get that far. Most often the majority of thefts are classified as 23H, all other larceny, but if analysis is done on location types and property types, a good percentage of these should be shoplifting. This does nothing to support strategic operations. There are further issues with burglaries and larcenies, simple and aggravated assaults, robberies, etc. Average-sized, local agencies need to be given much more incentive to report NIBRS accurately and perhaps vendors need to be called to task to better serve agencies. I could go on and on, but you get my point. Thanks for the thoughtful commentary and I will soon do what I can to support your efforts. Thanks. Deb Piehl
Is NIBRs data the same data that some departments upload to cityprotect.com?
I've been using that data for some analysis, but it has challenges. Aside from the spotty coverage, there are some holes in the data. For example, the Santa Clara County Sheriffs department provides block-level addresses for each incident, but not the city. If you want to get incidents by city, you have to do a reverse geocoding lookup, which is prone to error.
Is there a better way to get the data straight from the FBI?
Thanks!
Thanks Jeff: I added it to my "Violent and Property Crime Rates In The U.S." Len.
The fix is twofold . Congress should prohibit any federal funding for jurisdictions that do not participate in the NIBRS program. And report writing software should automatically upload to NIBRS
Then, there's the conservative sherriffs and police chiefs in the rural/small town places (where I've spent my life) who aren't all that excited about working with the Big Bad Feds, anyway (till they need them), so they might slow-walk things like NIBRS.
Maybe it doesn't matter all that much if we capture every single crime. The Biggies - Homicide, for example - mostly get reported/investigated. Surely, many smaller crimes slip through the cracks one way or another, so there's some built-in slop. Logically, the slop can be assumed to be consistent over time, thus in relative terms, the data are valuable.
THIS year's, or THIS month's or THIS day's crime data doesn't seem all that valuable to me, anyway (altho' the media just looooves 'em). The real point is Crime Trends Over Time where the bigger datasets tend to cancel some of the short-term statistical noise.
Maybe?