Digital Government and Data Theater

The Allegory in The Data

Published in

The Startup

10 min readFeb 28, 2020

Panoramic view of the Roman theatre in Palmyra. This photo was taken by Eusebius (Guillaume Piolle). — Panoramic view of the Roman theatre in Palmyra.

Governments are continuously taking on more and more automation strategies. This effort has been idealized, mythologized, at the city level, into the concept of the smart city, broken down further into terms such as smart buildings, and digital twins. It is difficult to explain what a smart city and related terms are even years after the terms were tossed around by big consulting companies as part of service offerings branded with the term to profit from the digital transformation of civic responsibility [10]. This corporatization is even moving directly into Urban Planning [8,9]. Among civic and built environment experts, you’ll find a complex set of often opposing answers about just what it meant, and what it means today. A common thread in the evolving notion of the smart city and digital twin is the expectation and hope for data driven digital enablement that solves some of our toughest challenges.

The reality however is that the disparate but interoperable functions of the digitally enabled built environment requires a holistic understanding by all those involved in the design of the built environment. The core challenge for us as architects, engineers, designers, planners, and active citizens is the requirement of continual refactoring of data driven design processes which define how we apply digital technology in constructs of our society. Society continually morphs over time and defies ontological edict as cultures and values shift. Unexpected and ephemeral shifts such as the potential for Covid-19 driven lockdowns continually challenge the resilience of cities, but also the ways in which they understand and utilize data.

5G represents a renewed value proposition for digital connectivity and the data fabric of society. The volume and velocity of data stored, processed, and analyzed in decision making will require new kinds of policy shaping. Our faith in our inconsistent understanding and interpretation [11] of data represents a risk to society which can cause real harm to entire segments of the population if the challenges at hand aren’t fully understood by the policy makers.

Since President Obama’s executive order for open government and transparency in 2009 and the creation of data.gov, we’ve seen an explosive expansion of data portals made available at all levels of government around the world [1]. What followed was the explosive proliferation of digital services using this data, as well as new services creating new types of data both public and private.

The landscape of Open Data, often touted by civic leaders as a measure of digital maturity and commitment to digitized transformation of governance, has grown very quickly [7]:

Cyganiak and Jentzsch, http://lod-cloud.net

Officials are becoming increasingly self-aware of data pitfalls, developing their data vocabulary to include healthy skepticism of data and its use. Even as far back as 2011, tech savvy city administrators were understanding the risk:

“I have been accused in this first couple months of promoting democracy by spreadsheet. The [Chicago open] data portal has been called a junk drawer. I think we need to think hard about it — the cynic says, get rid of instinct and gut feel in decision making and pair that with data. The more optimistic way to say it is there’s a street-level knowledge, a sort of narrative feel, that needs not to be totally displaced by a cyborg future of doing it by data. I don’t have answers for that. I think both of those things need to be in that. Ultimately, you fill a pothole with filler, not with data.”

- Former CTO John Tolva, Mayor’s Office, City of Chicago [2]

And yet, the risk remains. The Urban Institute’s study titled 2020 Census Who’s At Risk of Being Miscounted? [3] summarizes “Demographic changes over the past decade will make the population harder to count. And under-funding, under-tested process changes, and the last-minute introduction of a citizenship question could result in serious miscounts, potentially diminishing communities’ rightful political voice and share of funding.” The last part of diminished voice and share of funding is especially concerning for protected segments. Under-representation means that a percentage of any one person is being counted. We only need to look back into history a short distance to recall that human beings were counted as three fifths of a person for the color of their skin. These margins of error are exacerbated by institutional bias including race and status [4]. Allocation of funding from big infrastructure to items as simple as school lunch allocations have profound impacts on culture and the success of cities, neighborhoods and citizens. As cities race to digitize and renew, in the face of fierce global events such as Climate Gentrification, at-risk classes find increasing threats to their well-being [5].

This may all feel new. However, this risk is much older than the current fervor for digitization and data driven design. It may be simply human. There has always been data. Cave paintings are debated to be sign posts indicating and conveying available resources as information graphics displaying quantitative and qualitative data from one prehistoric human to another. In the article What the caves are Trying to Tell Us [6], Sam Kriss writes, “The birth of writing is always associated with magic, gods, and spirits: maybe the shamanism of the Paleolithic really worked; maybe these symbols were part of a global communications system that used the eternity of abstraction and the immortality of animal souls to bring all of humanity together, whispering across the underworld in a stone-age cybernetics.”

Cave painting at Serra da Capivara National Park, Brazil

What Kriss doesn’t mention however is that such cave paintings could have just as easily been purposeful misdirection. Whenever we encounter cybernetics, we find opportunities for hacking. Cybernetics is the science of communication and control of machine and living things, a term that is more and more appropriate for social media such as Twitter and Facebook, as exemplified in the manipulation of information to sway human behavior. A Paleolithic cyberpunk data hack of magic and illusion based in misdirection would appear like shamanism with the effect of a demotion of facts replaced by simple acceptance of a theater of someone else’s measure of reality. It seems Fake News could have been Paleolithic. Whatever their use, cave paintings, along with all information graphics, logos, semaphores, glyphs, hand gestures, facial expressions, language; Data in all its forms is fallible in its ability to deceive as well as be debated and ultimately misinterpreted.

At the center of all of this is a broad assumption of truthfulness of data. Veracity of data relies on many dimensions of concern and context. Veracity is not the same as accuracy. To understand the difference, a more detailed exploration is needed. We can use modern day graffiti as an example. Like cave paintings, we can glean information from the ubiquitous graffiti of the cavernous urban experience. Conclusions from the information however is a complex and contextual challenge. Although, like cave paintings, we could study the symbolism and iconography of various forms of graffiti, cities often use a simple graffiti count to measure the impact of graffiti and associated social implications. However, the way in which we count graffiti is as telling as the graffiti count itself.

*Chicago Open Data Portal Graffiti Count*

Chicago Open Data Portal Graffiti Incident Count By Year/Month

Graffiti is logged as data by most cities primarily by counting incidents when the graffiti is covered or removed. This is quantitative, and seemingly simple enough. A truck goes out, workers repaint a wall and move on to the next spot. Yet there are large swaths of cities where the graffiti is treated differently and even neighborhoods which invite graffiti as a spectacle to bring foot traffic to their area. Many neighborhoods deal with gang related graffiti that is safer to allow to go unreported out of fear of reprisal.

The graffiti incident counts available on many open data platforms then becomes a very different kind of data set, one which indicates economic prosperity of areas that can command the erasure of graffiti, rather than a measure of the actual graffiti activity in the city. This hints at a digital divide, ready for either exposure in the data or complete misinterpretation. Therefore, the count is not the measure of graffiti activity in the city, nor is it about the ability of the city to respond to incidents, it’s about the broader social and economic challenges that lead to graffiti that are more elusive to count.

In a case such as this, the data doesn’t even need to be false to be misleading. Interestingly, and perhaps unsurprisingly such a misinterpretation of thinking such data would give a meaningful indicator of graffiti activity is indeed bias driven. You might think such a conclusion from the data is biased towards wealthier or safer neighborhoods, but we’ve made a different mistake than simply a social and economic bias mistake.

Although one might conclude that economic bias is to blame when thinking about such a data example, the bias in this case is very different. This bias is Survivor Bias. Survivor Bias is simple: failures of data are ignored because the data of the failures is not collected. Only the surviving data is measured. In this case, only data of the graffiti that has been successfully cleaned away is collected. A misunderstanding of the complexity of data from the way it’s collected to its intended use has far reaching consequences in civic settings. Someone with the best of intentions may not understand the complexity of counting graffiti and use the data at face value to determine appropriation of funds and location of the workers and trucks to respond to graffiti reports. Those that need graffiti response the most may not be getting any at all.

Now apply this type of misunderstanding and risk of assumptions to more critical civic systems such as allocation of federally funded lunches for children where an uncounted daycare may translate into the miscount of subsidies required, ignoring people who literally need more help than others in simply being represented. In this case, the bias has very real implications and potential for harming people, enabling the digital divide to exacerbate more tangible social and economic divides.

As much as we may be at risk from Democracy by Spreadsheet, we must remain vigilant of data that has already been interpreted and displayed for specific goals in argument and evidence. The theater of Data, and especially Open Data in all forms of its life-cycle, represents unfettered opportunity for bias in the administration and planning of cities and constituent components.

The origin of Fake News may be much closer to home than most of us would like. As much as data may be used to fool, it is even easier for the many forms of bias to allow us to fool ourselves.

References

[1] Bill Schrier, Government Open Data: Benefits, Strategies, and Use, 2014 https://depts.washington.edu/esreview/wordpress/wp-content/uploads/2014/07/2014-Government-Open-Data.pdf

[2] Tolva, Code For America Speech, 2011 http://techpresident.com/short-post/quote-day-cities-cyborg-future

[3] 2020 Census Who’s At Risk of Being Miscounted? Urban Institute http://apps.urban.org/features/2020-census/

[4] Five reasons the addition of a citizenship question to the 2020 Census is detrimental to its goal https://www.urban.org/urban-wire/five-reasons-addition-citizenship-question-2020-census-detrimental-its-goal

[5] As Miami Faces Threats From Sea Level Rise, Some Worry About Climate Gentrification https://www.npr.org/2019/12/02/784225385/as-miami-faces-threats-from-sea-level-rise-some-worry-about-climate-gentrificati

[6] What the Caves are Trying to Tell us https://theoutline.com/post/2209/what-the-caves-are-trying-to-tell-us

[7] Cyganiak and Jentzsch, http://lod-cloud.net

[8] Google’s Sidewalk Lab smart city project threatens privacy and human rights: Amnesty Intl, CA says https://hub.packtpub.com/googles-sidewalk-lab-smart-city-project-threatens-privacy-and-human-rights-amnesty-intl-ca-says/

[9] How Ann Arbor, Michigan Became a Living Lab for City Mobility, Ford Motor Company, https://medium.com/cityoftomorrow/how-ann-arbor-michigan-became-a-living-lab-for-city-mobility-daf93672b5d0

[10] Does “We The People” Include Corporations? https://www.americanbar.org/groups/crsj/publications/human_rights_magazine_home/we-the-people/we-the-people-corporations/

[11] The “False Consensus Effect”: An Egocentric Bias in Social Perception and Attribution Processes LEE Ross, DAVID GREENE, AND PAMELA HOUSE Stanford University https://web.archive.org/web/20150927120010/http://web.mit.edu/curhan/www/docs/Articles/biases/13_J_Experimental_Social_Psychology_279_%28Ross%29.pdf

Digital Government and Data Theater

The Allegory in The Data

References

Further Reading and Exploration

Written by Jeff Berg