Any crime analysis done in the first half of the year really needs the disclaimer that “it’s early” and this analysis is no different. We are only 59 days into 2017 and a lot of the current trends could change over the next 10 months. That said, the available evidence suggests crime, driven by an increase in property crime and gun violence, is up considerably so far this year.

We can make that determination using my UCR Estimator. I’ve written about this before, but the UCR Estimator measures Calls for Service marked ‘RTF’ for 36 different crime types that likely feed into the city’s UCR counts. There’s a very strong correlation between Calls for Service and UCR totals, and this provides a running count much quicker than the city’s official data is produced.

The below graph shows the estimated annualized pace of UCR crimes in New Orleans for every day of the year since 2012. The city is currently on pace for 20,644 UCR Part I Calls for Service in 2017. That’d be good for a 10.5 percent rise in crime if realized and represents a roughly 7 percent rise from the end of February 2016 (note that there was an extra day in 2016).

ucr-by-day

The rise in crime is being driven by increases in just a few crimes. In fact, 18 of the 36 UCR crime types are up this year relative to YTD 2016 and 18 are even or down. The below table shows the top ten crime types in terms of increased crime in 2017. The top five (all property crimes) account for the entirety of the rise in crime this year though it’s hard to discount rises in shootings armed robberies, and assaults.

ucr-change

One possible explanation is the weather. This winter has been much warmer than normal in New Orleans culminating in an 82-degree Mardi Gras Day. There is some evidence that warmer weather produces more crime in New Orleans up to a certain point.

The below chart shows the average number of UCR Part I Calls for Service (Y-Axis) per temperature degree (X-Axis) between 2010 and 2016. There are more UCR crimes when the temperature is in the 70s than when it is in the 30s in New Orleans, but that effect stops around 79 degrees.

temperature-ucr

It has undoubtedly been a warm winter as shown in the below chart. I compared the average daily temperature for New Orleans (found here) with the observed temperature from Weather Underground.actual-temp

Using the trendline in the above scatter plot I measured the number of expected crimes per average temperature degree compared to the expected crime per observed temperature degree. The warmer weather adds about 60 more crimes than an average winter to date would have produced.

That’s about 1.8 percent of 2017 crimes to date. Last year’s winter, by comparison, was colder than average meaning that last year’s crime total may have been deflated by about 42 UCR Part I crimes. Add them together and it’s possible that about 3.3 percent of the total swing is due to warmer weather in 2017 and colder weather in 2016.

Another possible explanation for is NOPD taking an increasing number of crimes over the phone. The below chart shows the number of UCR Part I Calls for Service marked ‘Report to Follow’ (RTF) over 30 days that received a Code 0 priority meaning those incidents did not require an NOPD response.

Code 0.jpg

There was a large jump in Code 0 calls last August and that has persisted more or less throughout the start of 2017. Indeed, 45 percent of 2017 shoplifting incidents, 32 percent of theft incidents, and 28 percent of pickpocket incidents have been Code 0, suggesting it’s possible that crimes reported by phone are having an impact on crime totals. Of course only 1 percent of vehicle burglaries, 0.6 percent of simple burglaries, and none of the shootings, robberies or assaults were taken over the phone.

Finally, an  explanation for last year’s rise is definitely not behind this year’s rise: response times. Faster response times last year led to more crimes being reported (rather than marked ‘Gone on Arrival’ or GOA) which fueled the rise in crime in the first half of last year.

This year that is definitively not the case. The below chart shows the change in crime relative to YTD 2016 for UCR Part I crimes marked just RTF (red) and those marked RTF and GOA. If crime inflation from better response times were responsible for the rise in crime then we should see a higher percent change in red than in blue reflecting NOPD reaching a higher percentage of calls quicker this year. That’s not the case, however, as the two lines are marching in lockstep.

UCR Difference.jpg

Conclusion

One could build a narrative whereby easier crime reporting and warmer weather have led to a somewhat sizable increase in crime relative to YTD 2016. The former explanation is far more speculative than the latter though. The most likely explanation is that there has been more crime so far in 2017 than there was in the first few months of 2016 driven partly by warmer weather. Measuring the impact of other factors (such as reporting by phone) requires further research and evaluation.