This week:

Ireland’s General Election has finally come to an end. On my way from the polling station last Friday, I bumped into a (fellow nerd) friend. We got talking about Ireland’s unloved e-voting machines and the political saga that seemed to play out for years in Ireland.

In the end, the voting machines cost us over €50 million and they were eventually sold for scrap for the princely sum of €70,000.

So, this week, I try to answer the question: What went wrong?

To be honest, I struggled to limit this to only three answers. But here they are:

3 – Dodgy Database

2 – Dodgy Defences

1 – Dodgy Decision Making

 

 

(PS Thanks John Kennedy for sending me down this rabbit hole. I’ll never forgive you!).

References:

 

 

3 – Dodgy Database

The e-voting platform was built on top of a Microsoft Access database.

So what? When I started out as a software developer, I LOVED Microsoft Access. It’s like the Excel of the database world. And just like Excel, Access should be nowhere near a platform that is running a nation’s voting system.

So what? Building the system on top of Access is like building a skyscraper on top of sand. It will eventually fall over.

 

 

2 – Dodgy Defences

The machines were not encrypting the voting data but they were emitting specific radio waves during use.

So what? According to Dutch security researchers, if someone had physical access to the machine, they could adjust or replace the voting data with results they preferred. And if they were able to stand within a few meters of a voting machine (e.g. outside the walls of a polling station) with the right equipment, they could also see the votes of each individual because of radio waves emitted by the devices.

So what? So much for the security of the democratic voting process.

 

 

1 – Dodgy Decision Making

Trying to decipher when decisions were made and when costs were fully understood is something I’d never want to do again! But here are some key quotes from the CAG report:

  • “In 2000, the Department estimated a saving of €13 million over 20 years”. (Point 4.7, Page 17)
  • “In December 2001, after the tender, the estimated project cost was €31.7 million”. (Point 4.5, Page 16)

But here’s the best quote:

  • “However, the cost estimate at that stage did not impact on the project as the decision to proceed had already been made.” (Point 4.5, Page 16)

So what? As early as December 2001, it was clear that the up-front project costs would be at least twice the expected savings over a 20 year period.

So what? Politicians must love Back to the Future, because their motto is taken straight from that movie:

“Where we’re going, we don’t need a cost / benefit analysis“.