Influence without Surveillance

Project Overview

Parliament is meant to represent the people, but it does not always work as well as it should. Citizens want more input into the functions of government between elections; parliamentarians, conversely, often struggle with a deluge of messages that do not give a clear picture of what their constituents want or how to represent them.

RightToAsk lets people¹ suggest and vote on questions, which could be:

• directed to an MP to ask for an answer, or
• suggested for an MP to raise in Parliament.

RightToAsk shows parliamentarians which questions are popular and relevant to their role. It is informative, making suggestions they may not have thought of, and raising priorities they may not have been aware so many people cared about.

These questions are meant to be asked in formal scenarios such as Parliamentary committee hearings and question time, or answered in written form in-app. We aim to open a channel of communication between parliamentarians and citizens so that Parliament can be more directly and immediately responsive to input from citizens. It is intended to give citizens a way of putting things they care about on the agenda. It is also a way of aggregating expertise, allowing large clusters of citizens to amplify and support those whose insights are regarded as valuable.

RightToAsk aims to improve efficiency in terms of time spent at both ends: citizens can flick through a diverse list of questions and quickly express their opinion by upvoting questions they want answered. If they cannot find an existing question that reflects their priorities, they can suggest a new one. MPs, too, can easily search for questions directed at them or topics of interest and see a short, aggregated list of relevant and popular questions.

Votes are private, even from us. The risk of profiling and manipulation is both a vulnerability in our democracy and a disincentive for participation. RightToAsk uses additive-homomorphic encryption to add votes without revealing how individuals voted. It thus defeats political profiling and tracking, rather like an electronic equivalent of putting everyone’s vote into a ballot box and counting them only when their link with the voter has been broken.

However, RightToAsk does not attempt to hide who wrote which questions—this is partly to disincentivise aggressive or threatening ones, and partly to encourage connections between MPs and people who ask good questions.

Example scenarios

Parliamentary Committee Members

can use RightToAsk to get expert suggestions for good questions. Suppose Senate Estimates wants to examine the effectiveness of a recently-built government app, but doesn’t understand the app or its likely failures well enough to formulate effective questions. Knowledgeable technical people who have been examining the app can suggest important questions for the committee members to pose, and other citizens can vote them up so they receive attention.

People with important lived experience of a particular issue don’t need to organise in advance to make a large impact - one person can frame a relevant question, and others with similar experiences can amplify it.

MPs

can assess the importance of questions to their electorate. For example, if someone raises a question about an issue that matters to them, but nobody else upvotes it, then the MP can filter questions for their electorate alone and see that most of their constituents have flicked past it. Conversely, a new question that had received 100% upvotes might deserve attention even if the total number of votes was small. The MP may answer the question directly or raise it in Parliament.²

Journalists

and anyone else can adopt the questions for use in press conferences or any other setting.

Each MP’s contribution and responses will be public, both the questions they have raised in parliament (as suggested by the platform) and the ones they have answered. The idea is to incentivise participation by rewarding active participants with a public show of how well they are doing their job of responding to the wishes of the electorate.

Of course, MPs are also welcome to suggest questions to the electorate for feedback.

Goals

The primary goal is meaningful information flow from citizens to MPs. This should be valuable and efficient at both ends.

We step away from the notion of ‘voting on decisions’ towards raising awareness and asking questions about issues that are important to lots of people. We do not aim to distinguish something with 51% support from something with 49% support - instead, we try to raise up several issues, in the form of questions, that a large number of people care about.

The aim is to give MPs a succinct, clear list of most-popular questions, rather than a deluge of thousands of emails which take time to assess.

Citizens do the work of aggregating opinions and assessing others’ questions. There is some natural language processing for keyword searches and to present possibly-duplicate questions to citizens before they upload a new one. We use AI only to suggest things to human users, not to make the decision. This certainly does not remove the risk of algorithmic bias, but we hope it reduces it a little.

Final decisions on what to write, whether another question is equivalent, and what to vote for, rest with humans, thus providing a more trustworthy way of directing the most important questions to MPs (although, of course, people can be biased too).

1. Participants may be citizens, permanent residents, under-18s or (other) non-Australian-citizens. The term “MP” includes senators and members of state legislative bodies. ↩

2. The Victorian Parliament has recently introduced a special session for constituent questions: https://www.parliament.vic.gov.au/about/news/2458-constituency-questions-introduced ↩