Humans as Connective Tissue: The Hidden Costs of Government Digital Transformation
Human-centered design, by its very name, is centered around humans, but where does the dimension of labour fit into the human part of this discipline? With increasing promises of automaton through AI, in what ways are designers in the public service thinking about labour as a material of design, beyond an assessment of bureaucratic tasks? This is my contemplation on automation and the future role of labour in digital service delivery within the context of government digital transformation.
Fear & Apprehension
When looking at a process for redesign, I often ask myself: What is the ‘work’ being done? Put another way, what is the functional value of human labour in this process? In many task-oriented roles of public service delivery, the answer may be “very little”. Reviewing, checking, double-checking, assessing information. Many of these tasks, though not all, can easily be done by computers given the right technology. If this is true, then how do the politics of labour fit into the design of future-state services?
Executives rarely outline workforce reduction as a primary goal of digital transformation, and with good reason. Programs that no longer demonstrate the need for a specific number of full-time employees (FTEs) will see their future budgets cut. From an incentives perspective, this ensures that organizations don’t ruthlessly chase job cuts in the name of digital efficiency.
In the absence of an agenda seeking to cut jobs outright, the phrase I often hear is moving staff from “low-value work” to “high-value work”. This generally means eliminating the drudgery of tasks that computers could complete and focusing human effort on areas best suited for humans to do.
This explanation points to what low-value work is, but there is a real lack of clarity of what new ‘high-value’ work would entail.
For employees performing ‘low-value work,’ (one can question who determines what is low-value) there is a legitimate fear of their roles being automated, even if the organizational incentives are not strong. Employees may be apprehensive about being research subjects in an initiative that might result in their replacement. When designers ask, “Why are you doing this?”, they are often seeking to understand user needs and challenges. However, this can be a threatening question for employees, representing concerns about livelihood, value, and worth.
But is this shift from low to high value work happening or is the nature of the work changing in response to new digital systems?
Humans as Connective Tissue in Digital Transformation
A signal of how work is changing as it relates to technology comes from Kate Crawford, in Atlas of AI, where she states: “Instead of asking whether robots will replace humans, I’m interested in how humans are increasingly treated like robots.” In Amazon warehouses, workers “are there to complete the specific, fiddly tasks that robots cannot: picking up and visually confirming all of the oddly shaped objects that people want delivered to their homes. Humans are the necessary connective tissue.” (Emphasis added by me)
While Crawford’s analysis focuses on labour practices in Amazon and other large organizations, parallels can be drawn to public service delivery amid disjointed digital transformation efforts where humans often act as the connective tissue in the digitization of services. This is not the result of maximizing the use of human intelligence, but more about mending the seams that are too complex or costly to automate.
What does this look like in practice?
- A SharePoint site becomes the new standard for records management: Staff need to upload the files coming from multiple source.
- A new public facing application form is delivered : Staff need to manually review the information and validate it against a different system from a different department.
- An enterprise financial tool is mandated for issuing payments : It’s sold by a vendor and there are no APIs. Staff need to copy and paste information.
- An Excel sheet keeps the proverbial lights on : Staff use this application to fill all the gaps where there is no ROI for purpose built software. And let’s be real, government is run on Excel.
This connective role exists not because of value creation but rather the systems that should work together, don’t. It’s a socio-technical system patched together through human labour managing, tracking, and moving information. Knowledge work increasingly resembles an assembly line, aspiring to implement the principles of a Tayloristic factory floor. Exactly as Crawford describes on the Amazon warehouse floors.
Arguably, the above is simply an outcome of poor design. But I find it increasingly rare to see a holistic service design delivered across products and processes for both citizens and staff where functions are integrated within and between departments, let alone automated. A relentless focus on citizen experience can sometimes mean a de-prioritization of improving the working conditions behind the scenes, where staff are left making the system work. And modernization efforts across a service delivered by different departments let alone external organizations don’t always line up.
Human labour becomes more and more a mechanism to make things work, rather than an input for additional value creation. And the longer a department cannot identify the high-value work their staff should be doing, the more they are at risk of digital systems slowly eating away at their head count.
Designing High-Value Work
Automation does pose a risk to eliminating some jobs. However, in a world of limited budgets, it also presents a great opportunity to rethink and reimagine the role of labour in government. Designers working in government digital transformation should not perceive end users as the connective tissue of digital systems. They must understand the evolving role of labour in digital service delivery while designing the opportunities for high value work in the future.
— — —
A note on the use of AI: This article was written by me with editing suggestions from Chat.GPT. The post’s header image was created using Midjourney.
