You guys actually have a much bigger responsibility than I think people would let on. I'll maybe share that more in closing.
As an illustration of that, I love what Clay Christensen, a famous author and Harvard Business School professor, says, that when you destroy the value in one part of the value chain, it migrates to another. If you make software free, the value doesn't disappear; the value shifts over to the services, and now servicing the software is where the money will be located.
I think there is something similar around politics. When you knock the politics out of one part of the political chain, you don't destroy the politics, it just migrates to a different part. This is one place around data where you have an enormous responsibility. In a world of open data, you can presume that the information or data that government is creating will be made public. That used to be a political decision. It used to be a decision where a minister could say whether or not they'd share this or that data. But if we now presume that all data will be made public, we've now taken the politics out of that part of the chain. That doesn't mean the politics disappears; it just moves to a different part of the chain.
So I think one of the questions politicians will increasingly be asking themselves is, “If the data gets created, if that means it will be shared, I will have a lot more scrutiny over what data will get created in the first place.” Some people would argue that this is what happened around the long-form census, that we actually didn't want to have the data created in the first place because that meant there would be questions asked that government didn't want to be asked, or policies pursued that people didn't want to get pursued.
The data will become more political the more open it becomes. This committee needs to think about what the ramifications of that are.
I would follow it on to say that I would be very careful about presuming that data will lead to better decision-making. Having data-driven decisions does not mean a better decision. I only need to show you a map of a congressional district in Chicago, that looks like a tiny little filament running through nine different neighbourhoods, that makes absolutely zero sense. The reason that congressional district was created was to produce a very specific outcome.
That was a data-driven outcome; I want to be clear. You could never create that congressional district unless you had phenomenal data about who was living in what types of neighbourhoods and who you thought those people were and how they were going to vote. That was a data-driven decision. We could argue about whether it was a better decision or not. It was a better decision if you were trying to create the outcome that the decision created.
So we're not about to depoliticize any of this, and we're not about to end all of this. You guys have an enormous responsibility to be thinking about what the politics of data are, even if you're just talking about economic data. I don't want you to lose sight of that.