Voices in the Code: Citizen Participation for Better Algorithms

Posted on September 29, 2022 by Tiago Peixoto

Voices in the Code, by David G. Robinson, is finally out. I had the opportunity to read the book prior to its publication, and I could not recommend it enough. David shows how, between 2004 and 2014 in the US, experts and citizens came together to build a new kidney transplant matching algorithm. David’s work is a breath of fresh air for the debate surrounding the impact of algorithms on individuals and societies – a debate typically focused on the negative and sometimes disastrous effects of algorithms. While David conveys these risks at the outset of the book, focusing solely on these threats would add little to a public discourse already saturated with concerns.

One of the major missing pieces in the “algorithmic literature” is precisely how citizens, experts and decision-makers can make their interactions more successful, working towards algorithmic solutions that better serve societal goals. The book offers a detailed and compelling case where a long and participatory process leads to the crafting of an algorithm that delivers a public good. This, despite the technical complexities, moral dilemmas, and difficult trade-offs involved in decisions related to the allocation of kidneys to transplant patients. Such a feat would not be achieved without another contribution of the book, which is to offer a didactical demystification of what algorithms are, normally treated as a reserved domain of few experts.

As David conducts his analysis, one also finds an interesting reversal of the assumed relationship between technology and participatory democracy. This relationship has mostly been examined from a civic tech angle, focusing on how technologies can support democratic participation through practices such as e-petitions, online citizens’ assemblies, and digital participatory budgeting. Thus, another original contribution of this book is to look at this relationship from the opposite angle: how can participatory processes better support technological deployments. While technology for participation (civic tech) remains an important topic, we should probably start paying more attention to how participation can support technological solutions (civic for tech).

Continuing on through the book, other interesting insights emerge. For instance, technology and participatory democracy pundits normally subscribe to the virtues of decentralized systems, both from a technological and institutional perspective. Yet David depicts precisely the virtues of a decision-making system centralized at the national level. Should organ transplant issues be decided at the local level in the US, the results would probably not be as successful. Against intuition, David presents a clear case where centralized (although participatory) systems might offer better collective outcomes. Surfacing this counterintuitive finding is a welcome contribution to debates on the trade-offs between centralization and decentralization, both from a technological and institutional standpoint.

But a few paragraphs here cannot do the book justice. Voices in the Code is certainly a must-read for anybody working on issues ranging from institutional design and participatory democracy, all the way to algorithmic accountability and decision support systems.

***

P.s. As an intro to the book, here’s a nice 10 min. conversation with David on the Marketplace podcast.

artificial intelligence and problems of collective action

Posted on September 27, 2022 by Peter

Although I have not studied the serious scholarship on AI, I often see grandiose claims made about its impact in the near future. Intelligent machines will solve our deepest problems, such as poverty and climate change, or they will put us all out of work and become our robot overlords. I wonder whether these predictions ignore the problems of collective action that already bedevil us as human beings.

After all, there are already about 7.5 billion human brains on earth, about 10 times more than there were in 1800. Arguably, we are better off than we were then–but not clearly and straightforwardly so. If we ask why a tenfold increase in the total cognitive capacity of the species has not improved our condition enormously, the explanations are pretty obvious.

Even when people agree on goals, it is challenging to coordinate their behavior so that they pursue those ends efficiently. And even when some people manage to work together toward a shared goal, they have physical needs and limitations. (Using brains requires food and water; implementing any brain’s ideas by taking physical action requires additional resources.) To make matters worse, human beings often have legitimate but conflicting interests, like the need to gain sustenance from the same land. And some human beings have downright harmful goals, like dominating or spiting others.

One can see how artificial intelligence might mitigate some of these drawbacks. Imagine a single computer with computational power equivalent to one million human beings. It will be much more coordinated than those people. It will be able to aggregate and apply information more efficiently. It can also be programmed to have consistent and, indeed, desirable goals–and it will plug away at its goals for as long as it receives the physical inputs it needs. For instance, it could clean up pollution 24/7 instead of stopping for self-interested purposes, like sleeping.

However, it still has physical needs and limitations. It might use fuel and other inputs more efficiently than a human being does, but that depends on how good the human’s tools are. A person with a bulldozer can move more garbage than a clever little robot that works 24/7–and both of them need a place to put the garbage. (Intelligence cannot negate physical limits.)

Besides, a computer is designed by people–and probably by individuals arrayed as corporations or states. As such, AI is likely to be designed for conflicting and sometimes discreditable goals, including killing other people. At best, it will be hard to coordinate the activities of many different artificially intelligent systems.

Meanwhile, people already coordinate their behavior in quite impressive ways. A city receives roughly the amount of bread it needs every day because thousands of producers and vendors coordinate their behavior through prices. An international scientific discipline makes cumulative progress because thousands of scientists coordinate their behavior through peer-review and citation networks. And the English language develops new vocabulary for describing new phenomena as millions of people communicate. Thus the coordination attained by a machine with a lot of computational power should be compared to the coordination accomplished by human beings in a market, a discipline, or a language–which is impressive.

One claim made about AI is that machines will start to refine and improve their own hardware and software, thus achieving geometric growth in computational power. But human beings already do this. Although we cannot substantially redesign our individual brains, we can individually learn. More than that, we can redesign our systems for coordinating cognition. Many people are busy making markets, disciplines, languages, and other emergent human systems work better. That is already the kind of continuous self-engineering that some people expect AI to accomplish for the first time.

It is of course possible to imagine that an incredibly intelligent machine will identify solutions that simply elude us as human beings. For instance, it will negate the physical limitations of the carbon cycle by discovering whole new processes. But that is an empty supposition, like imagining that regular old science will one day discover solutions that we cannot envision today. That is probably true–it has happened many times before–but it is unhelpful in the present. Besides, both people and AI may create more problems than they solve.

a flowchart for collective decision-making in democratic small groups

Posted on September 23, 2022 by Peter

I few days ago, I proposed that Jane Mansbridge’s great book Beyond Adversary Democracy can suggest practical tools that would assist democratic groups as they make decisions. Such tools should be tested and revised, based on experience in the field.

As a first step, I provide this flowchart (above). The first step is to conduct a survey. The questionnaire would have to be carefully designed, but it could be customized easily for other organizations. Members of the group would be asked what they care about, their attitudes about process, and their social identities and roles within the organization. The survey would yield data that could then inform how the group makes decisions about each issue that the respondents mention.

introducing Habermas

Posted on September 22, 2022 by Peter

This is a 29-minute video lecture* in which I introduce the core ideas of the great German philosopher and sociologist Jurgen Habermas. I made it for our current Introduction to Civic Studies course, but it’s available for anyone to use. It also summarizes the beginning of chapter 4 of my recent book, What Should We Do? A Theory of Civic Life. In the book, I proceed to raise numerous critiques of Habermas, all of which have some validity, although I continue to find his framework useful.

*New version posted on 9/29, with better audio.

an existential threat?

Posted on September 21, 2022 by Peter

John Mearsheimer made the case against Ukraine on a well-attended panel discussion at the American Political Science Association. I don’t agree with those who disapprove of giving him a platform. If he’s wrong, his views should be aired so they can be rebutted, and I think that’s what happened when my Tufts colleague Oxana Shevel and other co-panelists debated him at APSA. If he has any actual insights, we should listen and use them.

Hi new followers! Since u followed me after #APSA2022 where @kath_stoner, @darelasn, Olena Nikolayenko & I disputed Mearsheimer’s claim that NATO expansion made Putin 2 attack ??, I put together?summarizing my comments. Thx @PopovaProf & @dandrezner 4 live-tweeting panel! 1/21
— Oxana Shevel ????? (@OxanaShevel) September 18, 2022

I would emphasize a point that probably arose already in the panel discussion (some of which I had to miss). Mearsheimer said repeatedly that Russia faces an “existential threat” in Ukraine and is responding rationally and predictably. When he completed this phrase, it turned out not to mean a threat to Russia’s sovereignty, autonomy, or security within its borders. Mearsheimer meant a threat to Russia’s status as a great power.

I do not doubt that Putin and other Russian nationalists view a successful and European-oriented Ukraine as diminishing Russian global power. But this is the problem: Russia is not actually a great power.

In 2019, before the pandemic and the current full-scale war, Russia ranked right below Canada (population: 38 million) in GDP, and well below each of four European countries (Germany, the UK, France, and Italy). Its medium-term economic prospects were much worse than theirs because Russia is so dependent on fossil fuels. In population, Russia lies right between Bangladesh and Mexico (which has almost as big an economy), but its population has shrunk since 1992.

To be sure, Russia has a large military. In 2019, it ranked fourth in the world in military spending and sixth in the number of military personnel (at least on paper). But it arguably could not afford this expense. The Russian military cost 4.3% of its GDP, versus 3.7% in the USA, which nevertheless spent 11 times as much. That year, the proportion of the population that was enlisted in the military was four times as high in Russia as in the USA (an economic burden). And Russia was probably getting bad value for its military spending because of endemic corruption.

Russia has a vast supply of nuclear weapons, sufficient to end civilization. Nuclear weapons guarantee its security against a foreign invasion, and certainly pose a risk to other countries. Perhaps they are an asset in a conflict like the current one, because the US would be more likely to supply Ukraine with aircraft and long-distance guided missiles absent the nuclear threat. On the other hand, nuclear weapons are difficult to use to support conventional political aims.

It is a dilemma if Russia lacks the underlying basis for being a superpower yet views any events that diminish its great-power status as existential threats. However, the proposed solution of allowing Russia to act like a superpower is not only unjust; it is also unrealistic. Sooner or later, Russia must navigate the difficult road that Britain and France began to travel after 1945, reluctantly and incompletely acknowledging that they could no longer have empires. The USA should also make that transition, but in our case, the imperative is justice rather than necessity, since we actually retain the capacity to project global power.

applying Beyond Adversary Democracy to solve problems

Posted on September 19, 2022 by Peter

Jane Mansbridge published Beyond Adversary Democracy in 1983. This book has been cited thousands of times and has deeply influenced political theory, certainly including my own work. At the recent annual meeting of the American Political Science association, Mansbridge won the Benjamin E. Lippincott Award for a “work of exceptional quality by a living political theorist that is still considered significant after a time span of at least 15 years since the original date of publication.” Beyond Adversary Democracy was the book in question, although Mansbridge would also deserve the Lippincott for Why We Lost the ERA (1986). She gave a beautiful and interesting lecture on the future of political theory.

During her speech, Mansbridge recalled that she had been inspired to write Beyond Adversary Democracy to help the radically democratic organizations of the 1970s–the many co-ops and communes that sprang up in that era–to address the challenges of self-governance that seemed, sooner or later, to wreck most of them. Someone asked whether any group had ever used her book to solve such problems. She said: never!

Perhaps no one has developed, tested, improved, re-tested, refined, and disseminated a practical toolkit based on the theory of the book. Such a product would be useful not only for radically democratic organizations but for any group that employs democratic decision-making for some purposes and at some times. For instance: a department in a university.

The first draft of such a tool would be an experiment, requiring testing and improvement. The draft might start like this:

A friendly person who is independent of the organization should conduct an anonymous survey (derived from Jenny Mansbridge’s own instrument) to identify points of widespread agreement as well as conflicts of deeply felt interest within the group. “Interests” mean not only self-interested goals and needs (like raising one’s pay) but also ideals, such as committing the organization to a certain strategy.

The person who conducts the survey should then provide a public list of any issues that are not conflictual within the group. Members should not devote time to discussing these issues. Instead, they should delegate them to one person or a small team. The people who are put in charge of each non-conflictual issue should be required to report back periodically and to keep records, and they should be open to advice and complaints, but they should decide what to do without the whole group’s involvement. Rotating such responsibilities can be smart, but it can also be wise to give these tasks repeatedly to people who enjoy them and are good at them.

The survey should also generate a public list of issues that are contentious. The whole group should discuss these issues with a trusted moderator who can stay fairly neutral. An important purpose of such discussion is to help everyone better understand the contours of the disagreement by allowing individuals to voice their divergent values and needs. Some time should be spent looking for creative solutions that could satisfy most people. However, the group should be prepared to resolve such issues by voting anonymously after a finite amount of time spent on deliberation. Although majority rule cannot satisfy everyone, it is better than trying fruitlessly to reach consensus when interests actually diverge.

Once a vote is taken, the fact that a minority has lost should be explicitly acknowledged, and the group should formally thank them and resolve to try to make it up to them in some way later on. Then the group should shift to an issue that has been identified as non-controversial. They should explicitly mark this transition. “Well, that was a tough discussion, but now let’s talk about office furniture. Does everyone agree that Al can take care of that? Thanks so much, Al. Everyone, please contact Al if you have any special concerns.”

The survey should be designed to reveal whether differences of interest fall along identity lines. For example, women might be tend to differ from others on matters related to maternity. Or people of color might have specific concerns in a predominantly White organization. When that is the case, it is important for the people who hold leadership positions to represent both or all of the relevant identity groups. It may be both fair and necessary to reward them for serving on committees, since this burden can fall disproportionately on disadvantaged members of an organization.

On the other hand, when there are no differences of interest, or when differences do not fall along identity lines, then it is much less important for the decision-makers to be demographically representative. It becomes correspondingly more important to give tasks to willing volunteers, to individuals who have experience for the task at hand, or to people whose turn it is do more for the group.

These are principles or maxims, but they could be turned into flowcharts with diagnostic questions and suggestions. For instance: “Does this disagreement seem to divide your group by race, or gender, or job title, or some other identifiable characteristic? If so, what is the characteristic? Does the committee include people on both sides of that difference? How should we reward individuals who join the committee to make it more representative?” (etc.)

This kind of application by no means exhausts the value of Beyond Adversary Democracy, which is mainly a rich contribution to theory. For instance, the book can influence how we read Aristotle or Rousseau. However, applying insightful theories can improve the world; and the experience of applying them can further enrich theory.

Added later: a flowchart for collective decision-making in democratic small groups. See also: the New Social Movements of the seventies, eighties, and today; friendship and politics; needed: pragmatists for utopian experiments

cities in an era of migration

Posted on September 15, 2022 by Peter

(Montreal) I’ve had opportunities to visit several cities since July and have enjoyed watching migrants in those spaces.

In Cordoba, Spain, the cathedral was once one of the world’s great mosques. I watched many Muslim visitors, especially young Francophone Arabs. I wondered how they interpreted this space, with its Islamic heritage and its boldly Catholic symbolism. (In a terrible act of vandalism, Charles V dropped a cathedral right in the middle of its coolly harmonious aisles.) Meanwhile, Grenada now has a mosque for the first time since the renaissance, serving its substantial and growing Moroccan population.

Forty-five percent of Augsburg’s population consists of migrants. I heard second-hand about an African migrant’s strong critique of structural racism, which reinforces what I have learned on other recent visits to Germany. (I was in Weimar last November.) However, superficially, Augsburg appears to be a lively and diverse city in which people from many part of the world–including, now, thousands of displaced Ukrainians–interact quite productively. It’s also nice that Augsburg doesn’t seem to draw many tourists, despite being very attractive and interesting. That means that the diverse population seems committed to the place. For instance, they are rapidly learning German.

Iceland’s national citizen population is about 366,000, and each year before the pandemic it was receiving about 2.3 million visitors (counting the ones who stayed at least one night). Meanwhile, citizens of the Schengen zone can easily get Icelandic work permits. As a result, Reykjavik has turned into one of those global transit zones, where a sample of the world’s wealthier countries–plus a few refugees–parades through public spaces, being greeted and served by people who are almost as diverse as the visitors. Although Dubai is Iceland’s opposite in many other respects (from climate to politics), it is another example of this category.

Now I’m in Montreal, which I always enjoy as a bilingual city that is also a magnet for migrants from the whole world. I love the “bonjour hi” greeting, which invites one to respond in French or English. (As someone whose French is not bad, but is generally much worse than the locals’ , I’m not sure what is expected of me.) I also enjoy watching the trilingualism of many recent immigrants. The men who served me my “shawarma poutine magique mix” communicated amongst themselves in Arabic, and with their customers in rapid-fire English and French. The actual dish combined tater tots, shawarma meat, hot sauce, cheese, and fresh parsley in an apt melange.

why don’t colleges allocate more resources to access?

Posted on September 14, 2022 by Peter

You would think that when a college or university gets a financial windfall, it would spend as much of its new funds as possible to make itself more accessible. It could cut tuition prices, increase financial aid, and/or expand the number of students. But George Bulman finds that none of these things happen.*

Bulman investigates the results when institutions see highly varied returns on their investments, from a 19% increase in an endowment in a single year to a 19% loss. Even in a given year, different comparable institutions can see disparate returns. Bulman finds that when their investments do well, colleges and universities spend more money on their programs, become more selective, allow their tuitions to rise, but allocate no additional money to financial aid, and actually admit and fund fewer students of color. The overall decrease in racial diversity is statistically significant.

Bulman doesn’t really speculate about the reasons. One could model institutions as decision-makers that are trying to maximize their own selectivity and rankings and use windfall money for that purpose. That model fits the data, but I would offer a different explanation that reflects my informal observations better.

I think that a host of groups within any given institution have needs. They make arguments for spending money on everything from student housing to research administration. Often these arguments have merit and an idealistic ring. For instance, students at several universities that I know are advocating more campus housing to relieve rent pressure on nearby neighborhoods that are subject to gentrification. They get this idea from their genuine engagement with those neighborhoods. They don’t want housing for themselves in a narrow way.

However, as a result of many such claims, all available revenues are quickly used up. The new expenditures tend to make the institution look more impressive, increasing applications and allowing the admissions office to become more selective. In essence, it’s a problem of actual internal constituencies trumping the interests of an abstract constituency: potential students.

What should we think when we read this kind of announcement?

Princeton University will enhance its groundbreaking financial aid program, providing even more generous support to undergraduates and their families as it works to attract talented students from all backgrounds.Most families earning up to $100,000 a year will pay nothing, and many families with income above $100,000 will receive additional aid, including those at higher income levels with multiple children in college.

To put this in context, I would note that Princeton’s endowment of $4.5 million per student should generate an average payout of about $225,000 per student per year. Princeton could double or triple its student body and offer full scholarships to all the additional students. Instead, it spends its funds on a range of activities, many of them meritorious, and many of which increase its luster, thereby allowing it to reject 94.4% of its applicants—all the while soliciting its alumni to support financial aid. Again, I would interpret Princeton’s priorities not as an intentional choice to buy selectivity, but as a result of many internal constituencies making valid claims on resources.

(Tufts’ endowment is about $200k per student, which should generate about $10k per student in an average year: a different story.)

See G. Bulman, “The Effect of College and University Endowments on Financial Aid, Admissions, and Student Composition,” NBER Working Paper 30404 http://www.nber.org/papers/w30404. See also Four perspectives on student debt forgiveness; the weirdness of the higher ed marketplace; etc.

Pirate Care, a Syllabus

Posted on September 14, 2022 by David Bollier

The following is one of a series of features from The Commoner's Catalog for Changemaking: Tools for the Transitions Ahead, by David Bollier. The Commoner's Catalog, a compendium of dozens of vanguard commons projects and movements, is available through book stores and online (for free) at https://commonerscatalog.org.

Alarmed that certain types of caring for people has been criminalized, a large group of Europeans assembled a course syllabus in 2019 on what they call “Pirate Care.” As the convenors of the project explained, “We live in a world where captains get arrested for saving people’s lives on the sea; where a person downloading scientific articles faces 35 years in jail; where people risk charges for bringing contraceptives to those who otherwise couldn’t get them. Folks are getting in trouble for giving food to the poor, medicine to the sick, water to the thirsty, shelter to the homeless. And yet our heroines care and disobey. They are pirates.”

Hence the idea of “pirate care” – and the need to offer humanitarian or lifesaving care even if the state chooses to criminalize it. The Pirate Care syllabus, developed by Valeria Graziano, Marcell Mars, and Tomislav Medak, fashions itself as “a research process – primarily based in the transnational European space – that maps the increasingly present forms of activism at the intersection of ‘care’ and ‘piracy.’ It proposes new and interesting ways for intervening in "one of the most important challenges of our time, that is, the ‘crisis of care’ in all its multiple and interconnected dimensions.”

The care deficit of our time can be seen as the failure of various market/state institutions – for healthcare, housing, food, social support – to help those in desperate need.

In general, care is often the most vulnerable human service because markets often do not value care unless there is demonstrable "consumer demand." State bureaucracies like to offer regularized units of service to people, as if they were machines. Large, rules-driven institutions are simply incapable of "care."

And so care remains largely unpaid or poorly paid work. It is often gendered (and marginalized) as "women's work" or as work for people with nonwhite skin.

The Pirate Care syllabus aims to elevate care as something that arises, and is maintained, through commoning. In this sense, care is about resisting state power and markets, and in so doing, revealing how power relations in a society are grossly unequal. This helps explain why providing care is often akin to a political act, such as the Good Samaritan helping an "enemy."

The Pirate Care syllabus explores how providing humanitarian assistance to migrants and refugees represents an elemental human act of connection and compassion. The syllabus offers readings about "pirate care" and suggests practical responses for understanding it through structured reflection, direct action, and “collective memory-writing.” The latter is the process of reminiscing and writing memories as a way to heal collective and personal emotional wounds, especially in the aftermath of a violent historical past.

The practitioners of pirate care see themselves as “experimenting with self-organization, alternative approaches to social reproduction, and the commoning of tools, technologies and knowledges.” This can mean civil disobedience and the risk of persecution for providing “unconditional solidarity to those who are the most exploited, discriminated against, and condemned to the status of disposable populations.” The Pirate Care Syllabus presents itself as an open, evolving “tool for supporting and activating collective processes of learning from these practices.”

I liked how the syllabus explains the political and power dynamics of care:

1. Caring is not intrinsically “nice”; it always involves power relations. Processes of discipline, exclusion and harm can operate inside the matrix of care.

2. Care labor holds the capacity to disobey power and increase our collective freedom. This is why when it is organized in capitalist, patriarchal and racist ways, it does not work for most living beings. We are in a global crisis of care.

3. There are no wrong people. Yet, caring for the “wrong” people is more and more socially discouraged, made difficult and criminalized. For many, the crisis of care has been there for a very long time.

4. Caring is labor. It is necessary and it is skilled labor.

5. Care labor is shared unfairly and violently in most societies, along lines of gender, provenance, race, class, ability, and age. Some are forced to care, while some defend their privilege of expecting service. This has to change.

6. Caring labor needs full access to resources, knowledge, tools and technologies. When these are taken away, we must claim them back.

Pirate Care, a Syllabus

Posted on September 14, 2022 by David Bollier

The care deficit of our time can be seen as the failure of various market/state institutions – for healthcare, housing, food, social support – to help those in desperate need.

And so care remains largely unpaid or poorly paid work. It is often gendered (and marginalized) as "women's work" or as work for people with nonwhite skin.

I liked how the syllabus explains the political and power dynamics of care:

1. Caring is not intrinsically “nice”; it always involves power relations. Processes of discipline, exclusion and harm can operate inside the matrix of care.

2. Care labor holds the capacity to disobey power and increase our collective freedom. This is why when it is organized in capitalist, patriarchal and racist ways, it does not work for most living beings. We are in a global crisis of care.

3. There are no wrong people. Yet, caring for the “wrong” people is more and more socially discouraged, made difficult and criminalized. For many, the crisis of care has been there for a very long time.

4. Caring is labor. It is necessary and it is skilled labor.

5. Care labor is shared unfairly and violently in most societies, along lines of gender, provenance, race, class, ability, and age. Some are forced to care, while some defend their privilege of expecting service. This has to change.

6. Caring labor needs full access to resources, knowledge, tools and technologies. When these are taken away, we must claim them back.