A recent disagreement between the City of New Orleans and the Metropolitan Crime Commission (MCC) provides an interesting analytic opportunity. The MCC on May 20th released its annual “Orleans Parish Criminal Justice System Accountability Report,” which measures “performance accountability” of the NOPD and the DA’s office “by monitoring arrest trends and the outcomes of all felony arrests.”
It’s an interesting report that attempts to measure the efficiency of New Orleans’ entire criminal justice enterprise. Coverage of the report, however, has largely focused on one of the report’s conclusions, about why arrests may be down while crime is up. The MCC writes that:
The police force shrinking to its lowest level in 40 years is, at the very least, partially responsible for the decline in arrests over the last three years.
This comment drew a rebuke from the Mayor’s Office in the form of a press release, which said the declining arrests owed to a smart and overarching policing policy shift. The city claimed the MCC report “erroneously links the downward trend in overall arrests to NOPD force size.” The reduction in arrests, the release explained, “is the direct result of smarter policies adopted and implemented by the city over the past five years.”
The city went on to quote NOPD Chief Michael Harrison saying that “the MCC report is strong proof that the policy shift, not the size of the NOPD, has purposefully and successfully reduced the number of arrests citywide over the past five years.”
This disagreement harks back to the 1981 article by baseball statistician Bill James mentioned in a previous post, about why Murder is a Bad Statistic. James wrote: “Why do people argue about which shortstop has the best range or which catcher has the best arm? Why not figure it out? You can get a pretty good idea by abstracting information from the available data.”
Abstracting information from available crime data enables an analysis of whether the overall decrease in arrests between 2012 and 2014 is the result of declining NOPD manpower or a sign of a smart, well-executed policing strategy.
Although this analysis provides evidence for both arguments, it fairly clearly illuminates how NOPD’s manpower decline — detailed in the table below — is impacting its ability across a wide array of issues.
A good first step in analyzing this question is to compare the number of arrests per reported Uniform Crime Report (UCR) reported crime. UCR breaks its crimes up into property crimes (burglary, theft and auto theft) and person crimes (murder, rape, armed robbery, simple robbery, and assault). Arrest totals dating back to 2010 can be accumulated from the MCC’s criminal justice accountability reports from 2013 and 2015 (linked above), while UCR crime totals from 2010 to 2014 can be acquired at NOLA.gov.
This comparison isn’t perfect, as not all person crimes identified in UCR would be classified as violent crimes in the MCC report. But it is the best available tool for the sake of this analysis and provides an interesting optic for an initial look at the issue.
UCR crime totals for both person and property crimes sub-categories between 2010 and 2014 are provided in the tables below. These tables also show how many arrests were made, a ratio between the number of arrests and UCR crimes report, and how that ratio has changed from year to year between 2010 and 2014.
Looking at the data this way highlights two important trends relative to this debate. First, as NOPD’s manpower decreases, so does its ability to make arrests. UCR crime totals for both person and property crimes have increased each year since 2010 while at the same time there have been decreases in the ratio of arrests per crime each year. If NOPD’s ability to make arrests was constant over time then one would expect a relatively steady (rather than decreasing) ratio of arrests per crime. That is not the case here.
Each murder, assault or robbery in New Orleans produced, on average, 22.18 percent fewer arrests in 2014 than that crime would have in 2010. Property crimes saw an even bigger percentage drop as an average property crime in 2014 produced 34.82 percent fewer arrests than a property crime in 2010.
The second trend illustrated by the above tables is that, as Harrison claimed, NOPD seems to be placing a focus on reducing violent person crimes. This is suggested by a smaller percentage decline in the ratio of arrests per person crime than the decline with property crimes.
Combining the clear manpower decline with an apparent focus on person crimes strongly suggests that NOPD is making strategic choices regarding its enforcement capabilities. Looking at changes in 911 calls for service data for two crime types — drug violations and curfew violations — between 2011 and present illuminates these choices even further.
Drug violations (both distribution and possession) are typically identified through undercover work, uniformed patrols and citizen complaints. Analyzing calls for service data showed that 98.5 percent of all drug violation incidents were listed as Report to Follow (RTF), suggesting that most drug violations result from police work rather than citizen complaints. By comparison, aggravated battery by shooting — a category that mostly comes from citizen complaints — produces RTF designations in only 60.9 percent of all calls for service as nearly 40 percent of calls are either duplicates, unfounded, etc.
NOPD changed its approach to policing drug violations sometime in the first half of 2014, according to a June 2014 article. Per the article, previously undercover police officers working in “street-level narcotics units” in each police district were merged into uniformed “general assignment” units tasked with patrolling the district. The above article cites several unnamed sources who claimed the change was a direct result of NOPD’s declining manpower.
Charting changes in drug violation calls for service between 2011 and mid-May 2015 (below) highlights the real world impact this strategic shift has had on NOPD’s work against the city’s drug problems.
This chart shows the annualized pace of citywide drug violations over a 180 day span. It is broken down with the pace from January 2011 through May 2014 in blue, and the pace from June 2014 (when the strategy shift was noted) until the end of June 2015 in red.
The pace of drug violation incidents in mid-May 2015 was 37.3 percent lower than it was at its peak in December 2013 before the pace rose slightly in June. Drug violations as a whole increased slightly from 2011 to 2012 and again from 2012 to 2013 before plummeting in 2014 and being on pace as of June 2015 to decline again.
Looking at the data this way shows clearly the impact of removing street-level narcotics units. It also helps illustrate how publicly available data can be used to assess police strategy.
Changes in trends for another crime further highlights the challenge posed by NOPD’s manpower decline.
While 98.5 percent of drug violation calls for service end in an incident report, 98.5 percent of curfew violations end in a Necessary Action Taken (NAT) designation. Like drug violations, the vast majority of these incidents appear to be generated by proactive police action rather than citizen complaints. Unlike drug violations, however, these incidents rarely end in arrests.
The changing pace of curfew violations from July 2011 to June is presented in the chart below. The data starts in July 2011 because, for some unknown reason, curfew violation calls for service are not listed in 2011’s data set until July.
There’s no getting around NOPD’s declining enforcement of the curfew over time. Curfew incidents are generated by officers on the street interacting with violators. This declining trend is further reinforced by the below table showing changes in the total number of curfew violations each year since 2011, with 2011 and 2015’s totals estimated from the available data.
If the estimates for 2011 and 2015 are correct, that would indicate a nearly 80 percent decline in curfew violation enforcement over the last four years. Fewer officers on the streets means less ability to enforce curfew violations, simple as that.
This post set out to determine whether a decline in arrests is the result of manpower issues or a deliberate policing strategy. It is impossible to escape the conclusion that, while there is evidence of a policy shift toward focusing on violent person crimes, the reduction in manpower seems to be a driving force behind NOPD’s changes.
This problem can perhaps best be seen in the below table. This table contains the number of incidents for five crimes driven by “proactive” policing between 2011 and late July 2015. These incidents are: curfew violations, suspicious person incidents, DWIs, drug violations and illegal gun violations. The percentage change marks how far each category has fallen from the previous year with the last row showing declines from 2011.
Fewer officers on the streets necessarily leads to choices in enforcement. Limited manpower has meant that crimes such as drug and curfew violations are receiving far less attention in mid-2015 than they were in 2011. Although NOPD is clearly employing smart policing strategies to reduce the number of unnecessary arrests from traffic incidents and out of state/parish warrants, the lowered manpower appears to be having a strong impact on the department’s ability to fight crime.
This post was previously published by The New Orleans Advocate.