First things first: 2016 was a bad year in terms of NOLA gun violence.

There have been 579 people shot in 479 shooting incidents with roughly 175 murders as of the time of this writing (December 29th). I have data on gunshot victims, incidents and murders beginning in 2010, and 2016’s count through December 28th is the 2nd most gunshot victims (587 YTD in 2011), the most shooting incidents and the 3rd most murders (198 and 189 respectively YTD in 2011 and 2012).

Unfortunately the historical record suggests 2017 could be an even worse year without a significant reduction in the amount of gunfire on the streets of New Orleans.

Forecasting murder is difficult and inherently imprecise. I do not know how much gunfire will occur next year, but let’s assume for the sake of argument that things will not get worse. In order to forecast what 2017 might look like, therefore, we must address two issues:

  • What will likely happen in terms of the city’s murder count if the amount of gun violence stays the same? And…
  • How likely is it that a significant drop in gun violence will occur?

There are a few ways to measure gun violence. Murder, as I’ve argued many times, is the least good way to measure fluctuations in gun violence over time due to all the randomness involved in whether a shooting is fatal. Shooting incidents is better, but even shooting incidents miss a fair amount of gun violence where there are no injuries or fatalities.

Indeed the majority of the time there are no injuries when there’s gunfire, and a person being fatally shot is even more rare. It’s impossible to say for certain how much gun violence we miss by only counting shootings, but firearm discharge incidents (signal 94) in NOPD’s Calls for Service database can help fill in the gaps.

These are incidents in which an individual calls 911 to report gunfire. Note that these spike artificially around January 1st and July 4th, so I have “smoothed” the day before and after each holiday by replacing the totals with the daily average of 9. This helps control for fireworks being reported as gunfire.

Going back to 2010 there have been 21,085 firearm discharge reports, 3,015 shooting incidents, and 1,136 people fatally shot through December 28th. That means that, on average, there should be 1 shooting incident for every 7 firearm discharge incidents, and there should be 1 person fatally shot for every 18.5 firearm discharge incidents.

This makes it possible to create metrics for an expected shooting total in a given year and an expected fatal victims total for that year based on the number of firearm discharge reports. Sometimes there are fewer than expected shootings and sometimes there are more, but over time things have a tendency to even out.

Consider the case of 2013, 2014 and 2015. Shootings swung wildly from year to year over those three years with 362 incidents in 2013, 432 in 2014 and 392 in 2015. Measuring firearm discharge reports, however, shows gun violence rose from 2013 to 2014 and was stable from 2014 to 2015 suggesting randomness was likely largely responsible for changes in shooting incident totals from 2014 to 2015.

Year Discharges Shootings Ratio
2013 2,555 362 7.06
2014 2,908 432 6.73
2015 2,881 392 7.35

In general when there are more shootings than expected in one year then there will be fewer than expected the next. When there are more fatal shooting victims than expected then the next year there will be fewer than expected. These can work in concert, such as in 2014 when there were more shooting incidents but fewer fatally shot victims than expected.

The expected shooting total can be derived by dividing the number of firearm discharge reports in a year by 7. This shows that 2013 was expected to produce 365 shootings, 2014 was expected to produce 415 shootings, and 2015 was expected to produce 412 shootings. Shootings dropped 10 percent and murder rose 10 percent from 2014 to 2015, but the amount of shootings and fatal shooting victims should have been relatively steady. New Orleans had more shootings but fewer murders than expected in 2014 and fewer shootings with more murders than expected in 2015.

Which brings us to 2016 and 2017.

2017 Forecast: Most Likely Fewer Shootings, More Murder

The first question when forecasting gun violence going forward is whether shootings and murder rose in 2016 because of bad luck or whether it reflects a genuine change in the level of violence.

Sadly it was almost certainly the latter.

Consider the below two charts which measure the rolling count of expected shootings over 180 days (below left) and expected fatal shooting victims over 180 days (bottom right). The expected totals are devised by dividing the number of firearm discharges by 7 and 18.5 to arrive at the expected totals for shootings and fatal victims respectively.

expected

The rise in 2016, particularly over the last six months, is stunning.

There were more shootings and fatal shootings over the last six months of 2016 because there was more gun violence in New Orleans not because of bad luck. In fact, the level of gun violence over the last six months of 2016 was higher than any other six month stretch on record (dating to 2010). The below table shows the number of firearm discharge reports, the total number of shooting incidents and fatal shooting victims, the expected totals for those counts, and the difference for each.

expected-table

While gun violence held relatively steady in 2014 and 2015 it undoubtedly rose in 2016. And the amount of gun violence rose considerably more than the rise in murder from 2015 to 2016. When shootings occur more often than expected one year they usually occur less often than expected the next. The same is generally true with fatal shooting victims and expectations as well.

The forecast for 2017, therefore, begins with the assumption that gun violence levels remain at this elevated level. At this level we would expect slightly fewer shootings in 2017 but more fatal shootings. This level of gun violence should have produced nearly 180 fatal shooting victims on average in 2016 which could have translated to around 190 or more murders.

Looking at it another way, average luck would have produced about 8 fewer shootings in 2016 and 13 more fatal shooting victims. Historically these things tend to even out, so 2017 is likely to see slightly fewer shootings than expected but more of them will be fatal.

Playing devil’s advocate, it’s possible that increased confidence in NOPD is leading people to call 911 more often when they hear gunfire artificially inflating the number of firearm discharge calls. While this seems like a possibility, it provides an unsatisfying explanation for why firearm discharge reports are skyrocketing at the exact same rate as actual shooting incidents. In addition, the level of satisfaction in NOPD as of the September 2016 was at 64 percent, only 4 percent higher than it was in 2014.

Alternatively, one could argue that medical techniques have improved since 2011 meaning more lives should be saved this year than just five years ago. Using firearm discharge data only back to 2013, however, suggests the city still should have had about 7 more fatally shot victims in 2016. Even if medical techniques have improved over the last five years it still seems more likely than not that a similar amount of gun violence in 2017 will lead to more fatal shooting victims.

There may be even more alternative explanations as well, but the expectation seems inescapable that without a significant reduction in gun violence murder will rise again in 2017 in New Orleans. After two straight years of rising murder that’s not an encouraging forecast.

Will Gun Violence Slow Down?

It’s impossible to forecast future shooting counts with certainty. Gun violence patterns are complicated, and it’s certainly plausible that the past year’s rise in gun violence will be an unexplained one-time blip. The recent history (going back to 2010) of New Orleans gun violence, however, suggests it’s difficult to change shooting patterns.

Gun violence has occurred at fairly stable intervals in New Orleans at essentially four tiers since 2010. The first tier took place from 2010 until early 2013. During this time the city averaged 450 shootings and nearly 190 murders per year. The second tier came about as NOLA for Life picked up and gun violence citywide fell an impressive 20 percent. Shootings dropped and murder fell with it though the effect was beginning to wear off in the first half of 2014. The third tier lasted from the first half of 2014 until early 2016 when the city averaged about 410 shootings per year. The fourth tier began around the middle of 2016 and lasted to today with gun violence rivaling and even exceeding the worst the city experienced from 2010 to 2012.

The below chart shows a rolling count of shootings over 365 days citywide with the four tiers in different colors. tiers

These tiers show relatively stable levels of gun violence in New Orleans over the last 7 years. The city averaged about 450 shootings over 365 days fairly consistently from 2010 to 2012. NOLA for Life had an impact in 2013 but the effect wore off a bit in 2014 with the city averaging about 415 shootings over 365 days until the middle of 2016.

The last six months have produced 273 shootings which, spread out over the course of a year would result in an astounding 546 shootings. The above discussion on expected versus actual shootings suggests there may be some drop in shootings in 2017, but there’s no reason to think there will be a large scale reduction absent some unforeseen intervention.

But there’s also no reason that gun violence cannot fall in 2017 though it would likely take a renewed effort from law enforcement and the city.

The rise in gun violence in 2016 could just be an unexplained blip rather than a long term trend. The problem is that gun violence patterns in New Orleans have been stable and stubborn to change over the last 7 years suggesting 2017’s most likely outcome will look similar to 2016.

Effective anti-gun violence initiatives have worked in the past, but it would require new initiatives and ideas that don’t appear to be in the works as of late 2016. NOLA for Life was effective because it took numerous gang members who were heavily involved in gun violence and the drug trade off the streets in a short period of time. This, combined with call-ins of unindicted gang members, appears to have had a considerable effect on gun violence levels citywide.

That momentum was unsustainable, however, and the effect wore off within just a few months. Little else over the last seven years appears to have had much of an impact on citywide gun violence, so replicating the concepts of focused deterrence used in NOLA for Life seems like a logical place to start.

Ultimately, absent a significant reduction in gun violence it appears likely that New Orleans will see another jump in murder in 2017.