Benyamin Meschede-Krasa is DefundNPD’s Data Analyst
Institutional racism is built into the structures of policing across our country and Newton, Massachusetts is no outlier. In Newton, we currently rely solely on the police department to analyze and report policing and crime data and thus far they have ignored trends of racist policing. An array of documents containing information about the state of policing in Newton were recently made public when they were shared with the Newton Police Reform Task Force.
Nowhere in our analysis of the phase 1 documents did the police highlight racist patterns policing. Below we describe our analysis of one of these documents, the NPD field interrogations and observations.
Field interrogations are a major way the police interact with the public, either in response to a call or if an officer perceives suspicious behavior. Each field interrogation is documented by date/time, location, race, gender, age, whether it was officer initiated, and the reason for the interrogation. An example of the records for a field interrogation are shown in Table 1.
Date/Time | 9/12/2015 20:49 |
Location | ALBEMARLE RD & NORTH ST (NV) |
Race | W |
Gender | M |
Age | 75 |
Reason | Call for a male sleeping on a bench |
Officer Initiated | No |
Table 1. Sample field interrogation record
In total there were 210 field interrogations from September 2015 to May 2020. We grouped them by race/ethnicity and then normalized to the expected number of field interrogations based on the population of Newton using demographics data on the city website.
Racial/Ethnic Group | Count | Fraction of Interrogations | Proportion of Demographic | Normalized level of policing |
Asian | 6 | 0.029 | 0.130 | 0.22 |
Black | 44 | 0.21 | 0.035 | 5.99 |
Latinx | 12 | 0.057 | 0.049 | 1.17 |
Native American | 1 | 0.005 | 0.001 | 4.76 |
White | 138 | 0.657 | 0.803 | 0.82 |
unknown | 9 | 0.043 | 0.031 | 1.38 |
Table 2: All field interrogations by Race/Ethnicity. In total there were 210 field interrogations between September 2015 to May 2020. Column 1 contains counts of field interrogations by racial/ethnic group. Column 2 is the fraction of field interrogations for each demographic. Column 3 is the fraction of each demographic in all of Newton. Column 4 is a metric for the level of policing each demographic faces based on field interrogations, calculated by dividing column 2 by column 3. If field interrogations were being conducted propoionally the normalized level of policing would be equal to 1.
Table 2 shows the level of policing faced by major racial/ethnic groups in Newton. We calculated the Normalized Level of Policing for each racial/ethnic group by normalizing the proportion of field interrogations to racial/ethnic group demographics (Table 2). If a group’s fraction of field interrogations was equal to demographic proportion in Newton, the “Normalized Level of Policing” would be equal to one. Values above (below) 1 indicate a higher (lower) number of police interrogations than would be expected based on the population of the racial/ethnic group in Newton. We report that field interrogations are not proportional to the racial/ethnic makeup of Newton, with Black people experiencing 5.99 times, and Latinx people experiencing 1.17 times the expected number of field interrogations. White people, on the other hand, experience 0.82 times the expected number of field interrogations (Table 2).
Figure 1: The Normalized Level of Policing in Newton across racial/ethnic groups highlights higher than expected levels of policing of Black and Latinx people based on the population of those groups in Newton. The expected value of 1 is shown by the dotted line which represents a proportionally unbiased level of policing. See Table 2 for underlying data.
Compared to white people, Black people are 7.3 times and Latinx people are 1.42 times more likely to be interrogated by police.
Officer Volition Exacerbates Racism
When the same analysis is conducted for the subgroup of field interrogations that were initiated by officers, we observe higher levels of policing of Black and Latinx people. Field interrogations can be in response to a call or by an officer choosing to approach and interrogate someone. We explored officer initiated interrogations in order to assess whether the racist patterns observed in Figure 1 could be explained by a high incidence of racially motivated 911 calls. Table 3 describes the 74 officer initiated field interrogations.
Racial/Ethnic Group | Count | Fraction of Interrogations | Proportion of Demographic | Normalized level of policing |
Asian | 1 | 0.0135 | 0.130 | 0.10 |
Black | 18 | 0.243 | 0.035 | 6.94 |
Latinx | 7 | 0.095 | 0.049 | 1.93 |
Native American | 0 | 0.00 | 0.001 | 0.00 |
White | 46 | 0.622 | 0.803 | 0.77 |
unknown | 2 | 0.027 | 0.031 | 0.87 |
Table 3. Officer-initiated field interrogations by race/ethnicity. In total, there were 74 officer-initiated interrogations between September 2015 to May 2020. See Table 2 for a description of the methods. Expected value for “Normalized level of policing” is 1.
Table 3 demonstrates that police are initiating interrogations with Black and Latinx people at a higher rate than the average over all field interrogations calculated in Table 2. For officer-initiated interrogations, Black people are experiencing 6.94 times, and Latinx people experiencing 1.93 times the expected level of field interrogations. White people, on the other hand, experience 0.77 times the expected number of field interrogations (Figure 2)
As compared to white people, Black people are 9.01 times and Latinx people are 2.47 times more likely to be interrogated by police when police initiated the encounter. The increased level of policing of Black and Latinx people when officers initiate the interrogations highlights that the biased outcomes stem from racism embedded in NPD. This refutes the claim that racially biased 911 calls are the source of biased policing in Newton, a line used frequently by NPD during FY2021 budget deliberations.
Figure 2: Officer volition exacerbates already higher than expected levels of policing of Black and Latinx people in Newton. Plotted is the Normalized Level of Policing for racial/ethnic groups based on officer initiated field interrogations, which are higher than all field interrogations. The expected value of 1 is shown by the dotted line which represents a proportionally unbiased level of policing. See Table 3 for underlying data.
These field interrogation data should be qualified in two ways. First, we report normalized proportions of field interrogations based on demographics which assumes each interrogation to be a unique individual. If individuals had multiple encounters with the police and the distribution of those individuals did not align with racial biases found above, then our results could be an overcount or an undercount, depending on the race of the individuals having multiple encounters with police. NPD likely has more information about each field interrogation and may be able to compute the same statistics above, without counting individuals multiple times which would prevent this potential issue. We welcome their analysis of these data that we do not have access to. Second, data on Black and Latinx proportions are likely an undercount because there were interrogations with people of “unknown” race or ethnicity. In our predominantly white city, where whiteness is the norm, it is likely that those labeled as “unknown” are people of color.
The definition of what is “unbiased policing” an or “what is the expected number of field interrogations” can also be further explored beyond the definition given here. We assert that the Newton police should not be interrogating BIPOC more than they are interrogating white people, especially when officers are initiating the interrogations and we believe that is a reasonable expectation of the police force.
Inflated Police Spending
Data aggregated by activists at Defund NPD demonstrates that we are spending more than we need to on the police. We pooled data from Newton and 23 representative cities based on criteria of proximity to Boston, total population, and crime rates (Edwards, 2020). The police budgets were found on each city’s website and used to calculate police spending per capita (Table 4, Figure 3). Additionally, we collected data on the 2020 total crime rate to compare per capita spending and crime rates (Table 4, Figure 4)
Figure 3: Police budget per capita in comparable cities in Massachusetts. Newton is the 5th highest spender on police, per capita.
Newton has the highest overall police budget of these cities and is the 5th highest per capita spender (Table 4, Figure 3). By defunding to the median per capita police spending, we would have $4,069,136.08 to spend on life-affirming public safety initiatives. This represents almost double the 10% cut to police spending that DefundNPD called for in the last budget cycle.
Furthermore, per capita police spending is not significantly related to the total crime rate (Figure 4) which questions the assumption that giving police more money directly reduces crime rates. It also questions the assumption that diverting police funding to other public safety institutions would result in a spike in crime. This information urges us to shift the conversation to funding other initiatives that may be more directly impactful on holistic and equitable public safety.
Figure 4: Crime and police spending are not significantly related . Each data point is one of the above listed cities. The blue line is a linear regression with shaded 95% confidence intervals computed via bootstrap. The regression line with a slope of zero lies within the confidence interval, demonstrating that there is no significant relationship between the total crime rate and police spending in the 24 cities analyzed.
| Population | Police Budget ($) | Police Budget per capita ($ per person) | Total Crime Rate (per 1,000 residents) |
Shrewsbury, MA | 37,973 | 5,523,925.00 | 145.47 | 3.1 |
Brookline, MA | 59,234 | 9,011,040.00 | 152.13 | 11.7 |
Franklin, MA | 33,230 | 5,336,655.00 | 160.60 | 3.6 |
North Andover, MA | 31,296 | 5,253,719.00 | 167.87 | 8.1 |
Melrose, MA | 28,193 | 5,073,129.26 | 179.94 | 5.8 |
Beverly, MA | 42,312 | 7,781,690.00 | 183.91 | 7.4 |
Arlington, MA | 45,624 | 8,427,498.00 | 184.72 | 5.0 |
Billerica, MA | 43,784 | 8,495,213.00 | 194.03 | 4.8 |
Bridgewater, MA | 27,395 | 5,558,278.00 | 202.89 | 8.1 |
Andover, MA | 36,403 | 7,541,641.00 | 207.17 | 4.9 |
Reading, MA | 25,337 | 5,316,250.00 | 209.82 | 5.9 |
Natick, MA | 36,229 | 7,634,780.00 | 210.74 | 13.2 |
Framingham, MA | 72,032 | 15,224,815.00 | 211.36 | 14.84 |
Wakefield, MA | 27,135 | 5,947,868.00 | 219.20 | 8.1 |
Wellesley, MA | 29,673 | 6,553,273.00 | 220.85 | 7.1 |
Lexington, MA | 33,792 | 7,567,030.00 | 223.93 | 3.6 |
Needham, MA | 31,248 | 7,063,740.00 | 226.05 | 5.2 |
Somerville, MA | 80,434 | 18,242,995.00 | 226.81 | 14.8 |
Marshfield, MA | 25,905 | 5,954,122.00 | 229.84 | 4.5 |
Newton, MA | 88,904 | 22,804,754.00 | 256.51 | 7.8 |
Milton, MA | 27,616 | 7,439,070.00 | 269.38 | 2.7 |
Watertown, MA | 35,954 | 9,960,086.00 | 277.02 | 8.7 |
Belmont, MA | 26,330 | 7,472,519.00 | 283.80 | 6.6 |
Waltham, MA | 62,979 | 20,270,701.00 | 321.86 | 11.9 |
Table 4: Police budget, population, and per capita police spending in the 24 representative cities in Massachusetts. Newton is the 5th highest per capita spender and has the highest police budget overall.
Troubling Police Priorities
In our analysis of the phase 1 documents we found some troubling reports on the Newton police’s activities which raise questions which we hope will be further investigated by the Newton Police Reform Task Force.
Is NPD Policing Homelessness?
On page 4 of the NPD Accreditation Reports by Bureau, the Patrol Bureau reports on the Charles River Task Force which “hopes to deter and prevent crimes, and give people who use the recreation paths for walking and running a sense of security” from the homeless population living along the river. It further states that the police are working with the Department of Conservation and Recreation and have “cleared out trees and other obstacles that helped create a safer and cleaner environment,” presumably in order to limit areas where homeless people can safely sleep with some shelter and privacy. We submitted records requests to the police for reports or data collected on this work but were told that no such records exist. The lack of record keeping on the methods being used to police homelessness is troubling. Many questions remain unanswered:
- Why is the city of Newton addressing homelessness with policing and tree removal?
- How much money has been spent so far in these initiatives and how long has this initiative been in place?
- How is the public safety of homeless people being considered compared to the public safety of walkers and runners?
- What holistic initiatives has the city of Newton taken to address the needs of any homeless people in Newton?
- Are homeless people being criminalized for their survival?
Is NPD’s Directed Patrolling is Racist?
Page 4 of the NPD Accreditation Reports by Bureau also gives some detail on the directed patrolling being carried out by NPD. Directed patrolling seeks to predict areas where crime is likely to occur in order to focus police resources in those areas. This would work if crime could be unbiasedly measured but it is well known that nationally police disproportionately suspect Black and Brown people of committing crimes. Therefore, reported crime may not represent the reality of where crime is occuring. In 2019 there were 7500 more directed patrols than in 2018 (06 – NPD Accreditation Reports by Bureau) and directed patrols made up a majority of police department call logs. A records request for the code or software used for algorithmic prediction of crime were denied. In response to a records request for the variables used for algorithmic prediction of crime, Bureau Commander Steven R. Smith reported that “the basis of directed patrols is determined based on multiple data sets. These include motor vehicle accidents, crime events and special concerns” (Directed Patrolling Records Request). Considering the racially and ethnically biased field interrogations, modelling based on crime events is likely to introduce bias in directed patrolling. Additionally, the vaguely cited “special concerns” may open the door to human introduced biases. Many questions remain unanswered, some of which are
- What methods, models, or algorithms are being used for crime prediction and directed patrolling in Newton? Why are they not publicly available, even upon request?
- Have the methods used for directed patrolling in Newton ever been assessed for racial or ethnic bias by NPD or an external review?
- Are checks in place to limit human introduced bias into the architecture of crime prediction algorithms?
- Are the data used to fit crime prediction algorithms assessed for racial or eithnic bias?
- Have the outcomes of directed patrols been racially or ethnically biased?
Are Police Entering Students’ Homes, Unrelated to Policing?
According to the pages 63-64 of the NPD Accreditation Reports by Bureau, until 2019 police officers were conducting “residency checks” for Newton Public Schools (NPS), presumably to assess whether a NPS student actually lived in Newton. Residency checks included police entering and inspecting resident’s homes for matters unrelated to policing. This is a clear example of a job the police have taken on that is not related to the public safety of people in Newton. Police involvement in residency checks was discontinued due to concerns about officers entering homes to do the checks, but “it is expected [NPD] will once again do the checks in 2020, with a new modified version” (06 – NPD Accreditation Reports by Bureau, 2020). In the Community Services Bureau report for accreditation, no further information is given on how the checks were run, if there were complaints, or what modifications would be made for the residency checks to occur. We firmly believe the police have no role in questioning a student’s residency in Newton, especially by means of invasively entering their living spaces. These are just some questions left unanswered by NPD’s self-reporting:
- How long was NPD conducting residency checks until they were temporarily halted in 2019?
- How were students/families selected for a residency check?
- Were BIPOC disproportionately investigated?
- Were police armed during residency checks?
- Were there any complaints filed by those subjected to residency checks?
- Were police being paid overtime during residency checks?
- How much money has the police spent on residency checks, cumulatively and annually? What is the source of this funding?
- Did the police reinstitute residency checks in 2020?
- What modifications were introduced to justify reinstituting residency checks?
- Are residency checks occuring during the ongoing pandemic and with what precautions?
At the moment we do not have access to data that would elucidate the use of police resources in the three described areas. We hope the Newton Police Reform Task Force will bring light to these unanswered questions.
References:
06 – NPD Accreditation Reports by Bureau. (2020). http://www.newtonma.gov/civicax/filebank/documents/104966
07D – Field Interrogations and Observations. (Sept 2015 – May 2020). http://www.newtonma.gov/civicax/filebank/documents/104976
Directed Patrolling Records Request. (20 Aug. 2020). https://drive.google.com/file/d/1LIPy6l4TPjjPZE7gELOtJVQuzoDTOXB_/view?usp=sharing
Edwards, Rebecca. (5 Aug. 2020). Massachusetts’s 20 Safest Cities of 2020. SafeWise, www.safewise.com/blog/safest-cities-massachusetts/
Andover, MA FY2021 Budget. (2020). https://andoverma.gov/ArchiveCenter/ViewFile/Item/1324
Arlington, MA FY2021 Budget. (2020). https://www.arlingtonma.gov/home/showdocument?id=50783
Belmont, MA FY2021 Budget. (2020). https://www.belmont-ma.gov/sites/belmontma/files/uploads/fy20_expenditures.pdf
Beverly, MA FY 2021 Budget. (2020). http://www.beverlyma.gov/wp-content/uploads/2020/06/20200605110426755.pdf
Billerica, MA FY2021 Budget. (2020). https://www.town.billerica.ma.us/DocumentCenter/View/8348/FY2021-Budget-Bookpdf
Bridgewater, MA FY2020 Budget. (2020). Fhttps://www.bridgewaterma.org/DocumentCenter/View/2395/TOWN-MANAGER-FY2020-EXPENSE-AND-REVENUE-BUDGETS-040119–updated
Brookline, MA FY2020 Budget. (2020). https://www.brooklinema.gov/DocumentCenter/View/18642/Sec-IV_FY20-Deptl-Budgets
Framingham, MA FY2021 Budget. (2020). https://www.framinghamma.gov/DocumentCenter/View/37852/FY21-Budget-Summary-Sheet-5-18-2020
Franklin, MA FY2021 Budget. (2020). https://www.franklinma.gov/sites/g/files/vyhlif591/f/uploads/appendix_c1_historic_data_-_budget_breakdown.pdf
Lexington, MA FY2021 Budget. (2020). https://www.lexingtonma.gov/budget/pages/fiscal-year-2021-recommended-budget-financing-plan
Marshfield, MA FY2021 Budget. (2020). https://www.marshfield-ma.gov/sites/marshfieldma/files/uploads/marshfield_2019_tr-web_copy.pdf
Melrose, MA FY2021 Budget. (2020). http://melrosecityma.iqm2.com/Citizens/FileOpen.aspx?Type=4&ID=10029&MeetingID=4631
Milton, MA FY2018 Budget. (2018). https://www.townofmilton.org/sites/g/files/vyhlif911/f/news/2018_annual_report-milton_final_5.3.19.pdf
Natick, MA FY2021 Budget. (2020). https://www.natickma.gov/DocumentCenter/View/9955/FY-2021-Town-Administrators-Amended-Preliminary-Budget
Needham, MA F2021 Budget. (2020). https://www.needhamma.gov/DocumentCenter/View/20762/Section-3
Newton, MA FY2021 Budget. (2020). http://www.newtonma.gov/civicax/filebank/documents/103602
North Andover, MA FY2021 Budget. (2020). https://drive.google.com/file/d/1UJG90tQKJBwVqhg-GTeHqfBTimFRGMgL/view?usp=sharing
Reading, MA FY 2021 Budget. (2020). https://www.readingma.gov/sites/g/files/vyhlif1116/f/pages/fy20_town_manager_budget_0.pdf
Shrewsbury, MA FY2021 Budget (2020). https://shrewsburyma.gov/DocumentCenter/View/7163/FY-21-Financial-Plan—Fiscal-Projection-1—2020_02_14
Somerville, MA FY2020 Budget. (2020). https://drive.google.com/drive/folders/1QyVyvQk3OEqJqH5baWbYrFPxvu005Rdz
Wakefield, MA FY2021 Budget. (2020). https://www.wakefield.ma.us/sites/g/files/vyhlif3986/f/uploads/annual-appropriations-booklet-fy2021_0.pdf
Waltham, MA FY2021 Budget. (2020). https://www.city.waltham.ma.us/sites/g/files/vyhlif1386/f/uploads/fy2021_budget_-_council_approved.pdf
Watertown, MA FY2020 Budget. (2019). https://www.watertown-ma.gov/DocumentCenter/View/27036/Annual-Town-Budget-FY2020
Wellesley, MA FY2021 Budget. (2020). https://wellesleyma.gov/DocumentCenter/View/19038/53-Budget-2-FY2021-SOURCES-USES
This is extremely well done.
Figures are referenced in the text, but do not seem to be provided.
Firstly, kudos for the first order work with the available information. Yours is an essential task well done. In preparing for subsequent rounds of quantification, it may be insightful to see what isn’t seen in the figures above:
Silence In The Numbers, Part A – Otherness: My personal experience was that the perpetrators of crimes against our property were not owner-residents of Newton. They were not rental-residents of Newton. They were from someplace else.
To the extent my experience is Newton normal, the analyses above do not necessarily imply conclusions about racially motivated behaviors of either an officer initiating an interrogation or of a citizen whose report resulted in an interrogation. Other explanations would also fit the numbers and might better fit the external reality.
Silence In The Numbers, Part B – Mobility: My experience living elsewhere is that longer-term residents of a community are known by sight, known by name, and known by their past behaviors. Citizens more recently dwelling somewhere are not as readily recognizable. And the less well known draw more attention, at least in part because the law of “we know where you live” applies to son of the high school classmate of an observant officer. Provincial. Quaint. People.
To tease at the possibility of that in the numbers, one would compare the Normalized Level of Policing with similar figures in a statistically significant group of situationally-peer Cities. That is not a suggestion that we hunt to assuage ourselves that Newton has no less of a bias problem than the US on average. Rather, it is part of the diligence that residents and voters owe one another in coming to grips with the reality of the realities described.
Others likely have better hypotheses worth examining before concluding the first order insights ought to determine specific public policies. The good news is that the work has just begun ..
+1 on you
The Newton Police Reform Task Force will hold an open session on Wednesday, September 16 at 6:00 pm according to:
https://www.change.org/p/creating-true-diversity-on-the-newton-police-reform-taskforce/u/27689567
@David O Knuttunen – Figures has been added back in. Be sure to hit your browser Refresh buttons to force the updated page to be reloaded.
Defund NPD will be giving a presentation to the Newton Police Reform Taskforce on Wedesday, 9/16 6:00-6:45PM
More Information: http://www.newtonma.gov/gov/mayor/taskforce/default.asp
Zoom Link: https://us02web.zoom.us/j/86129354250?pwd=ek54V2lrTlF6MENDK1A0SnZNa09wdz09
In the mean time you can learn more about Defund NPD at our website: defundnewtonpd.com/
@Benyamin,
Forgive me if you already answered this, but did you request and receive the NPD practices and procedures manual? The number of field interrogations, as a criminal law appellate practitioner, appears to me to be surprisingly low. That makes me want to question what events actually trigger a report for a field interrogation.
@ Lisap
That is a good point! NPD has not made those document’s public but I will follow up with them. I also thought the numbers were low for almost a 5 year period, but I would be surprised if NPD withheld certain field interrogations from the data shared with the task force.
Black people are 9X more likely to be interrogated by Newton police of their own volition, versus 7.3 times for all interrogations, compared to interrogations of white people. And since the 7.3 figure apparently includes all interrogations, including those done by police of their own volition, the difference must be even greater.
This is shocking but not surprising. It’s consistent with recent high profile incidents: the Newton PD concluded from its investigation of the Tim Duncan incident that it was handled properly. Presumably, given the same set of circumstances, they’d handle it the same way; no lessons learned. The Newton PD has never apologized for its indefensible 2015 racial profiling of the Asim family, in which the Newton PD incredulously insisted a woman driver had a beard. And the Newton PD stood by passively while a white guy in a pickup truck drove his truck at youth peacefully protesting in front of City Hall. Finally, repeated allegations of racism within the Newton PD by a Black officer have been ignored for years.
Important work, Benyamin.
@Benyamin,
I similarly doubt that the reports are being withheld. Rather, I think it would be useful to know what events trigger the reporting requirement and what the NPD manual says about when a written report must be generated.
Benyamin,
Thanks for this work, it helps us get ahold of several important issues.
I agree with Lisap. The numbers of encounters are way low. I also agree with her that this isn’t likely about withholding, but perhaps misclassifying or something about triggering reports.
You cover a lot of ground here, and I think it is critically important that we use the data we have, and get the data we don’t have, to address problems and not necessarily try to fit any preconceived narratives.
Even with incomplete data, I think it is important for anyone looking for conclusions to actually look at the spreadsheet at the actual stated reasons for the calls:
http://www.newtonma.gov/civicax/filebank/documents/104976
Once you look at the data, it’s much harder to put a high level summary statistic on the results and feel confident you’ve found the root cause of a problem. Correlation is not causation, as they say.
You mention homelessness, for example. The interrogations and Observations report spells out many encounters involving homelessness. We can and should have a public discussion about homelessness in Newton, Metro West, and Boston Metro as it pertains to city policy. And not just discussion, action. But the police don’t set policy. They get the calls when the rest of the system has failed. I don’t think the main question should be, “is NPD policing homelessness?” but rather “Why does homelessness exist in Newton (or better, the United States) and what will we do about it?”. You ask, “Are homeless people being criminalized for their survival?”, but I would propose that we’ve failed as a society if we force anyone to the edge of survival.
The Pew Trust has called homelessness “A Pileup of Inequities”, and I think the term is quite apt. It involves issues that are both systemic and individual. Race, poverty, violence, inequity, mental and physical health, lack of opportunity, substance abuse, crime, lack of social safety net, and many more. It’s not easy to solve. It’s much easier to avoid. When we fail (and for me that includes camps along the Charles River or Methadone Mile) we force the police to “address” the festering social problems that remain.
On the topic of directed patrols, many of your questions involve the algorithms used to direct those patrols. I suspect Newton Police, or many departments, are data driven to the point that the word “algorithm” even makes sense. Nor is that an especially good goal. Algorithms can be as biased as people. People make good decisions as well as bad ones. I am certainly frustrated with NPD’s current production of data, specifically traffic data, that could help in city planning. But I think that’s a limitation, not necessarily a bias.
As for residency checks, I’d love to hear the history about how NPD got that job.
I guess here’s where I come down: to make any meaningful determinations about improving NPD, and beyond that Newton’s response to a wide range of social problems, we first need transparency. From NPD. From all social services as well. We all need to understand how they do what they do, in detail. Yearly reports. More data. Published policies and procedures. How things change when things go wrong and things go right. A culture of accountability and constant improvement, both in NPD and the City.
That’s the foundation on which you build a system you can trust, verify, and adapt to meet a wide range of social and public safety needs based on shared values, with NPD as a component.
Benyamin,
This is a great start on evaluating the global field activities of the NPD. Congrats.
1) As I read thru your analysis, I was wondering what percentage of people subject to field interrogations were confirmed Newton residents. If they were not City of Newton residents, then the “Proportion of Demographic” and “Normalized level of policing” are invalid? Perhaps, you should you use larger Boston metropolitan area population stats.
2) What are key variables in determining the budget of police department in similar large, wealthy suburban city (e.g. West Hartford, CT)? I would assume variables could include longevity of police officers’ careers, number of required patrol cars per mileage of street, number of alarmed homes, age of police buildings and scheduled improvements, technology upgrades, number of required overtime activities, (e.g. funerals, major college sports and activities, etc.) and officer training.
2) I wonder if the homeless living along river banks (whom I never seen in my 30+ years in Newton) is under the jurisdiction of City of Newton or state Department of Conservation and Recreation (therefore State Police).
@Mike Halle
Thank you for the thoughtful response.
I don’t quite follow your logic on the reason section of the field interrogations. I agree that the cause of these disparities can’t be concluded from the results here, but the outcomes are clear. In Newton, Black and Latinx people are bring interrogated at a much higher rate than white and Asian people. Proving causation is near impossible in public policy or policing, unless a natural controlled experiment happens. We can’t wait for that to happen to act in Newton. When 24.3% of officer initiated interrogations target Black people, who make up 3.5% of Newton’s population I feel confident calling it out for what it is: racism. An analysis of the reasons why white people versus Black and Latinx people are stopped could be enlightening. I hope this report serves are the beginning of a conversation and that policing data in Newton are further analyzed more transparently, outside of NPD.
On the issue of homelessness, I agree that the root source of homelessness should be the priority for where resources are being spent. That is why I believe we must address that survival behaviors of the homelessness are criminalized and any resources being spent on breaking down camps along the Charles or waking up people sleeping on benches is money not being spent on fixing the root of the problem. I bring up homelessness more to ask questions about what is actually happening, as the Charles River Task Force is named as a priority of the patrol bureau. We know that on the frontlines, homelessness is being addressed by police across the country, and if it is happening in Newton too we deserve to know. You say, “we force the police to ‘address’ the festering social problems” but I don’t understand why this is normalized or expected. Why wouldn’t HHS or Newton Housing & Community Development Division be responding to homelessness on the frontlines?
On directed patrolling, I fully agree that algorithms shouldn’t be the goal. My understanding from the linked NPD records is that reported crime is being used to predict where future crime may occur, which would require some sort of prediction algorithm. If there is one in place, it should be rigorously assessed for bias. As you mentioned, human bias can be introduced into algorithms in many different ways.
And on transparency I agree too. It’s clear that there are many questions not being asked and that policing data is not being rigorously assessed for biased outcomes. I believe every department should be looking for biased outcomes in their work. Across the country, structural racism rears its head in discipline in schools, policing, incarceration, support of minority-owned, and representation in elected officials. We would be naive to assume that Newton is insulated from it all.
Benyamin, I think the connection of racism to the number of police field interrogations and observations (FIOs) is far more complex than just “police targeting”. Putting all the blame on police lets all of us off far too easily.
I honestly wouldn’t expect FIOs to have the same racial breakdown as the rest of predominantly affluent Newton. Nor would I expect the percentage of convictions. Or victims of crime for that matter. Some of that is the police and more broadly the legal system. But we also know that many of the underlying roots of crime (poverty, lack of social support, substance abuse, housing insecurity, family instability) impact minority and vulnerable communities disproportionally. Societal conditions pave the way for the statistics you present long before the police talk to anyone. These conditions are the systemic ones that we as a nation have failed to address.
That isn’t to say police bias is or isn’t there, it’s just that an understanding of it is less straightforward than expected value analysis on a limited and likely incomplete number of incidents can reveal. What happens after the FIO is pretty important. Or what happens during “apples to apples” comparisons between different subjects of interest. Just as important is the flip side of policing: do all communities receive the same level of service from the police?
I specifically say “communities” rather than races because this is an issue beyond race. Every person should be treated with the same dignity, respect, and presumption of innocence by police and everyone else. Black, White, Latinx, any ethnic group, any nationality, any religion or no religion, any gender or sexual preference, rich, poor, old, young, any physical and mental ability or deficit. All of us, all equal and valued under the law, under our government, and within our society. Sure, it’s utopian, but it’s still an ideal that can guide us.