ICTlogy

I am collaborating with ParticipationAI on finding out what would be an ideal model of stakeholder engagement in the life-cycle of public procurement of Artificial Intelligence.

So far, I’ve put together a collection of works related with AI in public Administration, in a very broad sense. Here it comes. All suggestions are welcome.

Working bibliography

Industrial garage door blocked, courtesy Morgane Perraud

Citizen labs —as it happens with open innovation, social innovation, etc.— are very interesting tools for opening up what is now the monopoly of public decision-making. They contribute to enable a citizen participation ecosystem with open, distributed, autonomous public infrastructures, so that an ecosystem of public governance is possible.

Citizen labs (fab labs, living labs, innovation labs, etc.) have already been around for a while. On the one hand, they seem to be a more than a hype and some have a long track of interesting experiences. On the other hand, they do not seem to clearly take off, be part of mainstream citizen activism and, most especially, have a fluid relationship with the Administration.

The main three challenges of citizen labs nowadays are the following:

1. Getting to know how their actual functioning, what works and what does not work, why things succeed or just do not happen, which actors are more relevant and what is their role, what different relationships between actors and tasks are more productive, etc. Of course, there already is quite a bit of literature about the topic, but I don’t think that “the” model has already been found, if it even exists. Related to this, which is about planning and operating, nor is there yet —and this should indeed exist— a way in which (a) the activity and performance of the lab is assessed and (b) the impact (not the outputs) is evaluated. As said, we do have terrific examples of “best cases”, but most of the times they are merely descriptive and, when they try and dig into outputs and outcomes, they usually are too singular or specific to be applied elsewhere.
2. If citizen labs are good —and I truly believe that they have some impact and at different levels— it is only natural that governments should promote them. But there are many doubts about how can bureaucracies promote citizen labs. Citizen labs dynamics are fragile, and they are specially sensible to excessive planning and tight structures. But bureaucracies need (or at least they so far work this way) detailed planning and quite inflexible structures. The paradox, thus, is how the Administration (a bureaucracy) should promote social innovation dynamics by means of fostering citizen labs without spoiling or denaturalizing them, how citizen lab dynamics and their results can be extrapolated and mainstreamed into the entire Administration without distorting them.
3. Sum of the previous two is the issue of how to one goes from the pilot or the project into systemic transformation. When speaking the citizen lab lingo, one speaks about concepts such as designing, prototyping, piloting, replicating and scaling. The feeling, though, is that what happens in the laboratory is fascinating (and I really believe that it has enormous utility in itself in terms of raising awareness and creating social fabric). But mainstreaming citizen labs methodologies and activities is way different from piloting or even replicating. There is this feeling that citizen labs is for the impossible intersection of people (a) interested in politics, (b) knowledgeable about technology and innovation and (c) with plenty of time to spend. And the feeling is felt both at the street level as inside the Administration. Anything about innovation has to transform reality and daily procedures; in politics or decision-making, social innovation means transforming the Administration. I do not think it is already clear how citizen labs are transforming the Administration. Or, in other words, how their impact is embedded in new political and government practices, how they are incorporated into the enormous machinery of public administration.

It seems to me that, for the ones working in citizen participation and social innovation, we still have a long road ahead to provide more solid evidence, standard tools to “easily” develop and run citizen labs and citizen labs’ based social innovation projects, and strategies to have a (deep and thorough) impact on the Administration in particular and in public decision-making in general.

Marcus Carlsen Häggrot, Goethe University, Frankfurt a.M.

Nomads are usually excluded from the electoral process as they cannot be assigned to a specific constituency. Maybe we should reconsider the concept of constituency, especially when residence is decreasingly important in an increasingly mobile society — and most especially within the European Union, with so many expatriates.

Single member plurality systems:

• Pros: popular self-government, accountability, eliminates extremist parties
• Cons: unequal power over policy, vertical inequality, anonymity

Two round election systems put voters in a trade-off between maintaining their integrity or having to vote (in the 2nd round, considering their 1st option did not pass) the lesser undesired one.

José Luis Martí, Pompeu Fabra University, Barcelona. Globalizing democracy, deterritorialisation and Crowdlaw

A new reality:

• Growing complexity
• Globalisation
• Digital revolution: deterritorialisation
• Crisis of democracy: dissatisfaction with institutions, populism, concentration of power, etc.

Globalising democracy: in the XVIIIth century, due to a new social, economic, technological and political scenario lead to a scaling-up of democracy, from the city level to the state level. Maybe, the new scenario coming on the XXIst century should lead us to the scaling-up of democracy, from the state level to the global level.

On the other hand, we are witnessing the (new) rise of cities, the nearest administration to the citizen, able to coordinate between cities. The paradox is that cities are increasingly able to address global issues, while their demos is obviously not global. We certainly have to rethink the traditional approach to democracy and participation. Again, the need to deterritorialise democracy.

And, besides the territorial factor, the deliberative phase and the voting phase of democracy get increasingly intertwined and their differences blur. e.g. liking a proposal on an online participation platform can mean nothing, can mean just endorsement of the idea and thus remain at the deliberation stage, or can boost the proposal and, past a threshold, make it binding, thus entering the voting phase.

Elise Rouméas, CEVIPOF/Sciences Po. Do I go with my party or my beliefs?

Compromise: a decision-making procedure based on reciprocal concessions. There are many reasons for compromise in face of conflict, and many times they have to do with the ethics of voting.

What happens when, to reach a compromise, you “betray” some of your main principles? Tactical voting as an internal compromise. What is wrong, if anything, with tactical voting?

Two main objections:

• Wrong attitude: dishonesty, “gambling”.
• Wrong outcome: mediocrity, obscurity. Not a true revelation of preferences.

A positive case for tactical voting: we have the moral obligation to vote tactically when we have campaigned for a strategic voting.

There also is the idea of reaching a second-best outcome when the optimum is not reachable.

Ismael Peña López, Government of Catalonia

Discussion

Andre: are you proposing a guild-based democracy, with different levels of participation? Can we shift to a commons-based democracy?

Andrei Poama: what is the role of representative democracy and elected members in such a democracy?

Ismael Peña-López: we should certainly promote a commons-based democracy in the sense that anyone should have the tools to make collective decisions. Horizontal networks have proven to be effective, if appropriately facilitated, in diagnosing, deliberating and negotiating. Representative democracy institutions still have a crucial double role: (1) to nurture and take care of this democratic commons and (2) to provide the holistic vision required to connect the different dots and to be able to draw the big strategies, maybe too abstract for grassroots participation methodologies.

Laurentiu: there’s the statement that shifting from hierarchies to networks does not necessarily means losing power. How to back this statement? Based on what theory?

Ismael Peña-López:

• Hierarchies are very sensible to voting with one’s feet: digitization scales-up the power of voting with one’s (e-)feet. Exiting the system (or circumventing it) is stronger than voice.
• Network effects are stronger than economies of scale.
• The estate/administration is the central node.
• Networks are based on a different currency: the gift economy.
• Institutionalising informal participation.
• Enforcing through law and budget.

Chiara Destri: are there externalities in such distributed models? What about accountability? Ismael Peña-López: in a distributed system, accountability is not accurately allocated to anyone and externalities (positive and negative) can go wild as they are difficult to bring inside the system. Institutions thus have the duty — and may be the best positioned — to institutionalise what is going on in a distributed network for collective decision-making, in order to internalize externalities and to allocate accountability.

Ludvig Beckman, University of Stockholm

You cannot be enfranchised if you do not have the real ways to participate in elections.

By what kind of principle can we define the demos in a democracy? Is it the status of citizenship the same as the demos?

If you are part of the demos it’s because you are affected by the decisions of this very same demos. How can we define how one is affected by such decisions? The fact that the state can coerce you to abide with the decision made, then you are affected. You are a subject if, according to the law, you have some duties abiding from the decisions made.

The problem is that not always your jurisdiction coincides with the extent of the law.

Andreas Brøgger Albertsen, Aarhus University

Using voting advice applications (VAA) affects turnout. There is evidence that affects party choice and political references. VAA usage affects knowledge. And use of VAA depends on education, income, age, etc. and has discrimination effects.

How are individual voter’s political preferences affected by receiving advice from a voting advice application?

We find that VAA increases the likelihood of changing your vote if you receive incongruent advice in relationship with your own prior views.

We should strive for differential effects counteracting existing inequalities. That is, to use VAA to affect those less prone to vote so to help them to take the decision of voting. VAA should also be improved to include ethical issues usually not covered by this kind of applications, including the ethics of influence. Also include the impact of VAA in candidate choice, not only party choice.

Toni Gibea, University of Bucharest

Role of experimental ethics in participation and voting. That is, how specific (social) experiments or experiences can affect one’s own judgement and, thus, how we are affecting people’s decisions. What are the ethics behind this? Should be taken into account.

Sometimes, you don’t need to take specific actions to affect judgement and people’s decisions. If a given political candidate states that they will be implementing policies leading to the exclusion of a given minority, is that harming that minority? Are voters of such candidate actually contributing to harm that minority? How ethical is that?

There is a debate whether reasoning improves intuition (dual-process model) or, on the contrary, reasoning finds ways to back and support former intuition (social intuition model).

Chiara Destri, CEVIPOF/Sciences Po. Voting citizens and the ethics of democracy

There is some failure to address actual democratic institutions when providing a justification of democracy: representation, mass participation organised through political parties; failure to to answer to the citizen “incompetence challenge”; failure to account for a democratic understanding of political obligation.

Double role of citizens: as rule-takers, and as rule-givers, that is, “rule of the people, by the people, for the people”. What are the duties citizens have in their role as rule-givers?

What is the distinctive content of democratic citizens’ political obligation? What is the political obligation of parties? What is the political obligation of representatives?

Voting is an action which:

• It is outcome-oriented: citizens vote for someone or something; contributory theory of voting.
• It expresses citizens’ attitudes and beliefs.
• It involves a relation between each citizen, what she votes for and other citizens: electoral results affect all, one votes together or against other citizens (both a cooperative and competitive dimension).
• It comes as the end of a process involving other aspects (public debate, political campaigning, deliberation).

Informed voting as due diligence: voting is a contribution to a result and a relation to other citizens. There is a duty to vote “well”, to get informed before voting. The outcome-oriented dimension of voting requires citizens to be collectively responsible for the outcomes. The relational dimension of voting requires citizens to be individual responsible with respect to their fellow citizens as co-authors of the law. It is consistent with pluralism and reconsiders citizens autonomous political rights.

Parties have important motivational and epistemic functions. they organise political competition. The simplify the political discourse and develop policy platform. Make information publicly accessible. They are “catalyst to public justification”. Are venues of deliberation. Partisanship structures and supports political commitment.

Representatives also do have a role. Representative democracy is quite different from direct democracy, both in the functioning and the justification. The constant tension between the democratic ideal and its representative institutionalisation: accountability to citizens, accountability to parties. Representation as a performative process whereby interests and political entities are also created and not simply taken as given.

Discussion

Toni Gibea: there is an interesting paradox in fake news where partisans are able to correctly identify fake news created by their opponents, but are genuinely unable to identify the fake news disseminated by their own parties.

Ismael Peña-López: about VAAs, one of their problems is where they get the reference data from. They usually get them either from what a given party voted in the Parliament, or what a given party stated in their electoral programme. But there are deviations between those two references: parties can state one thing in their electoral programme and to the opposite when at the Parliament (for several reasons, legitimate or illegitimate). How to measure this bias/gap? How to include it in VAAs?

Pierre-Ettienne Vandamme, Université Libre de Bruxelles

Add a layer of ethics of voting and deliberation around elections:

• Enrich the message conveyed through votes.
• Stimulate public debates, before and after vote.
• Foster a more reflexive and public-spirited ethics of voting.
• Focus attention on policy proposals.
• Clarify the specific mandate conferred to elected representatives.

Need to think about institutional ways of fostering an ethics of voting.

Traditional secret ballots send the message that all motivations are equally valid.

Need to think about devices that both respect privacy and protects voters, incentivizing “deliberation within” on relevant considerations.

Laurentiu Gheorghe, University of Bucharest

Inequality splits society and opens the gates to demagogs.

Big data and artificial intelligence helps in identifying the major trends of society, the major interests of society. This is in general good, but can be used in evil ways: to change the major trends of society, to affect the major interests of society. That is, to massively manipulate society.

We should regulate this in some way: we have to preserve the freedom of building one’s own reasons to vote and the sense of that vote.

Miljan Slavic, University of Belgrade

The proceduralist approach: how we vote is very important. The way we design voting procedures/institutions determines the legitimacy and the outcome of the elections.

Discussion

Elise Rouméas: discrimination should not come free, it should have a cost both for politicians and voters that chose a discriminating programme/option.

José Luis Martí: Sebastian Linares talks about the “democratic oath” as a programmatic compromise with values.

Andrei Poama, University of Leiden

Is deliberation better than voting?

Voting insulates vulnerable citizens, citizens that do not know or just cannot argue in front of other citizens on a deliberation. But they may be good enough at casting an informed vote. Voting protects vulnerable citizens from influence from others (which may be good, but also bad).

Disenfranchising someone from their right to vote has been traditionally justified for criminals, although in many democracies disenfranchising is not allowed. There are other debates about enfranchising/disenfranchising vote for children or elderly people.

Alexandru Volacu, University of Bucharest & Bucharest Center for Political Theory

People usually have a negative view about the ethics of voting.

About individual duties concerning voting, some people believe that they have the duty to vote as a sense of responsibility on what would happen if only “the others” voted; another point of view is that it is a right that costed a lot to have recognized, and thus it would be disrespectful not using it; last, many people believe that voting grants a right to political critique —and, inversely, if you do not vote you should not critizise what you don’t like.

There is the debate whether people have the duty to “vote well”. But it may be more correct to speak about some instances where one can “vote badly” (e.g. most people would believe that selling your vote is not ok), but “voting well” is much more difficult to define.

A usually accepted of “bad vote” is when it goes against your own interests, taking “bad” as non-rational.

About institutional design, there’s the open debate on compulsory voting, allocating voting rights, the design of electoral systems, and the secrecy or openness of voting.

Jonas Pontusson, University of Geneva

There are cases where inequality has not increased (or actually decreased) and nevertheless voters have shifted towards populist/fascist options. So it is difficult to identify one single simple issue as the cause of the raise of populism.

We have a large number of studies that confirm that the poor are less represented, that they vote less, and that middle income (not middle class) voters do not have the weight in policies that they would have considering their number. The different of affluent voters and poor voters is huge and in favour of affluent voters.

Left parties seem to be increasingly shifting from poor voters to middle class / affluent voters. This implies a dispossession of poorer voters, leaving them with lesser options, while middle class voters have much more where to choose from.

On the other hand, people tend to penalize candidates without a certain level of education or skills, and also penalize people earning above a certain threshold (e.g. twice as much as the average income). The problem being that people usually know the educational or professional background of a given candidate, but not their income.

Dominik Gerber, Stockholm School of Economics in Riga. Sustaining democracy: citizen’s duties and the problem of demandingness

Across the world, citizens are losing faith in democracy (perceived performance of democracy).

Four approaches of the value of democracy, based on two axes: instrumental/non-instrumental and epistemic/non-epistemic.

[I really could not follow this presentation]

Interesting readings on the epistemic value of democracy:

• List, Goodin (2001). Epistemic democracy. Generalizing the Condorcet Jury Theorem
• Warren (2017). A Problem-Based Approach to Democratic Theory
• Estlund (2013). Epistemic approaches to democracy. In Encyclopedia of Philosophy and the Social Sciences, Volume 1

Discussion

Chiara Destri: when we speak about parties, their characteristics and their behaviour, we should tell parties as an entity from parties as part of the party system. Quite often parties push in one way while the party system pushes to a different way.

Q: delegation of vote, how would that be considered in terms of voting well/badly? Volacu: cannot see anything “wrong” in vote delegation, always considering that there is no trading in it.

Q: usually, intuitions come first and then we rationalize them. What about if we have bad intuitions? Would that be voting badly even thought we honestly rationalized our voting?