In government reform, there are always more Death Stars

January 24, 2021 1:55 AM

I call this Death Star thinking. Many of us grew up profoundly influenced by Star Wars. What’s the lesson too many people take from that first and most influential episode in the franchise back in 1977? One incredibly well-placed shot into the thermal exhaust port and the entire apparatus of our oppression explodes spectacularly. All we really needed were the plans to the Death Star and a very talented fighter pilot guided by the truth (“the Force”.) Never mind that there are countless Death Stars ahead of us as the Imperial war (and the franchise) continues.

Jennifer Pahlka: Death Star Thinking and Government Reform.
posted by MartinWisse (26 comments total)

I thought this article was pretty great. I've had the same thought that the author articulated in the introduction: if you're an outsider to a complex, social system, then the answer to any question that begins "Can't you just..." is probably no. There should be a new Betteridge's law for that.
posted by Alex404 at 3:12 AM on January 24, 2021 [20 favorites]

I love the way this article takes the reader through the details of a complex system, showing that the rules laid down to govern its development were all (1) absolutely correct and sensible in isolation, and (2) inevitably responsible, when combined, for the bloated mess of the transport protocol.
posted by Harvey Kilobit at 3:17 AM on January 24, 2021 [5 favorites]

This is an important essay and well worth the read. So many current problems are due to strict requirements put in place to address past problems, which have subsequently become outdated. I see Death Star thinking as not only a Silicon Valley, or tech, or engineer’s disease issue. It pops up whenever a desire for big change is joined with (an often understandable) limited patience for the slow work of iterative improvement.

What I don’t see tha author mention is the need for a continuous effort to the unsexy work of reviewing old laws and standards and removing the cruft.
posted by meinvt at 5:11 AM on January 24, 2021 [5 favorites]

The SNAP and MACRA examples: very thought-provoking!
posted by brainwane at 5:26 AM on January 24, 2021 [1 favorite]

Thanks for posting this.

I think a large part of the issue is that engineered systems built from a predetermined list of requirements can work just fine when the users are expected to receive days of training on the system. However, that is no longer a world in which we live. Systems are expected to be "intuitive" and require little to no training to use.

I was watching a Computer History Museum historical video about SABRE (the airline reservation system designed in the 60s that is still in wide use in the travel industry) on YouTube the other day and it struck me when the narrator proudly said that it only took two days of training for reservations agents to become proficient in its use.

That was fine when large businesses and government both had a much larger proportional headcount than they do today because there were people whose sole job was to operate the computer systems. Now that it's just part of everyone's job "for efficiency" that kind of design is completely unworkable, yet people still insist on doing things that way.

I won't argue that either way is better in the abstract, but it's clear that iterative design with constant reference to user metrics is the only way to produce a large IT project meant for wide use successfully given the unstated requirement of having all organizations be in some sense "lean." No organization is going to spend the resources necessary to train large fractions of their workforce or employ people to do nothing but enter user data from paper forms or whatever else the general public can handle for themselves.

The culture of computing has changed, so software development has had to change as well. Not everyone has gotten the message quite yet.
posted by wierdo at 5:45 AM on January 24, 2021 [6 favorites]

meinvt, I think "find patterns in the barriers they face and working with the administrators of the program to address the operational and policy drivers of those barriers." to me is the author talking about (among other things) reviewing and changing regulations.
posted by brainwane at 5:53 AM on January 24, 2021 [2 favorites]

Lord of the Rings also has a single point of failure for the bad guys.
posted by Nancy Lebovitz at 6:12 AM on January 24, 2021 [9 favorites]

My wife talks about this as the “Rosa Parks problem” which is that in the popular narrative, racism’s death knell was sounded when one black woman sat down at the end of a long day. The popular narrative is not that Rosa Parks was one of a number of organizers working in concert and that their strategy included boycotts and civil disobedience and going to jail.

Death Star thinking goes hand in hand with the idea that one person can change the world. Which is implicitly meant as “one person acting by themself can change the world in historical ways.”
posted by gauche at 6:23 AM on January 24, 2021 [20 favorites]

I agree with gauche that this kind of thinking goes way beyond government technology projects. It shows up here in every U S. political thread when people start listing their One Weird Trick to fix the American system of government (get rid of the Senate! ranked choice voting! pack the Supreme Court! Medicare for all!) It also underlies our purity tests for politicians and impatience when they don't fix everything to our satisfaction immediately. Similarly, our search for quick fixes for personal change like weight loss, exercise, and productivity.

Real change is not one big act. Real change is small acts repeated many times.
posted by medusa at 6:52 AM on January 24, 2021 [30 favorites]

Lord of the Rings also has a single point of failure for the bad guys.

A ton of stories have this, movies in particular. It is a key element of the power fantasy that one good shot or act of will by one virtuous individual can undo a whole system of villainy involving multiple powerful actors. It is thrilling and compelling and ultimately not at all effective at advancing any kind of systemic change in the real world.

I actually suspect that some of the recent rise of fascism has to do with a generation of people, mostly men, having been given a mode of heroism that has no resemblance to reality.
posted by gauche at 6:58 AM on January 24, 2021 [20 favorites]

This reminded me of a quote I come back to often from Zen Mind, Beginner’s Mind:

After you have practiced for a while, you will realize that it is not possible to make rapid, extraordinary progress. Even though you try very hard, the progress you make is always little by little.

I fall really easily into the mental trap that if I just do one significant, heroic effort I can make sweeping changes to my meditation practice/weight/fitness/work projects. It just doesn’t work like that. Facing a long, incremental battle and seeing it through takes a very different mindset.

Translating this to politics is really difficult; it’s an arena where the players change frequently, and it’s so easy to sell the public on the heroic and quick solution to their problems. How do you keep everyone’s eye and resolve on a goal that’s decades away?
posted by brism at 7:23 AM on January 24, 2021 [7 favorites]

Real change is not one big act. Real change is small acts repeated many times.

I love this.
posted by gauche at 7:30 AM on January 24, 2021 [6 favorites]

Reminds me of how Ivy League Universities are promoted, that their students can "change the world". And indeed graduates of prestigious engineering schools often have these kinds of ideas.
posted by Monday, stony Monday at 8:45 AM on January 24, 2021 [1 favorite]

She sort of hints at this in the article but part of the reason for Death Star Thinking is VC culture and its obsession with elevator pitches and being able to articulate a vision for how your company will disrupt some industry if you "just" (INSERT PARADIGM SHIFTING TECHNOLOGY INSIGHT HERE).

The VC playbook is very well optimized for analyzing one piece in a complex chain of loosely coupled free market entities and seeing how an incremental improvement can fit the needs of some customers, create some new opportunities by meeting that need, and whether the company in question is ready to exploit those opportunities. It's less adept at working through broader systemic change or in driving players who are looking for some targeted outcome beyond making more money.

As a small case example there's the story of Alloy, the private company started by Reid Hoffman of LinkedIn and employing two of Pahlka's colleagues from the USDS (Mikey Dickerson and Haley Van Dyck) to help Democrats build out better voter databases. It didn't have nearly as many regulation/procurement challenges as a bunch of other CodeForAmerica projects, but it still failed based on the fact that it thought it could build a better mousetrap than the state Democratic parties who it needed to have as its customers and wound up alienating the very people that it depended on to be successful.

So much of public service work is building partnerships with folks that you might otherwise think of as competitors whereas so much of private industry is treating everyone as either being a customer (to be exploited), a supplier (to be exploited) or a competitor (to be crushed).
posted by bl1nk at 8:51 AM on January 24, 2021 [9 favorites]

Death Star thinking is also behind the corporate hiring of outside consultants, who fly in on the Powerpoint Falcon and blast their way to massively disruptive changes because obvs this was the problem. And then you have new problems. And the consultants have your money. (And there is usually a round of layoffs.) And then the consultants convince you to hire them again, and so on.
posted by chavenet at 9:09 AM on January 24, 2021 [12 favorites]

As noted political theorist Jason Mendoza put it: "Any time I had a problem, I threw a Molotov cocktail and, boom, I had a different problem."
posted by Huffy Puffy at 9:35 AM on January 24, 2021 [31 favorites]

I bet there are MeFites who have successfully made slow incremental progress in improving government-related processes and IT, and I would love to hear some of your stories!
posted by brainwane at 9:58 AM on January 24, 2021 [5 favorites]

Most of my work has been in the vein of supporting voter information and voter protection efforts for Democrats, and one example is using geolocation tools to make it easier to verify polling location info at scale. Another is building better ticketing systems to capture and track efforts at voter suppression.

But really the stories that I find genuinely inspiring are seeing what US Digital Response (which Pahlka also cofounded with other Code For America alums) accomplished this past year in their mission to help state governments build out tech for their COVID response. So much good stuff done in very short time by some really great volunteers.
posted by bl1nk at 10:31 AM on January 24, 2021 [6 favorites]

Oh, this is just about the kind of "government reform" that annoys software engineers, not actual material reform or restructuring to address material inequalities and injustices. Not ending imperialism just making Empire more efficient. "Can't you just build a better death star?"
posted by anazgnos at 10:31 AM on January 24, 2021 [2 favorites]

Brainwane. Yes, and OMG is it painful and slow and takes almost constant batting away of "new shiny thing" that leadership watched a sales pitch on... If we had more technical people come up thru the ranks into management, this would help.

As it is, pre-covid, we were a tiny tiny tiny overworked team with a flat structure (one layer of mgmt to agency leadership), but now due to federal covid funding we've quadrupled the size of the team and added 5 or 6 layers of mgmt and split the thing into a bunch of separate teams *who all still rely on ONE team for data input*. We had 2 brilliant experienced people who kept the thing running and had enough pull to shut down a lot of stupid ideas.

So the idiots believing that FHIR is going to save us are no longer 'water cooler-ing' with the people who could give them the standards smackdown about how it takes decades even with ONC, CDC, and CMSs best efforts to incentivize interop in health IT.

Props to orgs doing the dead slow work of health IT standards development (HL7 mostly), and the member orgs who fund and encourage their employees to actively participate, comment, and try to implement at connectathons. That shit is solid gold and those drinks, food, and casual commiseration are where real change happens.

Trying not to out myself IRL, but that's my POV from a corner of our frankeinsteined and duct taped health care IT world, mostly the teeny corner trying to actually use (coded, not human readable) data coming in across all EHRs (yea, I see you Epic walled garden, that doesn't count).
posted by esoteric things at 10:37 AM on January 24, 2021 [4 favorites]

Really interesting and clear examples that show the effects on every day users of these systems. As a total outsider to tech and government I always assumed layers of rules and laws impeded tech innovation and usability but it’s interesting to see it laid out to succinctly.

“This approach is incremental in nature, but creates a relentless machine for step-by-step change, in which improvements come far more quickly than by a massive process to rebuild systems entirely”

Government seems set up not to allow for incremental changes like this and I’m encouraged to see at least a few cases where it’s being done. I wonder how this can be encouraged more while also recognizing that legacy system integration are a real thing for even the most tech forward companies that have been around for a while, so government must be that x10.

“I frequently encounter public sector leaders who believe that, if only they could get the procurement right, a vendor could build them the right technology that will bring their agency or department into the 21st century.”

Like brism above, I also had the thought that this tracks to the way a lot of us think about changes in our personal and professional lives. Instead of trying out new skills, the idea that one has to start a whole new career or go to get a new degree. The propensity towards New Years resolutions—that one has to start over with a clean slate and be a new/better person. The idea that if I can’t do the “best” thing then I might as not do it at all. Is this a uniquely inpatient American perspective on life and the need for instant change or just part of the human condition?
posted by Bunglegirl at 10:58 AM on January 24, 2021 [4 favorites]

This article is brilliant and this thread is great, I can't wait to see what people say over the next few days. Thank you MartinWisse!

My job is pretty similar to Pahlka's discussion about the California SNAP program shell - though instead of writing new interfaces, I write user guides designed to bing-bang-boom a user through the system as quickly as possible. It would be nice if the system(s) were created with an eye towards the intuitiveness in interplay, but intuitiveness costs money, and why do that for your internal tech users (and not your customers, lol) when you can pay someone to write quick'n'dirties.

... all of that to give you context when I tell you I gasped out loud when I read:

Instead of trying to control or design or even understand systems, it is more important to design systems that participate as responsible, aware and robust elements of even more complex systems.

!!!!!
posted by snerson at 3:49 PM on January 24, 2021 [1 favorite]

Lord of the Rings also has a single point of failure for the bad guys.

Except in that story, when Frodo gets into position to defeat the bad guys in one dramatic blow, he fails.

It's the persistence on their long, slow journey and the forbearance shown to Gollum that matters, not a single exceptional moment of heroic effort.
posted by straight at 8:25 PM on January 24, 2021 [6 favorites]

I have similar misgivings about superhero movies because they promote the idea of a single champion to save us instead of us saving ourselves.
posted by hypnogogue at 7:00 AM on January 25, 2021 [1 favorite]

I think this is backwards - in every case they mention it is regulators making up regulations without thinking of unforeseen consequences, and my experience in interpreting government regulations is the same - that sometimes, the answer to a grievance is to do nothing until you understand exactly what the problem is, and even then the correct answer might be to make no laws.

Too often, government (at all levels) hears someone communicate a problem, and then they build laws and regulations, when sometimes the answer should be empathy but ultimately doing nothing about it, because laws have unforeseen consequences.

Another option would be to make better use of older, existing laws. They have many decades of thought, grievance, and testing behind them.

This is not to say that all new laws and regulations are bad, just that they need to do more upfront work before acting.
posted by The_Vegetables at 8:58 AM on January 25, 2021

Someone up thread expressed a concern about being outed as part of the governance apparatus in some capacity. i feel this. Not quite as much as i feel this article and discussion, though which is why i'm posting so long past the end of interest.

It also amuses me to think about posting this deep in the thread :D

The one shot exhaust port really makes sense to me as a metaphor for certain schools of thought on government policy. In particular, i enjoy watching efforts by politicians and politician-adjacent public servants to solve "red tape" by passing laws/regulations requiring 1 or 2 regulations to be removed for every new one created. Remove the Red Tape: There's no clear or easy way to determine, never mind communicate, which parts of a regulatory "system" are stale or harmful, so skip over understanding the black big ball of laws and rules and just blow it apart via it's hind end.

branewane asked about successful incremental change. I have some experience. i'm a few years away from being ready to speak to the whole experience in detail. But perhaps some vague generalities?

First, there are parts of any decently complex enterprise that are fundamental, irreplaceable, and boring. Start there. Don't hide what you're doing, just work next to things that no one wants to look at or be around for very long. Why look for the boring places in the governance framework? At least two reasons: one, someone up thread mentioned that standing in or near the limelight means fighting to keep the light where you need it. Two, there are very few things the government does in the regulatory space that have any significant lasting effect unless they operate gently over longer periods with consistent support (or benign neglect). If that sounds sinister, it's because it is: go fix the sinister things so they stop hurting people.

You need to be willing to work as hard as the next hard working person, because you will need to grind to get into position to be able to make a move towards/into the fundamental, irreplaceable, boring place. You have to prove yourself first (and then again and again and ...) otherwise no one will let you near the irreplaceable thing that gives them aching stress when they are forced to think about their ignorance of it. On the other hand, being used to grinding away at iterations is useful discipline for the kind of responsive-through-iteration policy development work the article advocates. So, ... win?

I'm a big fan of responsive-through-iteration, by the way. No one knows what their doing (love that catch phrase, FilmJoy). Approach each change in policy or process as a test and it will prepare you to immediately shut it down if it turns out harmful. There's no such thing as designing policy for unanticipated consequences. It's right there in the adjective.

Once you've dedicated yourself to proving you can be more reliably boring than everyone else, you'd better hope you were smart. Or at least insightful. Or lucky. Because in government, like in private enterprise, people usually don't realize what the foundation of their day to day actually is. (How much rebar is in the concrete in your basement?) If you judge/guess wrong, you will have spent a lot of effort maneuvering into position to slowly and carefully shape something that turns out be only needed by the government because it was popular at some point in time that everyone has forgotten by now. Then when you actually do change something that has a noticeable effect on people's work or lives, your likely to get reorganized into a part of the government that writes a lot of briefing notes and policy proposals.

But if you are next to the fundamentals, and you are insightful, you can calmly and gently create or recreate platforms that foster some themes in regulation and discourage others. Again, kind of operating out of sight. Maybe that's a good thing though. Operating in the dark will probably teach you how to listen carefully.
posted by Deep state creature at 8:13 PM on February 10, 2021 [2 favorites]