Why all the fuzz about automation?
Next: Open Ministry/Municipality
This is the third post in a series of posts about public services. The previous ones introduced the concept of automation budgets as a method of crowdsourcing what public services to automate and how to test automations.
This post is more on the background story. Why automate at all and how does that change public services.
Schrödinger Benefits
Public sector makes a lot of discretionary decisions where each person’s unique conditions are considered. This sounds as a flexible way to ensure social security should there be unforeseen circumstances. However, the volumes for discretionary decisions are high, indicating that perhaps the conditions are not so rare.
A case for automation stems from the fact that people are notoriously bad at making just decisions. Studies done in US and Israel have shown that the parole boards for prisoners work so that just after lunch when the glucose levels are high and the judges feel good, they release 65% of cases but many hours later then they are tired and low on energy, practically no appeals go through. Boards try to eliminate bias by having multiple people on them but in this case, it does not help as they all eat at same time.
A drastic solution would be by having each decision maker wear a mandatory ankle or wrist brace measuring glucose levels and ensuring snack breaks are taken. Unlikely to happen.
Since humans are the same everywhere, it can be expected that all human decision making has the same flavour.
Same happens in large private companies. When the portfolio board makes decisions, the first ones to discuss are always the big mainstream products bringing in the money. Everyone is focused, full of energy. At the end of the 4-hour session when there is no oxygen left in the room and new stuff comes up for discussion, in practice only two options remain. Either it is postponed to next session or rejected. For postponed minor proposals the loop repeats at the next meeting.
Another major reason for discrimination is the complexity of legislation and the language used. People without legal education find it hard to comprehend. Legislation is also opt-in. You have to ask for your rights.
The people who are uneducated, old, tired or sick have the least energy to push through and demand their services. The system favours people who have relatives inside the system and know the right liturgic expressions for fast processing or who are young and fit with time and energy at their hands.
With automation rights become opt-out.
It is possible that some politicians know this all too well and laws are passed knowing that most beneficiaries will never know or get them. They are dead benefits or Schrödinger Benefits – they are there but in a way that only chosen have access.
As another example see for example Tobias et. Al. “Quantifying the Cost of Decision Fatigue: Suboptimal Risk Decisions in Finance” where they write:
” Making decisions over extended periods of time is cognitively taxing and can lead to decision fatigue, which is linked to a preference for the ‘default’ option, namely whatever decision involves the least cognitive effort. Such effects have been demonstrated across a number of applied settings, including forensic and clinical contexts. An open question, however, is whether this necessarily leads to worse decision outcomes. Using 26,501 credit restructuring applications evaluated by credit officers of a major bank, here we show that in this real-life financial risk taking context credit loan approvals across the course of a day decreased during midday compared to early or later in the workday, reflecting a preference for the default option”
Automation means equality.
Digitalisation changes the nature of services
Automation brings many other benefits, not just equality. Digital public services will follow the same logic as any on-line services:
On-demand self-service. Being able to use digital services whenever you need 24*7.
Many ways to use the services – via apps, web, voice interfaces or just automated based on my status (these are called multi-modal services).
Being able to access services from anywhere. I may be on holiday and still use services when the need arises.
The services get better the more people use them. Digital services learn what works and what does not. The services get better and better all the time based on past feedback.
Personalised services. This is common on the commercial side since years and will be standard on automated public services as well. Services will over time get better and better at helping me, not a broad category.
Being digitalised means that other features common on the cloud also apply
If there suddenly are a lot of request for a specific service (for example due to recent news about a health issue), it can scale and everyone gets service
Same technical hardware and platforms are used for multitude of services which means that since demand varies between times, resources get used much more efficiently. There is less idle time in the server room where automated applications run.
Comparison between current model and automation budget
The Citizen’s and Enterprise (Automation) Budget concept are quite different from how public decisions are made now.
Today representatives make decisions based on proposals prepared by administration. Representatives have the right to make own proposals but if you are from an opposition party, this is merely performative as such proposals never go through. The coalition of parties that have the majority decide through negotiations a bunch of compromises that are tied together. On national level few party leaders in practice agree with each other consulting leaders of parliament groups, sometimes not even that.
Money in budgets needs to usually be spent that year.
When participants have direct control over their tax money, several ideas can proceed in parallel and the money does not need to be spent that year. A group of participants (citizens, companies) behind a proposal can form a short-term association or company to drive it. This organisation can invest the collected money and start a multi-year project for planning, negotiations with suppliers etc. As example, a small group of people in rural area can save and be savvy enough to get their road fixed that would never be important enough to warrant the great leaders’ attention.
One could call this opt-in democracy.
Next week we will dive into additional models for making public decision making more participant driven. Stay tuned.