Poor Alignment Between Tech and Politics

Jul 21, 2013

Ten thousand years of technological progress is enough to let us put a man on the moon. It’s not enough to prevent a public servant from deleting (or leaking, or collecting) digital data when they shouldn’t.

The latter feat may indeed be harder to achieve.

It’s a challenge precisely because government’s digital infrastructure is poorly aligned with the political system.

We have more than enough technology in government. Pervasive NSA surveillance wouldn’t be possible without up-to-the-second technology. Bureaucratic emails and records appear and disappear electronically, in Ontario and so many other places.

In some cases we need more technology. Senate expenses should always have been posted online, rather than recorded on clay tablets or conclave ballots.

Technology serves the executive branches of government, who actually “do”, rather than deliberative bodies, who “see that it is done”.

The robust security mechanisms of today’s computers are almost invariably hierarchical, so they meld naturally with bureaucracies. They leave core administrative control with individual executives, for good or ill.

Each appointed officer, executive, or manager is a single point of potential failure. Individuals misappropriate or lose documents, delete emails, metastasize programs, or process wonky expenses.

Of the 99 reasons a civil servant is permitted to delete an email, there’s only one that counts: because democracy lets them. Don’t ask why a civil servant deleted an email. Ask “Why did we make emails deletable?”.

The legislators who approve the supply of computing technology must retain central forms of technical control over it. Without it, parliamentarians must rely on “trust”, which is showing its age. Positive technical control can harmonize with democratic processes to discourage the worst excesses of technology-based scandals.

There is no complete technological analogue to the legislative body, yet. Many assemblies use electronic buttons for voting, or deploy voting machines for elections. These are just organizational devices, saving only labour and paper.

Numerous experiments in “Wiki Government” encouraged citizens to contribute to lawmaking. They offered citizens superficial control, but were fundamentally managed by bureaucrats and politicians. They usually ended quickly and with a whimper; that alone shows they were barely consultative.

True electronic democracy will require renovations at the lowest levels of hardware and software architecture, but it is doable. The goal is to align the most powerful technical privileges with the democratic mandate. A majority vote starts and stops the execution of otherwise restricted computing processes.

Encryption will play its less well known role of ensuring identity and non-repudiation. Digital signatures certify that email archives, access logs, and expense reports are unaltered.

But unless democratic bodies retain the power to get around encryption, individual functionaries will still be technical tyrants. It’s better that skilled technologists join procedure wonks and subject matter experts in offering testimony to parliamentarians about what computing structures to enact.

Only with this kind of architecture can citizens and representatives have fundamentally different expectations about their effective authority.

They can preserve the complete audit trail of government information independently of any number of saboteurs. The count or timing of votes, or the text of voted-on bills, can’t be changed without public debate.

They can allocate resources (including precious tax dollars) directly and publicly. In a networked world, the e-resolution that authorizes a particular payment can actually cause the transfer of funds to take place. The result is no more inconsistent executive discretion.

Realigning technical and political control can even help distribute the work of lawmaking more widely. When technical supremacy is in lawmakers’ hands, trust in public servants no longer needs to exceed the practical ability to verify their work.

Kirk Zurell is an industry member of REAP (Research Entrepreneurs Accelerating Prosperity) led by the University of Waterloo, and a graduate of Leadership Waterloo Region. His opinions are his own. Reach him at kirk@kirk.zurell.name.