It is obvious that this change is due to management policies at the NYPD. My observation is that when the NYPD increases the number of stop & frisks in a year the change in Not Innocent often drops. In years following, the NYPD corrects their "quality control" to increase the percent of Not Innocent results.
If you recall, "Not Innocent" does not mean that someone is found "Guilty." It means a suspect in a stop & frisk was arrested, given a summons or both. This "Not Innocent" result is the proof that the NYPD is effectively fighting crime with stop & frisks! Overall, the NYPD seems to be satisfied to have slightly less than 12% "Not Innocent" results versus 88% "Innocent" results from stop & frisks. (The 10 year average is actually 11.9% Not Innocent versus 88.1% innocent.) This works out to about 1 in every 8.3 to 8.4 stop & frisks has a "Not Innocent" result. This seems to be the NYPD's target.
Several people have claimed that during the Giuliani Administration crime dropped by a third. (One of these was Governor David Patterson on a local radio show, "Curtis & the Gov" AM 970, "The Answer.")
I came across a chart in a report which showed that the Giuliani Administration was doing about 55,000 stop & frisks, while the (early) Bloomberg Administration was doing slightly under 100,000 (around 2002.) I have not been able to find it again.
I also heard a show on WNYC (93.9 FM), This American Life, #414: "Right to Remain Silent", Stories about people who have the right to remain silent, but choose not to exercise that right—including police officer Adrian Schoolcraft, who secretly recorded his supervisors telling officers to manipulate crime statistics and make illegal arrests. On one hand it was positive in that crime has been reduced by about 75% since the 1990s. On the other hand it claimed the modern statistics may have be compromised by under reporting and misreporting. (A number of violent rapes, for example were reported as lesser crimes.) It also documents that the NYPDs approach to stop & frisk has been more quantity with little concern for quality.
You can listen here:
What do I mean by "quality" in a stop & frisk? "Quality" means the police stops a suspect for probable cause and arrests them, or gives them a summons, or both. Yes,it is obvious that some stop & frisks will be of people who are innocent, but quality control would reduce this.
I was looking at the summary data provided by the NYCLU and noticed that I had not used the summary data for 2002, because it did not have break downs by race. However, it really proves my point about quality versus quantity. In 2002 the NYPD did 97,296 stop & frisks, 80,167 people were deemed innocent, which means that 17,129 were Not Innocent (arrested, given a summons, or both.) This is the smallest number of stop & frisks in the series, with the highest percent of Not Innocent results, 17.6%. Only 82.4% were deemed innocent. This is to say that about 1 in every 5.6 stop & frisks resulted in "success"! This is the highest "success rate" in the eleven year series!
Here is my new, modified graph to display this:
In this graph, the "Not Innocent" result is the red columns at the bottom of the chart. These are the number of "successful" stop & frisks in that they result in an arrest, a summons or both. Putting these columns at the bottom helps us to see the values. The lowest value was in 2002 with 17,129 Not Innocent results. The highest was in 2010 with 82,963 Not Innocent results. That means that the maximum success was 4.84 times higher than the lowest.
The blue columns, sitting on top of the red columns, are the "Innocent" results. The low for the Innocents was also 2002 with 80,167 innocent people stopped. The high for innocent results is in 2011 with 605,328 innocent people stopped & frisked. From low to high, 7.55 times more innocent people were stopped & frisked in 2011 compared to 2002.
The two columns together show the total stop & frisks per year. Once again the low is in 2002 where 97,296total people were stopped and frisked. And the high is in 2011 where 685,724people in total were stopped & frisked. From low to high 7.05 times more people were stopped & frisked in 2011 than in 2002.
In summary, from low to high: 7.05 times more people were stopped & frisked and 7.55 times more innocent people were harassed to identify 4.84 times more "bad guys." This is just plain bad management! The pursuit of quantity over quality!
Graphed on the Right Hand Scale: Percents of Not Innocent Results by Year.
The purple line at the top of the chart shows the percent of Not Innocent results by year. The high is in 2002 with 17.6% of stop & frisks resulting in a Not Innocent result. That is about 1 in every 5.6 stop & frisks resulted in "success"! The low percent is in 2006 where only 9.7% of stop & frisks resulted in "success." That is 1 in every 10.3 stop & frisks results in "success." The next two highest years are 2010 with 13.8% and 2007 with 13.0%. The 2007 result is interesting since it follows the series low in 2006. This implies that NYPD management was doing something to improve the quality along with increasing quantity that year.
The light green line depicts the change in Not Innocent percent. The Greek letter delta (Δ) in "Δ Not Innocent %" is a math symbol meaning "change in." The green line starts at zero percent, because the year 2002 is the beginning, unity year. The percent shown for every other year is relative to 2002. In all cases the percents for the other years are lower than 2002. Notice that as the number of total stop & frisks increased in 2003, 2004, 2005 and 2006 the change in Not Innocent percent plummeted! In 2006 there were 506,491 total stop & frisks, 5.2 time more stop & frisks tan in 2002, but they only got about 2.9 times more "bad guy"s. ( 49,328 for 2006 versus 17,129 for 2002.) The results for 2006 were a 44.7% decrease in the percentage, from 2002 to 2006! (17.6% to 9.7%.) If the NYPD had gotten the same percentage "bad guys" in 2006 as in 2002, they would have gotten about 89,000 Not Innocent results instead of 49,000! It is apparent that the NYPD noticed this disparity and made efforts to correct it in 2007. In 2007 the NYPD decreased the number of stop & frisks to 472,096 and increased the numbers of Not Innocent results to 61,160. This is the third highest Not Innocent percent in the series (13.0%) and implies that the NYPD made some efforts to improve quality of stop & frisks that in 2007.
.Notice that the green line drops slightly (indicating the NYPD had slacked off quality controls a bit) then jumps in 2010. In each of the years 2008, 2009 and 2010 the total numbers of stop & frisks increases steadily from 472,096 in 2007 to 581,168 in 2010. In 2010 the NYPD must have introduced some quality control efforts to increase the percentage of Not Innocent results to 13.8%, the second highest percentage rate in the series.
In 2011 the NYPD pursued quantity over quality and their "success" rates plunged again. 2011 is the 11 year peak in total stop & frisks, 685,724 but with a slight decrease in Not Innocent results from 82,963 to 80,396. Thus we see the purple line (percent of Not Innocent results) declines slightly and the green line (change in percent of Not Innocent results) plunges downward. In fairness to the NYPD, the Occupy Wall Street protesters were active from mid-September until the end of the year. The 2011 results may have been impacted by the protests.
In 2012, the last year in our series, the total number of stop & frisks decreased and the NYPD's "success rate" decreased too. 2012 has the lowest number of stop & frisks since 2008 ( 533,042) and a lower number of Not Innocent results than 2007 ( 59,742.) This is the second lowest percentage of Not Innocent results (11.2%) in the 11 year series. The green line again plunges downward.
There are underlying anomalies in the data, based on policy decision of the NYPD, who mostly pursued increased quantity of stop and frisks and only occasionally quality. The year to year data by race is influenced by these underlying policy changes. We must keep in mind when examining the data for each race that:
- There was a steady increase in quantity from 2003 through 2006 with little regard to quality.
- In 2007 quantity was decreased slightly and quality was increased considerably.
- In 2008, 2009 & 2010 there was a steady increase in quantity and in 2010 a definite increase in quality too.
- In 2011 there was a major increase of quantity with little regard to quality.
- In 2012 there was a decrease in both quantity and quality.