Open Data As An Open Letter (A Time and Place for Hopeful Maps)
Introduction
If you search for the phrase “data humanism” most browsers will provide you with over seven million hits. One of the hundreds of interesting links leads to a luminary and human-centered data-visualizer named Giorgia Lupi, who says…
“Ultimately, data is human made.”
She has an…
“interest…in levelling understanding so that we citizens can have access to, and develop an imagination about, all the information that’s already being collected.”
Thirty years ago—long before “data humanism” populated the web or the current sociocultural zeitgeist and its ugly cries were evident— author and geographer Ann Buttimer posited that…
“Humanism is defined as the liberation cry of humanity, voiced at times and places where the integrity of life or thought was threatened or compromised, or when fresh horizons beckon.”
___________________________
Herewith I argue that open data — a phrase now nearly ubiquitous — is de-facto the most human kind of data. When overlapped, interrelated phrases — open data, data visualization, information design, infographics, urban cartography, etc. — are methods and tools that become a Venn diagram (< verso) and a goal by which analysts and readers can examine their town or city data in a transparent and humanistic fashion.
Backstory
At the end of August 2020 Mayor Melvin Carter and City Councilors started to debate the City of St. Paul (MN) 2021 Budget. The St. Paul Police Department (SPPD) released open data from that month on crime counts that were collected within city boundaries. Then, on November 18, Councilors voted to establish a “Community-First Public Safety Commission,” with the Mayor offering its objective as “realizing safer outcomes.” There certainly will be devils in the detail facing this Commission in 2021 with the key to success being expansive, humanistic, and clear goals.
Goal
The unfolding chronology mentioned above captured my attention as an urban cartographer, one living in, writing about, and mapping U.S. cities, especially St. Paul. My illustrated essay herewith becomes an open letter to members of the Commission the goal of which is two part: first, holistically realize safer outcomes by thoroughly understanding our Mayor’s outcomes, and second, in a detailed fashion, make visible — to the Commission, St. Paul residents, and readers of this essay — crime data by way of a “mini case study,” i.e., displaying thousands of arrest events that occurred during one particular month in 2020. This effort starts with what can happen when open data are made available to the public and ends with some specifics that will be a possible “pitch” to Commission members to refine a mission statement in the light of data humanism.
§ § §
___________________________
To The City of St. Paul (MN) Community-First Public Safety Commission
Dear Leaders:
Re: Commitment
The Mayor should be commended for most of the core goals he and his team have set for 2021. At the beginning of his August 2020 address Mayor Carter offered this:
“[A] centerpiece of our work moving forward will be the evolution of our Community-First Public Safety framework.”
“As we dig into the work that our police officers do on our behalf… it’s critical that we look into our data.”
Then — towards the end — he added this:
“We have committed ourselves in Saint Paul…to a public safety strategy that doesn’t just center around responding as quickly as possible after something bad is happening, but to a strategy that invests in proven interventions to help reduce the likelihood that those bad things will happen in our community in the first place… All of our work…will be guided by the deep public engagement that has marked our administration from day one.”
Visualizing these commitments clearly will require additional communication partners. I will use some of my professional processes and tools to that end, i.e., word clouds, data-maps, graphs, infographics, etc.
___________________________
“Words count,” as the saying goes. I created a word cloud (whereby frequent language is larger) from the Mayor’s address (< verso). Together with the excerpts above this word cloud suggests that he and his team looked to be very thorough.
Re: A case study of crime data as mapped
Visual and narrative matter reported in this case study is by way of five infographic-style broadsides (labelled hereafter as IG and numbered 1–5) akin to those that are found today in newspapers and online. The use of infographics can be a powerful means to champion urban decision-making, for a discussion of that see Jonathan Corum’s Storytelling with Data.
____________________________
IG1 features open data customization by way of a spreadsheet in order to organize, analyze, and visually summarize crime incidents in St.Paul.
IG1 below shows where I started my look at St. Paul public safety events made available as SPPD open data for August 2020. Source from which data was offloaded is here. Total (gross) crime count that month was 5,659 incidents, about 3,000 of which were labeled “proactive policing.” That left the net count at a still-hefty 2,691 incidents. I’ve used IG conventions to call out key details via captions. There is the terribly sad fact that August is several months after the killing of George P. Floyd, Jr. in Minneapolis, so St. Paul crime events of late summer 2020 will in some ways differ from those reported during previous years. For example, urban homicide counts were up nationally, but in St. Paul the count was down, from a rough average of three to one that August.
____________________________
IG2 is my analysis and first draft of multiple data-maps of the status of crimes per grid-zone last August.
The top row of IG2 below offers Commission members some context: each map displays key boundaries of St. Paul. Still more context is offered by a set of four maps on the second row (right side): they are from a meeting presentation as viewed by City Councilors on November 11, 2020. Link to council-meeting is here. At the time the SPPD deputy chief presented these maps and graphs as part of update on the status of ongoing Community-First Public Safety efforts. These and similar geovisualizations surely will be some kind of focus of the Commission. The bottom row offers three choropleth maps as the initial cartographic triptych of this case study, thefts, auto thefts, and narcotics events.
____________________________
IG3 describes some detective work with the support of traditional — as well as “novel” — visualizations.
This IG below suggests that Commission members look at St. Paul data visualization as a “cartographic detective” might. The focus here — initially by way of the top row of maps — is on assaults (all kinds, aggravated, domestic, etc.). The spatial geometry of data measured is what the SPPD refer to as their “grid,” a mesh of 200 zones draped atop everything that falls within city limits. The units are roughly one-half mile in size. Scale here is unusual by its high granularity: one has to give SPPD credit for due diligence as few metropolitan police departments in the U.S. seem to post open data at this level of detail. Top-right are maps that could be a worthy challenge to leaders who might be willing and able to put on a “Sherlock cap” and read the long caption about data clusters and outliers that I uncovered. This is a novel kind of map that displays statistical visualization known as localized spatial autocorrelation (acronym is LISA, see Luc Anselin’s LISA analysis for more information) and, at first glance, may seem more detailed than necessary. But two observations come to mind: (1) if one resides within St. Paul grid-zones troubled by violent crime, then one will have a keen interest in its very detail, i.e., by the frequency and location of assaults as shown (data becomes most interesting when one “makes it personal”); and (2) non-traditional visualization tools used here (and later) yield previously-unseen evidence of patterns such that new questions arise. Comparing and contrasting is what is required of another map in IG3, assaults, top-left, with a specialized demographic census map, bottom. One question immediately comes to the fore: what is the story these maps tell of the many red-tinted assaults within high-or-medium concentration of African-Americans who live in their own neighborhoods of St. Paul? Facts do not lie, so leaders and readers likely can intuit a provisional answer, one that gives the reader pause and perhaps may lead to some ethical discomfort.
____________________________
IG4 pertains to what happens if and when map data are spatially plotted: more patterns start to emerge.
This IG below is again a bit of a “Sherlock challenge” for leaders, one that takes a deeper dive into public safety data. These visualizations are called parallel-coordinate-plotting (see Robert Kosara’s Parallel Coordinates for details) that utilize another technique considered by analysts as visually and statistically innovative. Map A features a dozen and half grid-zones that are highest for six crime categories as revealed in detail by its companion — a “broken-line-plot” to its right. Likewise map B and C display several other public safety categories at risk (personal safety and auto theft). At the other end is Map D that shows dozens of low-crime grid-zones. Clustering of these grid-zones often is present, interesting since patterns seen can suggest emerging questions or hypotheses.
____________________________
IG5 transitions leaders and readers from Wilder Research literature to a cartographic framework I envision for a hopeful future that seeks to improve public safety in St. Paul.
At core this IG is an effort whereby I ponder how leaders and partners might transform these many still-tentative visualizations — ones that look and feel both cognitively and emotionally hurtful to residents — into a new kind of map that I’ll call hopeful. As previously suggested each might be as much a “hypothesis as map;” nevertheless, such a map could be a new beginning.
There exists a comprehensive view of sociological literature on the topic of violent crime authored by Lindsay Turner, a researcher at St. Paul-based Wilder Research. I summarized her findings as a one-page table below
The core column that I titled Action/s in the above table served as a window for her research summary that suggested to me possible and specific paths to new maps that help one understand crime. At first glance there looked to be at least four “candidates” for hopeful maps, one per each of the following numbered rows…
- By row 4: look at St. Paul residents who share a neighborhood but have unequal incomes, since literature states that that circumstance may precipitate higher rates of violent crime. (Then: act on implications.)
- By row 6: examine neighborhoods for possible high “population turnover” (e.g., say ones with homeless encampments) as the literature states that such turnover is known to increase risk of violence. St. Paul’s deputy police chief confirmed this in his November 2020 presentation, see IG3. (Then: pursue consequences.)
- By row 8: locate health clinics along with volunteer organizations in neighborhoods, then create spatial buffers around each: literature suggests they reduce crimes nearby. (Then: visit with clinic leaders and ask for their help.)
- By row 7: locate schools and colleges in neighborhoods, then visualize proximity by map buffers around each: as in the case of health clinics (as reported by Turner) schools also have a calming effect.
(Row 7 is the focus of the map featured in IG5.)
Action: greatly expand “after-school” programs, perhaps “weekend” programs as well, in high-crime zones. (Current St. Paul programs — Healing Streets, Sprockets, Rec Check — are useful but are not hitting the “on-site, after-school” mark.) IG5’s map displays high-crime grid-zones (there are five, each called-out by a black line) as best targets for such an initiative. One key detail: cartographic “buffers” (each one-quarter-mile in diameter, in blue) have, at their center, as few as one and as many as three schools per grid-zone.
Re: A pair of takeaway ideas
Proposing new ways of mapping open data pertaining to St. Paul public safety has great value but such an effort may require more resources. That’s where Commission members enter the picture, community leaders who can — and I argue should — help enact these best practices and human-centered strategies.
Data policy expert Jane Wiseman suggests that city leaders might assess themselves by examining the current status of their data management processes, see her published model below.
The above model leads to a question that the Commission might ask: How does St. Paul fare regarding its open data right now?
Another idea from Wiseman is community leaders exploring the creation of a position titled Chief Data Officer in order to help their constituents and partners travel down a better road to structuring and visualizing city data. Mayor Carter has floated a proposal to launch a brand-new “Office of Violence Prevention” in St. Paul, a potential umbrella under which I would think that a Chief Data Officer could be integrated.
Re: Conclusion
____________________________
Given a previous version of a hypothetical word cloud I now have another (< verso), this one marking with yellow highlight the attributes that appear most relevant from the perspective of yours truly, urban cartographer and vested St. Paul resident (I live in ward 2, grid-zone 131).
A key phrase — public safety — remains a huge challenge. Like any and all big-city leaders those in St. Paul are quintessential risk-managers and the SPPD will likely conduct various kinds of hyper-specific crime analytics, e.g. anticipatory, enforcement, predictive, etc. My final takeaway only remains salient by its brevity, morphing the Mayor’s word cloud back to a sentence: it is now offered as an eighteen-word addendum to a mission statement, below, one that I propose Commission members consider, with emphasis to be placed on analyzing and visualizing critical data with new tools as a means to increase public safety supported by “data-humanist” values:
The Commission was recently assigned a firm deadline (May 31, 2021) and now is reported to number 48 members. Such a great human resource! I conclude with this question: Shall we agree that 2021 is the time and St. Paul is the place to use open data to transform old maps into ones that are more novel, human-centered, and hopeful?
Re: Closure in the context of January 2021
In my introduction various excerpts were offered from two generations of knowledge-savvy “humanist pioneers” — Lupi and Buttimer — each a luminary in her respective field. I close with one more quote — right now as timely as can be — from Dr. Buttimer…
“The Renaissance of humanism calls for an ecumenical rather than a separatist spirit: it calls for excellence in special fields as well as a concern for the whole picture. It beckons sensitivity to what the “barbarism” of our own times might be. And challenges all to seek ways to heal or overcome that in responsible action…”
Further reading
Buttimer, A. 1990. “Geography, Humanism and Global Concern,” Annals of the Association of American Geography, 80 (1), pp. 1–33.
Lange, A. 2019. “Can Data Be Human? The Work of Giorgia Lupi,” The New Yorker, May 25.
Hutson, M. 2020. “The Trouble with Crime Statistics: It’s surprisingly hard to say what makes crime go up or down.” The New Yorker, January 9.
McClure, J. 2020. “Saint Paul reimagines delivery of its public safety services.” Villager, St. Paul, MN, December 9.
Tufte, E. 1990. Envisioning Information, Graphics Press, p. 67.
Acknowledgments
____________________________
This essay’s narrative matter and its illustrations — save Jane Wiseman’s figure — are assigned Creative Commons 2021.
LISA map language assistance courtesy Dr. Julia Koschinsky. Copy editing assistance courtesy Kathy Heuer and Brendan J. Byrne. Errors that might remain are mine and mine alone.
