This increase, assuming it’s confirmed, would place the city on pace for an overall decline in crime over the first half of 2015 relative to 2014 crime levels, but ahead of crime levels from 2011-2013.
Translating Calls for Service to UCR Estimates
Calls for Service and UCR are different data sets, but analyzing Calls for Service patterns creates a relatively accurate estimate of UCR patterns with a much quicker turnaround than waiting weeks or months for UCR .
To perform this analysis, I listed all Calls for Service for Person and Property Crimes with a “Report to Follow” disposition. These crime types are defined in a city presentation on UCR. This list includes: aggravated assault, aggravated battery, arson, armed robbery, auto theft, burglary, carjacking, homicide, sexual battery and rape, and theft.
The number of Calls for Service incidents is tabulated for each day starting Jan. 1, 2011 and ending June 30, 2015. Calls for Service and UCR numbers are going to differ for several reasons. Most significantly, UCR counts victims, while Calls for Service counts incidents. In addition, although Calls for Service will capture the vast majority of incidents, it will not capture all of the crimes reported through a more thorough system such as UCR.
The table below contains raw numbers for these two data sets between 2011 and the first quarter of 2015. As this table shows, tabulating Calls for Service in this manner steadily captures, on average, 93.1% of reported UCR totals. This method isn’t perfectly accurate, but it is reliable enough to use to gauge big-picture crime trends without having to wait two months for an official UCR release.
Estimating the Second Quarter of 2015
This method shows that if we can count Calls for Service, we can confidently estimate UCR totals in very little time. Which brings us to Person and Property crime trends for the second quarter of 2015.
Calls for Service data from April through June 2015 showed 4,211 reports of Person and Property crimes over that period compared to 4,060 over the January-March period. Estimating how many UCR crimes will be officially reported can be accomplished by dividing 4,011 Calls for Service by the 2011-2015 average accuracy of 93.1% to arrive at an estimated 4,524 Person and Property crimes in the second quarter.
This estimate, if realized, represents a rise of 3.7% from the 4,386 crimes reported through UCR in the first quarter of 2015. That would mean Person and Property crimes are continuing to occur at a reduced rate compared to 2014, but the gap is narrowing slightly.
The trend of Person and Property crimes in New Orleans over the last 4.5 years as described by Calls for Service data is shown in the graph below. The 365-day pace of Calls for Service data shows a relatively steady level of Person and Property crimes between 2011 and 2012, a small rise in 2013, a larger rise in 2014 and a moderate decline through the first six months of 2015. The 90-day pace of Calls for Service data highlights how Person and Property crimes peaked around August 2014, fell through April 2015 and have increased again over the last two months.
This methodology presents a new way of using publicly available data to evaluate overall crime trends without waiting weeks or months for government sources to provide official numbers. This method is not perfect, but it creates a good estimate of UCR data that can be of great use moving forward.