Digital Anti-Administrativism vs. Incremental Tech Modernization in the Public Sector - Scott Timcke

Public sector tech modernization isn’t simply a matter of swapping old systems for new ones. It’s more like performing heart surgery while the patient is running a marathon. Decades of established processes, intricate regulatory frameworks, and mission-critical systems that each day serve millions of citizens create complexity. Notwithstanding their flaws, these evolved systems keep our public services functioning.

Yet there’s a growing chorus of voices who dismiss slow incremental changes to this complexity as mere bureaucratic resistance or institutional laziness; government systems are outdated obstacles to growth, they say. Rip and replace. Embrace innovation. Disrupt. Code around damage. (And government is viewed as damage). These are the next generation of the ‘move fast and break things’ crew. I call these voices ‘digital anti-administrativists’. 

The digital anti-administrativism views are sophomoric because they think about products, not final purposes. Bear with me while I emulate Adam Tooze with my example of benign but complicated tech modernization.

The XM7

Five years ago, the U.S. Army initiated a program to replace the M4 carbine, a weapon which has been integral to their equipment and infantry doctrine since the 1960s. With advancements in body armor and the anticipation of near-peer conflicts, the U.S. Army determined that it was necessary to increase their rifle caliber. This led to the selection of the XM7 rifle, which units began fielding this year. 

The XM7 introduces changes to infantry operations. The rifle itself weighs nearly 1 kg more than its predecessor, and due to the heavier ammunition, each soldier now carries an additional 2 kg of weight while having 70 fewer rounds available. This trade-off has implications for soldiers. The increased bulk means they must have more endurance to carry the load, and must become better marksmen under this strain to effectively use their reduced ammunition.

A stronger cartridge has other consequences. The XM7 has more recoil when fired. The U.S. Army has over half a century of experience training 18-year-olds to use the M4 and its variants, which is much easier to control. When weapons are changed, much of this training doctrine and tacit knowledge needs to be adjusted. And so it takes longer to train a recruit to use the XM7 as effectively as the M4. The required marksmanship training has knock-on effects like whether basic training should be longer to accommodate this time on the range, or whether to pair down other training components to fit the existing timeframe. If basic training takes longer, then fewer recruits can be trained per year and so on. 

You can see where this is going. Knowledge, systems and skills must be reorganized, priorities set, costs and benefits weighed. 

Supply and Logistics 

The transition affects other infrastructure systems throughout the U.S. Army. Shooting ranges were built with the ballistics of the M4’s 5.56mm round in mind. Military facilities are equipped with racks specifically designed to hold M4s. Warehouses are optimized for the efficient transport of 5.56mm ammunition. This includes specialized crates, loading bays, and truck beds. As they were built for one kind of ammunition, so armories must be inspected, recertified, and rebuilt where needed.

The implementation raises other logistical questions. How hard is it to clean the XM7? How easy is it to replace parts? Are the manufacturers reliable and financially sound? Is there a long supply chain that could be disrupted, and how much of that supply chain is vulnerable? Are there sufficient weapon-smithing facilities? Beyond weaponsmiths, there are inventory clerks and programmers who can update ordering and inventory databases. Throughout there are financial compliance and audit requirements.

If a hegemonic military power with extensive resources faces such challenges in implementing a relatively straightforward weapons system change, this should temper any anti-administrativism expectations about using AI and machine learning in public sector service delivery. More so in resource constrained environments, like the African public sector.

Different Tolerances Towards System Failure

There is another point worth making. When the public mission is important, it is irresponsible to ‘rip and replace’. There are bigger issues to address in conjunction, if not well before any tech modernization.

As Iginio Gagliardone keeps saying, most high-tech solutions fail. There are too many variables to cover here. But because failure is so common, some public sector resistance to innovation is rational governance. And this is before one considers path dependency, infrastructural lock-in, and complex purposes where numerous interconnected components influence technological modernization. And so, being blasé about failure is both unreasonable and irresponsible. 

In the public sector tech systems are typically customized. Once a project pathway is established, attention shifts to integration with existing systems, scalability, maintenance, and staff training. This leads to a preference for proven technologies that comparatively minimize cost overruns. The process prioritizes continuity and aims to preserve institutional knowledge, reflecting the public sector’s unique responsibility to maintain stable, reliable services for all.

In contrast, private firms in competitive marketplaces have clear imperatives to innovate. As Marx wrote, they must ‘accumulate or be accumulated,’ with profits serving as rewards for risk-taking. When private firms fail, which they do fairly frequently, the damage is typically contained — shareholders bear the losses, physical assets can be repurposed, workers can find new employment, and customers have alternative options.

Any failure in public sector tech systems has far broader, open-ended implications, potentially affecting millions of citizens. This fundamental difference in scope, scale, and stakes, is often overlooked by anti-administrativism. 

Public sector tech modernization is incremental by design. Thankfully so. Long may that continue.