Levine What Should We Do - Pp. 1-50
Levine What Should We Do - Pp. 1-50
Levine What Should We Do - Pp. 1-50
I assume that you are a citizen, in the sense of someone who belongs to one or
more communities that you hope to improve. Your communities may range
from a city block or a religious congregation to the whole earth. As a citizen of
these communities, you seek to address their problems and influence their dir-
ections; but more than that, you want to make them through your thought, your
work, and your passion. You want to be a co-creator of the human world.
All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or applicable copyright law.
As a good person, you ask, “What should I do?” But as a good citizen, you must
ask, “What should we do?” The question becomes plural for two reasons: no one
can accomplish much alone, and we must reason together to improve our opin-
ions and to check biases and self-interest. Reasoning together is an indispensable
way to think well about matters of public concern. To be sure, the individual per-
spective never vanishes, because people should reflect on which groups they be-
long to (membership is not always explicit, obvious, or voluntary), which ones
they should seek to join, and which ones they should exit. Those are matters of
individual ethics. But to be a citizen, one must also adopt the plural “we.”1
The citizen’s question—“What should we do?”—ends with an action verb be-
cause it is not enough to form and express opinions; citizens must actually affect
the world. They must decide under conditions of uncertainty and limited time,
and then act. (Refraining from action is a form of action for which members of a
group are responsible.) Acting and then reflecting on the results is an important
way to a make a group’s discussions and learning serious.
Importantly, the question is not “What should be done?” or “How should
things be?” These are the predominant questions in political theory, political phi-
losophy, and the study of public policy. They are often far too easy. For instance,
what should be done about the threat to the global climate? Carbon should be
taxed everywhere to reduce production, and the proceeds should be used to mit-
igate the harm of rising temperatures, with most of the funds going to the world’s
poor. The carbon tax might not even have to be very high, and other tax cuts
could offset it.2 I think this proposal is correct, but it is also empty. I cannot actu-
ally tax the world’s carbon; neither can you. We are not influential in any group
that has that power. Our responsibility and accountability vanish if we ask what
Copyright 2022. Oxford University Press.
1 Portions of c hapters 1, 2, and 3 appeared in Levine 2019 and are reprinted here with permission.
2 Metcalf 2019.
What Should We Do?. Peter Levine, Oxford University Press. © Oxford University Press 2022.
DOI: 10.1093/oso/9780197570494.003.0001
EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI
AN: 3279159 ; Peter Levine.; What Should We Do? : A Theory of Civic Life
Account: s8335684.main.ehost
2 What Should We Do?
should be done instead of what we should do, where “we” refers to a group in
which we have tangible agency.
The opposite pitfall is to be satisfied with individual actions or shared volun-
tary decisions by small groups. A family or a few friends can agree to turn off
their lights and stop eating beef to save carbon, but 7.5 billion other human be-
ings will go on as before. The hard question is how any group in which I have
genuine agency can affect large problems. A good citizen does not shy away from
that question because it is difficult.
The “should” in “What should we do?” is also essential. Good citizenship is
not about doing what we want, or what our biases or interests or norms tell us
to do. It is about struggling to pursue the best ends with the right means. We
are responsible for altering our wants, biases, and interests to make them better.
Since good ends and means cannot simply be looked up, the pursuit of moral im-
provement is intellectually difficult. We know that we can be wrong about right
and wrong, for all the people who have sincerely believed in slavery, patriarchy,
and violence have been badly misguided—and we share the limitations of those
fellow human beings. But although we can be wrong, we must try to be right. The
good citizen does not shy away from the question “what should we do?” because
the answers are contested.
Finally, the word “what” in this question matters, because the good citizen
must be concerned about the options: their costs, risks, and probable outcomes.
Empirical information is essential in this analysis. When time is scarce, addi-
tional empirical investigation can be too costly or time-consuming, but the good
citizen is not afraid to face the implications of the available evidence.
By putting the emphasis on what we should do, I am not suggesting that eve-
ryone bears primary responsibility for addressing every issue. You may be the
victim of a social injustice that someone else has created or has the best oppor-
tunity to remedy. In such cases, it is most important for them to decide what they
are obliged to do to improve your situation. Not every problem is your problem.
Nevertheless, “What should we do?” remains an important question for vir-
tually all of us. Even if the main moral responsibility lies with someone else, the
only thing we can control is what we do. We may decide that we should demand
justice from another person or group, but making a demand is also a form of
action that we choose to take. In fact, making demands on “target authorities”
is the characteristic activity of social movements;3 and social movements are
composed of people who ask, “What should we do?” It’s just that their goal is to
compel other people to take more responsibility.
Furthermore, acting is not merely a price we must pay in order to improve
the world. It can also be a benefit that we reap, since exercising agency can be
3 Tilly 2004, 7.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
The Citizen’s Fundamental Question 3
tralists and decentralists, large and small states, and opponents and advocates of slavery.”
7 An example of an influential principle is Robert A. Dahl’s: “Every adult subject to a government
and its laws must be presumed to be qualified as, and has an unqualified right to be, a member of the
demos.” Dahl 1989, 127.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
4 What Should We Do?
ultimate sovereignty. Rather, it is a name for a whole array of offices and bureaus,
officials, rules, norms, and laws that may be in mutual tension and that inevitably
shade into other institutions that are not strictly governmental. Many non-state
groups and institutions exercise power over governmental bodies. The state is
in the midst of associations of all kinds, not the fundamental basis of all other
associations.
Second, even when we decide that a government’s policies about who should
have which rights are the most pressing issues, the main question for us re-
mains: what should we do about it? Although it is worth analyzing citizenship
rights in a given state and arguing that they should be defined or enforced differ-
ently, we cannot stop with such analysis. If the status quo is unjust, we should try
to change who has legal citizenship. To do that will probably require joining or
forming groups other than the state that can advocate change, whether they are
lobbies, social movements, or guerrilla armies. So my focus is on membership in
a wide range of groups, in a world in which membership is not always voluntary,
and in which people are often included or excluded against their will.8
To take this stance is to be “republican” in one ancient sense of that word.
Etymologically, a republic is the “public thing” (or “public good”). In this book,
I assume that a very wide range of goods belong to their associated publics: the
safety of a street belongs to the neighborhood; a country’s laws belong to its citi-
zenry; and the global climate belongs to us all. These publics have both the right
and the responsibility to exercise ownership of their respective goods—with the
question of who collectively owns which goods constantly arising. Collective
goods are compatible with (and often essential for) private goods. For instance,
the safety of a street is a public good even if all the buildings have private owners.
To own a public good is to ask what we should do about it. It means resisting the
passive question, “What should be done?” Classical republics were born when
the public wrested responsibility from a monarch, a clique, a clergy, or a foreign
power, thereby adopting an active voice.9
To do a good job of collectively governing any good requires appropriate vir-
tues and institutions. Some of these virtues are discursive: good speaking, rea-
soning, and listening. But discursive virtues will not suffice; people must also
work, sacrifice, and sometimes fight to defend their common goods. It is natural
to use the word “civic” for the skills, habits, virtues, and institutional forms that
enable us to govern common goods well; thus we speak of “civic discourse,” “civic
8
In the terms of Tully 2008 (8), this is a “civic,” not a “civil/national,” notion of citizenship.
9
Cf. Constant 1819: “the liberty of the ancients . . . consists of exercising collectively, but directly,
many parts of absolute sovereignty, [and the right] to deliberate, in a public space, about war and
peace, to ratify treaties of alliance with foreigners, to vote laws, pronounce decisions, examine the
accounts, actions, and management of officials, to compel them to appear before the whole people, to
accuse them, to condemn or acquit them.”
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
The Citizen’s Fundamental Question 5
engagement,” “civic associations,” “civic forums,” and “civic courage.” This book
is a contribution to civic republican theory, although I dissent from some uses of
that phrase.
It can also be seen as a contribution to democratic theory. “Democracy” is
defined in many ways, but I find it most useful as the name of a system in which
voting plays an important role and in which everyone who is a full member gets
an equitable share of the vote. To make voting legitimate and valuable may re-
quire additional features, from freedom of speech and rule of law to some degree
of social equality. These features buttress the central architectural element of a
democracy, the vote, whose purpose is to equalize power.10 It seems likely that
democracy, so defined, is the best system for governing the public goods that are
the laws of most (perhaps all) sovereign states. Therefore, states should generally
be civic republican democracies. It is a more complex question whether other
important entities—such as religious denominations, professions, academic dis-
ciplines, computer networks, extended family and kinship groups, markets, and
the earth as a whole—should be democracies. It is possible for people to own and
govern these things without ever voting. Still, all should be civic and republican,
because in each case, a group must define and govern itself and its goods wisely.
This book’s question, “What should we do?,” has a plural verb because politics
exists once people belong to groups of any kind, from small voluntary associ-
ations to nation-states. The plural question raises a new set of issues that are not
directly addressed in individual ethics.
To start, problems of complicity arise once you belong to a group. Consider
a case like the firebombing of Dresden by Allied forces during World War II,
which probably caused 35,000 civilian deaths on one night and did nothing to
advance the Allied victory over Nazism. The firestorm sucked oxygen out of the
air and caused civilians in shelters to die of asphyxiation. It was started by bombs
from one thousand airplanes. Eight thousand crewmen flew in those planes,
and “many thousands further were involved in planning and support.” Exactly
the same number of deaths would have occurred if 999 bombers had flown in-
stead of one thousand. Thus, as Christopher Kutz observes, each crewman or
10 Robert Dahl defines democracy as “voting equality at the decisive stage” (Dahl 1989, 109–11).
I find this narrow definition useful for analytic purpose; it allows us to investigate what is needed to
make democracy work well and what outcomes democracies accomplish. If we load other features
into the definition, the resulting system (e.g., voting plus deliberation, or voting plus social equality)
may be more attractive, but it’s harder to analyze whether voting equality is necessary or sufficient.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
6 What Should We Do?
You cannot have absolutely your own way—you cannot be a free lance; you
cannot be strongly and fiercely individual if you belong to an organization. For
this reason some folk hunt and work alone. It is their nature. But the world’s
greatest work must be done by team work. This demands organization, and that
is the sacrifice of some individual will and wish to the good of all.13
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
The Citizen’s Fundamental Question 7
14Hirschman 1970.
15See, e.g., Baron, Philip & Slote 1997.
16 The famous trolley problem (Thomson 1985) offers a fictional scenario that poses a conflict be-
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
8 What Should We Do?
must consider the group’s future capacity to act effectively. In many cases, a group
that tried to do the most possible good for the whole world would dissolve, be-
cause the level of sacrifice expected of its members would be too great, and they
would exit. Since the existence of a group permits deliberation and coordinated
action, which are impossible for individuals, dissolution may be too high a price
to pay.
Third, we can consider the virtues of a group—virtues understood, in an
Aristotelian way, as dispositions that are reflected in, and reinforced by, actions.
In other words, virtues are habits that are durable but can be deliberately shaped.
Groups as well as people can have virtues, such as courage, temperance, magna-
nimity, etc. To cite an example developed by Donald Beggs, the individual mem-
bers of a knitting club may not be concerned enough about people with HIV/
AIDS that they would take action on their own. But the group (possibly under
external influence) can develop a strong tradition of knitting for AIDS patients.
Then the group has durable dispositions or virtues, such as empathy and inclu-
siveness.17 Groups also have “epistemic” virtues or vices that make them good or
bad at understanding the world, for example, an ethnocentric group is bad at un-
derstanding outsiders.18 Developing and maintaining virtues requires different
strategies when a group (instead of an individual) is the thing that has character
traits.
We will revisit these issues of loyalty and complicity in chapter 7, equipped
with a more robust theoretical structure for thinking about collective action.
This discussion has assumed a simple dichotomy of individuals and groups. That
scheme must be complicated in three ways.
First, individuals do not really precede groups. Anyone who thinks in a lan-
guage is already part of a linguistic community. People who ask, “What should
we do as a city?” were probably influenced by the history, norms, and prevailing
values of that city, which had existed before they did. Even if they react critically
to the dominant opinions of the city, the city still influenced them by causing
their reaction. These are examples of the dependence of individuals on groups
that we never chose to join.
Second, groups are not merely accumulations of individuals. Consider these
statements: “A group just is the people who make it up.” “If a group can be said
to have intentions at all, its intentions must somehow be the intentions of its
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
The Citizen’s Fundamental Question 9
19 Epstein 2015.
20 Epstein 2015, 47–49.
21 Epstein 2015, 272.
22 Epstein 2015, 168.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
10 What Should We Do?
At all scales, citizens confront myriad concrete issues: poverty, tyranny, crime, racial
injustice, environmental degradation, and many more. These issues are too various
and contextual to be addressed by any single theory. However, beneath such specific
challenges lie three general categories of problems. Political theory that is useful to
citizens must address these three categories.
First, we often struggle to coordinate our separate interests and choices to ac-
complish valuable ends. Here I refer to problems of collective action. When these
problems defeat us, we fail to achieve goals that everyone would agree are worthy.
For example, even the many nations that agree that carbon emissions threaten the
earth’s environment have so far failed to enact agreements that would restrict their
respective use of carbon sufficiently. Most lament this failure; it nevertheless persists
as a collective-action problem.
Second, we often disagree about what outcomes we should value or what
means are ethical and appropriate. Some people hold extremely bad opinions of
those questions. Their values pose a threat, and the fact that they hold them re-
minds us that we, too, could be wrong. We may try to identify the right values by
discussing contested issues with other people, but such discussions can go badly
for a set of reasons that I will call problems of discourse. In 2015, the US Senate,
frequently called the “world’s greatest deliberative body,” defeated an amend-
ment that would have expressed the body’s opinion that “climate change is real;
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
The Citizen’s Fundamental Question 11
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
2
A Case
The Montgomery Bus Boycott
What Should We Do?. Peter Levine, Oxford University Press. © Oxford University Press 2022.
DOI: 10.1093/oso/9780197570494.003.0002
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
The Montgomery Bus Boycott 13
associations, and at least one important labor union, the Brotherhood of Sleeping
Car Porters. Such organizations represented social capital—a concept explored
in more detail in c hapter 4.
The sociologist Aldon Morris credited Black churches with creating forums
for discussion, identifying and training leaders, and offering symbols around
which individuals could cohere. In the decades before the civil rights move-
ment got going, some well-placed and sympathetic observers had argued that
the strength of the Black church was an obstacle to social change, because it di-
verted attention away from political activism.1 But the 1950s showed that deter-
mined leaders could redeploy the assets of the Black church to create a political
movement. According to Morris, the “presence of indigenous resources” did
not guarantee success, but leaders could “re-direct and transform” available as-
sets to “develop and sustain social protest.”2 Recent scholarship has emphasized
the importance of other forms of African American social capital in addition
to churches. Some categories of organization (e.g., Black newspapers) were al-
ready political before 1910, and some (e.g., beauty salons) were owned and led
by women.3
African American organizers had also accumulated relevant experience. As
early as 1942, the Congress on Racial Equality (CORE) had organized sit-ins and
other forms of nonviolent civil disobedience. In Baton Rouge, Louisiana, African
Americans had even conducted a bus boycott similar to the Montgomery bus
boycott in 1953.
These organizations, networks, and experiments nurtured leaders. For ex-
ample, by the time she was arrested on December 1, 1955, Rosa Parks had served
as a professional staff member of the Alabama NAACP and had been an active
participant in the League of Women Voters. When a young African American
woman, Mrs. Recy Taylor, was gang-raped by White men, Parks had helped
to form the Committee for Equal Justice for her, a significant effort to combat
both white supremacy and violence against women. Parks was employed as a
domestic worker by one of the most progressive White couples in Montgomery,
Clifford and Virginia Durr, who helped her to participate in trainings in nonvi-
olent civil disobedience at the Highlander Folk School in Tennessee. One of her
mentors at Highlander was Septima Clark, often called the “Queen Mother” or
“Grandmother” of the civil rights movement.
The boycott began after Parks boarded a Montgomery bus that was racially
segregated by law, with Whites seated at the front and Blacks at the back. Black
female domestic workers predominated as passengers on Montgomery’s buses
1 Brown & Brown 2003, 634, summarizing arguments by E. Franklin Frazier and Gunnar Myrdal.
2 Morris 1986, 78, 291.
3 Ling 2006; Jones 2010.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
14 What Should We Do?
and were frequently sexually harassed as they interacted with the White male
drivers. Parks sat in the “colored” section but was ordered (with several others) to
renounce her seat to a White rider. She refused and was arrested. The charge was
refusing to give up a seat in the “colored” section.
When Parks was arrested, she called her mother, who called E. D. Nixon, the
president of both the Montgomery branch of the Brotherhood of Sleeping Car
Porters and the state’s NAACP chapter. Nixon called Clifford Durr, the White
liberal lawyer. After Nixon and Durr posted bail for Parks and she agreed to con-
test her arrest, Nixon made a list of people he would call the next morning, in-
cluding two young Black pastors, Ralph Abernathy and Martin Luther King Jr.
Meanwhile, someone else called Fred Grey, a Black attorney in town, who, in
turn, called Jo Ann Robinson, who was a professor at the local Black college,
Alabama State, and the president of the Women’s Political Council, an African
American women’s group. Robinson assembled some of her friends on the
Alabama State campus late on the night of Parks’s arrest and used the college
mimeograph machine to make flyers announcing a public meeting in support of
Parks. It was Grey who called Robinson, but she had also worked with Nixon and
had served as a leader in King’s church.
These connections are examples of the ties among a network of African
American leaders (plus a few supportive Whites) in Montgomery. The day
after Parks’s arrest, about fifty local Black leaders gathered in the basement of
King’s church.4 Most of them probably knew each other already. Collectively,
they controlled notable assets, from Alabama State’s mimeograph machine to
several churches and a union hall that offered reasonably safe places to meet.
Montgomery’s Black-owned taxi companies could ferry commuters to work if
there was a bus boycott. Lawyers and people with financial assets were among
those who convened in the church. Their discussion is an exemplar of a group
deciding what they should do.
At first, their natural leader appeared to be Rev. L. Roy Bennett, who was
president of the Black ministers’ group in Montgomery. This group, the
Interdenominational Ministerial Alliance, also looked like the right organization
to coordinate any activism. However, Bennett exasperated most of the other ac-
tivists by lecturing at length and refusing to share the floor. The other existing
Black organizations in Montgomery either had narrow memberships or were
vulnerable if they took a prominent role. For example, Jo Ann Robinson could
be fired from Alabama State if the Women’s Political Council, which she led,
was seen as too political. A group of activists decided to create a brand-new as-
sociation, the Montgomery Improvement Association (MIA). They persuaded
a somewhat reluctant King to be its first president and elected Bennett as the
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
The Montgomery Bus Boycott 15
vice president, managing that transition through private conversations, not a re-
corded vote.5
The Montgomery Improvement Association first met in the Alabama Negro
Baptist Center, until the Center’s White donors demanded the radicals be thrown
out; then in the Citizens Club, until the city threatened to revoke its license; then
temporarily in the First Baptist Church building; and finally, in the Bricklayers
Union Hall. “Here the white community could not force us out, since most of
the members and all of the officers of the union that owned the building were
Negroes.” In this passage, King quickly names four existing Black organizations
that were able to offer the physical capital of space because of their social capital.
Overall, King writes, “Labor, civic, and social groups were our stanch supporters,
and in many communities new organizations were founded just to support the
protest.”6
The MIA’s elected officers met regularly and made key decisions throughout
the Montgomery campaign. Thousands of individuals from Montgomery’s
African American community joined and contributed to the MIA. At mass
meetings held in churches, everyone would vote by calling out “aye” or “nay” on
such matters as whether to start (or, later, to stop) the bus boycott. However, the
board had already decided on a course of action, and the mass votes merely rati-
fied their decisions.
As the MIA considered and reconsidered the question, “What should we
do?,” they had to address the difficulty of coordinating individual acts of resist-
ance. Before Rosa Parks, other individuals had refused to give up their seats on
segregated buses (including a young woman named Claudette Colvin earlier
in Montgomery in 1955). Individuals had protested segregation in many other
ways as well. However, if one or a few people resisted at a time, they would not
make much positive difference, and they would risk violent reprisals. On the
other hand, if thousands of people resisted at the same time, they would be safer
and they could have a major impact.7 When thousands of Black Montgomerians
boycotted the city’s buses, they did so safely and defeated the bus company.
Fear and pessimism could make the challenge of coordination more difficult.
By 1955, Black Americans had experienced 336 years of slavery and then vio-
lently enforced white supremacy.8 Many people responded by exiting the state
entirely. About one third of Alabama’s African Americans had participated in
the “Great Migration” to cities in the North.9 Their exit gave Montgomery’s civil
rights movement connections to family members outside the South and probably
5 Garrow 1986, 18–22; cf. Branch 1988, 137, for a somewhat different account.
6 M.L. King 1958/2010, Kindle loc. 1139, 1094.
7 This is the overall topic of Chong 1991. For the Montgomery Bus Boycott, see 101–2.
8 Thornton 2018.
9 Gregory n.d.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
16 What Should We Do?
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
The Montgomery Bus Boycott 17
bus was full. In other words, they did not originally contest segregation on city
buses, just the additional injustice of having to give up seats in the Black section
at the back. The NAACP was critical of this objective because it sought to abolish
segregation on all transportation and did not want to concede that separate but
equal buses could ever be acceptable.
However, even on the night when the MIA voted to boycott buses, King de-
picted their objective as much deeper and broader than changing who could sit
where on buses. He told the mass meeting of the MIA, “We, the disinherited of
this land, we who have been oppressed so long, are tired of going through the
long night of captivity. And now we are reaching out for the daybreak of freedom
and justice and equality.”13
A movement must also choose one or more organizations, agencies, or people
as its targets and create costs or difficulties for them. The MIA targeted the
Montgomery City Lines Inc. (the city’s bus company) by organizing a boycott
that would cost the company thousands of fares each day.
Other potential targets could have been chosen. For example, during the boy-
cott, Montgomery’s police commissioner joined the White Citizens Council, an
organization explicitly committed to preserving white supremacy. The commis-
sioner, the police department, or the White Citizens Council could have been
chosen as the main target, as could downtown stores or the Democratic Party.
The MIA chose the bus company because they saw a particular vulnerability that
they could exploit: the city’s buses depended on Black riders who could refuse to
ride and find other ways to work.
The MIA told the bus company and the City of Montgomery that Black com-
muters would go back to riding the city’s buses if: (1) Black riders did not have
to give up their seats to Whites; (2) drivers treated Black passengers courteously;
and (3) some Blacks were hired as drivers. These three points were their initial
demands. Meanwhile, Fred Grey and others involved with the MIA sued the city
and the bus lines to end all segregation in public transportation. They ultimately
won their case.14 They made different demands in court because a federal lawsuit
was a different context from a boycott. Demands are different from objectives
(often narrower and more modest), and they vary depending on the target.
The MIA also had to choose a strategy. Its main strategy was to organize and
sustain a boycott of the city’s buses for many months. Meanwhile, the MIA
sued to end segregation in federal court. When the police indicted eighty-eight
leaders of the boycott (including King), these leaders surrendered voluntarily
and made their imprisonment into an act of nonviolent resistance that brought
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
18 What Should We Do?
international media attention to their cause. They also presented their case to re-
porters and raised money from supporters across the country and overseas.
These were the strategies of the MIA, but they could have taken other actions
instead or as well. Later, the civil rights movement would often violate police
or court orders against marching so that the police would arrest them and they
would fill jails with thousands of supporters. They organized sit-ins, they regis-
tered voters, and they held large funeral ceremonies for martyrs. In Memphis in
1968, they supported a major strike. The civil rights movement could have em-
ployed violence against people or property, although it chose not to.
The movement also had to decide what to say and how to say it to various
audiences. The MIA made Martin Luther King Jr. into its main spokesperson,
and King gave speeches that are still widely read, as well as many interviews to
reporters. Soon after the movement’s success in Montgomery, he published a
book, Stride toward Freedom, that presented his early views more fully to a global
audience.
During the mass meeting that launched the boycott, King described Parks’s
recent arrest this way:
Mrs. Rosa Parks is a fine person. And, since it had to happen, I’m happy that it
happened to a person like Mrs. Parks, for nobody can doubt the boundless out-
reach of her integrity. Nobody can doubt the height of her character. Nobody
can doubt the depth of her Christian commitment and devotion to the teach-
ings of Jesus. And I’m happy since it had to happen, it happened to a person
that nobody can call a disturbing factor in the community. Mrs. Parks is a fine
Christian person, unassuming, and yet there is integrity and character there.
And just because she refused to get up, she was arrested.15
Like any exercise in public communication, this passage reflects choices. The
MIA did not have to choose Martin Luther King as its leader. Rosa Parks could
have spoken in addition to King or instead of him. Or King could have spoken
about Parks’s experience as a militant activist and her long record of fighting
sexual violence against Black women. Instead, he chose to emphasize the invol-
untary nature of her arrest and her Christian faith. During the boycott and later,
King invoked American history, democracy, law, Christianity, and racial justice
in specific ways to influence fellow members of the Black community, relatively
sympathetic Whites, majorities of Americans, and global audiences.
The leaders of the Montgomery Bus Boycott had to make concrete choices of
objectives, targets, demands, strategies, and rhetorical framings; and they had
to design procedures to make these decisions collectively and implement the
15 M.L. King 1955. He tells the story in a similar way in M.L. King 1958/2010, Kindle loc. 651–652.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
The Montgomery Bus Boycott 19
results successfully. They applied detailed information and insights about their
own circumstances. They also displayed an impressive understanding of more
general problems that typically confront civic actors. Here I will itemize these
problems and consider how they arose in the Montgomery case.
16 M.L. King 1963a. Cf. M.L. King 1963b on “the myth concerning time.” And see Luban 1989 for a
detailed study of King’s account of time in conjunction with Walter Benjamin’s idea of Jetztzeit, “time
of the now.”
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
20 What Should We Do?
Free-riders: It is often tempting to let other people carry the burden for a
public good, in which case the good may not be provided even if everyone wants
it. Examples range from a failure to clean the dishes in a shared house, to a deci-
sion not to vote and let others choose the government, to the refusal of nations to
limit their carbon emissions.
Although the term “free-rider” conveys moral disapproval, that is not appro-
priate in cases like the Montgomery bus boycott. Individual African Americans
could hardly be reproached for hesitating to challenge white supremacy, because
resistance would be dangerous and futile in the absence of a powerful and effec-
tive movement. It would be reasonable for Black individuals to hope that others
might resist and then benefit from their sacrifice. Meanwhile, the system of white
supremacy had overcome its own challenges of free-riding by paying the police
and others to enforce a racial hierarchy. Thus, free-riding is not always a moral
failing, and solving a dilemma of free-ridership does not always improve the
world. Nevertheless, free-riding is a challenge to any effort to coordinate human
action, whether for good or evil.
Free-riding bears a family resemblance to other situations that game-theorists
model as prisoner’s dilemmas, arms races, games of “chicken,” and other types.17
Understanding the differences among these situations can be important for
selecting the best solutions. However, for the purpose of this chapter, the point
is simply that people often make individual, independent choices that are rea-
sonable for achieving their respective goals (including understandable or even
praiseworthy goals) but that produce bad outcomes for the whole group.
Path dependence: Once enslaved people were transported to the future
United States, economic and political interests arose that favored slavery. The
national economy depended on cheap Black labor and had its own inertia. The
cost of banning slavery would have been low in 1650; doing so would have put us
on a different path. Much later, in the 1870s, when formerly enslaved people were
denied a share of the land that they had worked, a new economy developed that
had a political constituency and economic inertia. In short, to change to a system
of economic and social equality is costly once it exists.
This is an illustration of path dependence, which can also be observed in many
other contexts. For example, we might all be better off if, a century ago, cars had
been developed to use electricity instead of gasoline. By the twenty-first century,
electric cars would be fantastically fuel-efficient and convenient.18 But the petro-
leum path was chosen, and now to shift to electricity is very expensive and diffi-
cult—so much so that it might even be unwise.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
The Montgomery Bus Boycott 21
The Iron Law of Oligarchy: The founders of the MIA were committed to dem-
ocratic and egalitarian values. However, from its launch, the MIA was dominated
by pastors, professors, lawyers, and other Black Montgomerians who held ad-
vanced degrees—and, often, extensive prior experience in civil rights organi-
zations. During the remaining dozen years of Martin Luther King Jr.’s life, the
composition of the leadership of the national movement became a point of fre-
quent tension, with groups like the Student Nonviolent Coordinating Committee
(SNCC) chafing at the dominance of somewhat older and more seasoned leaders
like King and his colleagues in the Southern Christian Leadership Conference
(SCLC). Women were clearly underrepresented in the inner councils of the
movement, despite their enormous contributions. Ella Baker, who played central
roles in both SCLC and SNCC, observed that younger people often told her she
would have been famous and held more exalted positions if she had been male;
and she agreed that the movement had been “largely caried by women.” She op-
posed depending on strong, charismatic leaders.19 Another example of bias was
the decision to keep Bayard Rustin offstage because he was gay.
With the benefit of hindsight, we can say that the leadership of the national
movement could have done a better job of including more and more diverse
people. However, the challenge is intrinsic and cannot be explained entirely on
the basis of biases, such as sexism or homophobia. There is a general tendency
for small subsets of people who already possess the most experience to dominate
even the most egalitarian groups.
The sociologist Robert Michels coined the phrase “Iron Law of Oligarchy”
in his eponymous 1911 book, which focused on the democratic socialist par-
ties of Europe. Even though these parties were committed in theory to internally
democratic processes, they formed bureaucracies that empowered their senior
leaders.20 Michels believed that this pattern represented an “Iron Law,” because
organizations require specialization and hierarchy, and that undermines egal-
itarian goals. Even if Michels’s supposed “law” is just a tendency, it is frequent
enough to constitute a generic problem.
Problematic decision rules: The MIA reached decisions by deliberating in-
tensively within a small leadership group, who typically reached a consensus
that they then presented to the mass membership for a vote by acclamation. This
19Baker n.d. (but before 1973), 350–351. See also Payne 1989.
20Michels 1915, thesis first stated on p. 11. Doerr 2018 provides an insightful account of oli-
garchical tendencies among left-wing groups in Europe and the United States today that echoes
Michels. However, Hirschman (1970, 84) offers an analysis that undermines Michels’s claim of a
universal law. For Hirschman, oligarchy within small radical parties reflects ease of exit; members
prefer to leave rather than challenge a party’s leadership, because they have choices. In systems with
a few, large parties, members will have more incentive to exercise voice, and leaders may be made
more accountable.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
22 What Should We Do?
method worked well enough because the movement had momentum and deep
support.
However, when people disagree, a rule requiring consensus may favor sup-
porters of the status quo. Majority rule is an alternative that has natural appeal.
In the case of the MIA, genuine majority rule would have required putting con-
tested choices to a vote by the whole membership. But that method is not always
wise. In a majority-rule system, the majority can badly mistreat a minority—as
was the case in the state of Alabama as a whole. And when a vote is close, it maxi-
mizes the number of people who are dissatisfied. In a voluntary association like
the MIA, dissatisfied people may exit, weakening the organization. The founders
of the MIA demonstrated civic skill by gently moving Rev. Bennett into second
place below Dr. King without forcing a vote on the matter, which would have
been divisive.21
In fact, it has been proven that no system for turning votes into decisions can
simultaneously meet several obvious criteria.22 In practical terms, the main im-
plication of this finding is that we must often choose between voting systems
that force a two-sided choice (like referenda or two-party elections) or voting
systems that allow many to win small shares of power (as in Israeli elections). The
former systems disadvantage anyone who is dissatisfied with the forced choice.
The latter can lead to stalemate or unpopular minority rule. The choice of rules is
somewhat arbitrary and yet may determine the outcome.
Boundary problems: In the 1960s, Stokely Carmichael (later Kwame Ture)
began to emphasize that Blacks actually represented a majority in many
Southern counties and Northern cities. Voting rights would therefore enable
what Carmichael called (using a term that already had a long history) “Black
Power.”23
In this instance, Carmichael focused on the fact that the boundaries of
American political jurisdictions sometimes favored African Americans, if they
could vote. However, by the same logic, the boundaries of US states favored
Whites, since they then constituted a majority of every state except Hawaii. And
the United States as a whole had a White-majority electorate, which has posed a
consistent challenge for Black Nationalists in this country.
The generic problem is that political outcomes are relative to the boundaries
that we draw around communities. This is why the drawing of electoral districts
is so important to outcomes. Similar problems arise at larger scales as well. The
proper status of Northern Ireland would be decided differently if the decision be-
longed to residents of that province, residents of the island of Ireland, residents of
21 For a similar example with a theoretical explanation, see Mansbridge 1980, 67.
22 Arrow 1950.
23 E.g., Carmichael 1966.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
The Montgomery Bus Boycott 23
the United Kingdom, or residents of the European Union. The main point of de-
bate is which population should determine sovereignty. Such problems can arise
not only for political jurisdictions within which people vote on laws and leaders,
but also for other boundaries, such as school catchment zones, parishes and dio-
ceses, and even the membership bodies of private associations.
Principal/agent conflicts: An agent is someone whom a principal employs to
take care of her interests. For instance, I employ my dentist to take care of my
teeth. But my dentist’s interests may diverge from mine; she may want me to have
expensive surgery that I do not need.
The same divergence can occur in political contexts. The state of Alabama pre-
sented itself as an agent of all its citizens, which was clearly false with respect to
Black Alabamans and plausibly untrue for many Whites as well. Perhaps cases
like this are not best understood as principal/agent dilemmas, because the ex-
ercise of raw power is so evident. But the civil rights movement provides clearer
examples. Leaders like King sought to be agents of their community. This was
their genuine goal, but they did not always accomplish it in practice. They might,
for example, have valid reasons to try to protect their own reputations, influ-
ence, access to powerful politicians, and budgets so that they could continue to
lead, and that desire might cause their interests to diverge from the people they
tried to represent. Without naming individuals, Ella Baker criticized movement
leaders who “get so involved with playing the game of being important that they
exhaust themselves and their time, and they don’t do the work of actually orga-
nizing people”24 As with free-riding, the Iron Law of Oligarchy, and problematic
decision rules and boundaries, this challenge is intrinsic and not reducible to bad
choices by individuals.
Discourse Problems
Under segregation, race was deeply encoded in language, even in the way people
addressed each other. In the North, too, Americans talked about race in ways
that reinforced white supremacy.25 One of goals and accomplishments of the
civil rights movement was to change how we talk and think about race. But the
movement also had certain ways of presenting its own values that can be con-
tested. I have cited King’s presentation of Parks as a tired Christian worker in-
stead of a seasoned campaigner against rape. Another example is the frequent
use of patriotic and religious language by King and other leaders of the SCLC.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
24 What Should We Do?
Although I interpret their rhetoric as authentic rather than merely strategic, and
I see powerful arguments in favor of it, it is certainly contestable.
Put more generically, civic actors confront discourse problems, many of which
we can categorize as ideology, propaganda, and polarization.
Ideology: This word can be defined in various ways, but here I mean a set of
assumptions that distort one’s other beliefs. Jason Stanley writes that everyone
has an ideology, but “flawed ideologies” are those that serve as “persistent bar-
riers to the acquisition of knowledge.”26
Racism is an ideology, in this sense. We could also cite less objectionable
examples and cases in which ideology besets people on both sides of a valuable
debate. For instance, some people believe that the United States was once a wel-
fare state with a social safety net that has been badly frayed because of neolib-
eralism. And other people believe that the United States was once a society of
free, self-reliant entrepreneurs that is becoming socialist. They cannot both be
right, but they could both be wrong. The federal government of the United States
has controlled a fluctuating proportion of between 15 and 25 percent of Gross
National Product since 1955, with no evident long-term trend during that pe-
riod.27 That is just a discrete fact, but it complicates narratives of decline from
right and left alike. When adherents of any given view will not adjust their posi-
tion in response to inconvenient facts, we have a problem of ideology.
Propaganda: Although it would be difficult to quantify the degree to which
segregation relied on Confederate statues, Gone with the Wind, and textbooks
that told a sanitized version of American history, such forms of communication
may well have reinforced the system. More generally, it is possible to influence
public opinion by deploying resources. One can pay for advertisements, hire
writers and artists, erect monuments or billboards, and collect data to determine
which message will be effective with which audience. Money and power can shift
what people believe and value. When money is widely deployed to persuade,
it becomes unclear what people authentically value, let alone what they should
value.28
Polarization: Most Black and White Alabamans held dramatically divergent
views of the social system. This divergence was not the problem; racism was.
Whites needed to learn how things looked to African Americans. However, di-
vergence can be an intrinsic problem. Numerous studies have found that groups
on one general side of an issue will migrate toward more radical opinions when
or purpose. Hobbs 2020 suggests that calling a communication propaganda can be subjective and
depends on context, but she emphasizes the criteria of mass scale and intersection with entertain-
ment (8–9).
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
The Montgomery Bus Boycott 25
they form an enclave of like-minded people.29 Groups that span a wide spectrum
of opinion will often polarize into relatively radical opposing subgroups. One
of several reasons for this is that individuals want to be accepted into groups of
like-minded peers. Insofar as this problem arose in the civil rights movement, it
was most evident once the major groups began to splinter over funding, credit,
strategic choices, and issues like Vietnam.30 These disagreements were substan-
tive and valid, yet members of each component organization—from the NAACP
to SNCC—began to view the other groups in somewhat stereotyped terms and
intentionally chose divergent approaches.
29 A recent contribution to this literature is Gabbay, Kelly, Reedy & Gastil 2018. These authors find
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
26 What Should We Do?
32 In a binomial logistic regression using a national survey conducted by the Tufts University
Priority Area Research Group on Equity in Health, Wealth, and Civic Engagement with Ipsos
(n =1,267), I find that identifying as African American raises the odds of reporting that “you have
been unfairly stopped, searched, questioned, physically threatened or abused by the police” by 4.6
times when controlling for self-reported gender, education, age, English-language proficiency,
household income, housing type, and any mental health diagnosis, as well as Census estimates of me-
dian income at the county level.
33 Wimmer 2008, 975.
34 Barth 1969 begins his influential account with examples like this, before arguing that many
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
The Montgomery Bus Boycott 27
boundary would have failed for a similar reason, so Whites developed a new cat-
egory based on the convenient visual cue of skin color. The path from religion
to race went through an intermediary stage, the idea that some people could
never become full Christians—what Rebecca Anne Goetz calls “hereditary
heathenism.”35
Today, race is deeply ingrained. Many Americans identify a photograph of a
face with a racial category and have a consequent emotional reaction within one
tenth of one second after seeing it.36 Although race is socially constructed, it does
not follow that it is easy to reconstruct or deconstruct. It is now a powerful reality.
Identity groups (groups that define who people are rather than what they do or
think) often remain durable over time, notwithstanding gradual evolution due to
births and deaths, conversions and excommunications, marriages and divorces.
But groups’ beliefs and behavior often change, as do the rationales for the bound-
aries among them. Identity boundaries may not mark dramatic differences or
gaps in communication. On the contrary, people often accentuate boundaries
when they perceive interactions and similarities across the line.37 The Black/
White boundary in America is a perfect example: interactions have often been
intimate and pervasive, and both groups often share the same religion, language,
and many other aspects of culture, yet the color line has been profoundly impor-
tant. Boundaries reflect the incentives created by institutions (such as states and
markets), power differentials, network ties, and path-dependence, among other
factors.38
A wide range of criteria can be used to create identity distinctions, of which
skin color is only one example. Sometimes, a third party can be hard pressed to
detect a difference that is highly salient to the people on both sides of a boundary
that structures their interactions.39 There can be disagreements about where a
line lies, about what it entails, or about both. For instance, over the course of
US history, conflict has frequently focused on whether the Black/White binary
does, may, or should mark a disparity in status and rights. But disagreement
about where the line lies has been relatively rare. With some exceptions, African
Americans and Whites typically share the same view of who is Black or White.
That is notable because the racial boundary is defined quite differently in other
countries and did not exist until the 1600s.40
One response would be to question the very concept of identity.41 However,
that may be futile, given the pervasive use of identity boundaries to structure
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
28 What Should We Do?
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
The Montgomery Bus Boycott 29
will be effective in all instances of any category of problem. The most successful
strategies will vary and will require experience and judgment to select.
Each of these problems also has a normative dimension (using the word “nor-
mative” in its philosophical sense, to mean differences between ideas or actions
that are fair, just, or good versus those that are unfair, unjust, or bad). We need
normative judgment to decide whether situations that fit the criteria I have
described in this chapter are actually problems, and what would count as solu-
tions. For instance, firms that must compete for customers face what they could
plausibly call a collective-action problem. A cartel would solve their problem. It
would require an agreement for mutual accountability, coordination, and mon-
itoring. Yet we have good reasons to denounce cartels. We may see price-fixing
as a threat to the public interest (since a few producers will profit at the expense
of many consumers) or as a form of cheating (since competitors should demon-
strate fair play by competing over prices), or both. One could perhaps defend
business cartels under certain circumstances. For our purposes here, the point is
that calling something a “problem” almost always requires a normative argument
as well as a technical analysis.
The three sets of problems are connected. Yet none of the three is, in my view,
reducible to another. A very ambitious game theorist might assume that all
breakdowns of human interaction are problems of collective action.47 But game
theory takes as givens the values and identities of the players. In general, identi-
ties are not fixed but are created and changed by discourse. For example, a swath
of countries across the Middle East is currently embroiled in conflicts between
Shiites and Sunnis. If one takes the existence of those two opposing groups for
granted, then game theory may be helpful for explaining certain developments,
such as the arms race between Saudi Arabia and Iran and the tit-for-tat sectarian
violence against civilians. But identification with these two groups is a variable,
not a constant. Adam Gaiser writes that identifying with an Islamic sect is a “dy-
namic and conscious process of adoption, maintenance, and manipulation of
certain types of narrative identities in particular places and at particular times by
particular persons or groups of persons.”48 Other identities apart from sects (e.g.,
Islam as a whole, Arab versus Persian nationality, or pan-regional solidarity) can
be salient instead of the sectarian split. How people identify is a choice that can
be influenced by discourse.
In fact, some cultural critics may argue that all of our problems arise from
discourse, which produces our values and identities. But even if we all sin-
cerely agreed about something as important as climate change, problems of
47 An example is Tarko 2017, 7–11, which I mention because of its subject. It is a fine book, but
I disagree with Tarko that we only have institutions because of transaction costs or that Ostrom
would have thought so.
48 Gaiser 2017, 62.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
30 What Should We Do?
collective action would still be formidable. We would not only face the mother of
all tragedies-of-the-commons but also challenges of path-dependence (e.g., our
reliance on internal combustion engines), boundary issues (the countries that
use the most carbon are separated from the big producers), and indifference to
people we view as alien.
Nor is it plausible that we would automatically reach consensus about values
and coordinate our actions if we could eliminate power imbalances based on
identities. Seeking an example of a community in which everyone shares the
same identity, I considered Mount Athos, which is now an autonomous monastic
republic within Greece and the European Union. Every resident on Mount Athos
must be an Orthodox man; even female animals (with the exceptions of insects,
cats, and birds) have been banned for centuries. Thus there are no distinctions
of religion or gender on Mount Athos. The monks do represent different nation-
alities, but that does not seem to have been the root of the problem in 1913. The
issue then was whether the Name of Jesus as God is the same as God Himself.
The monks split on this question in a 661–517 vote (with many abstentions) that
provoked “not only verbal but physical violence.”49 Although one side was led by
a Russian noble, the tsar of Russia felt it necessary to send gunboats and troops to
support the other side, ordering the arrest, excommunication, and exile of hun-
dreds of monks. In game-theory terms, we would say that two sets of players had
reached a destructive impasse that was settled by an outsider. The situation was
highly problematic even though issues of identity were irrelevant.
A subset of these three sets of problems will typically confront any concrete
group of human beings who aim to influence their worlds. In many circum-
stances, the various problems will closely interrelate, causing webs or cycles of
challenges.
49 Larchet 2016.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
The Montgomery Bus Boycott 31
and despises” or else finds reasons to discount all evidence on their other side.
Later in the same paragraph, Bacon also recognizes what we now call positivity
or optimism bias: “it is the peculiar and perpetual error of the human intellect to
be more moved and excited by affirmatives than by negatives; whereas it ought
properly to hold itself indifferently disposed toward both alike.”50
Although some theorists have long been interested in such cognitive limita-
tions, a model of individuals as rational utility-maximizers was essential to clas-
sical and neoclassical economics. These disciplines made strong assumptions
about people’s ability to understand complex situations and predict the outcomes
of their private choices. Beginning in the late 1960s, Amos Tversky and Daniel
Kahneman shook that model with experiments that demonstrated that human
beings use “simplifying heuristics rather than extensive algorithmic processing”
to make decisions. Individuals err in predictable ways even when they want to
think rationally.51
Tversky and Kahneman’s program spread across the behavioral sciences and
constantly discovers new biases that are predictable enough to bear their own
names. Attribution Bias means explaining one’s failures as the results of difficult
external circumstances, while others’ failures must flow from their bad choices.
The Bandwagon Effect is a tendency to endorse ideas that many others are ex-
pressing. The Control Illusion is the tendency to overestimate how much we
control events. One could continue an alphabetical list through to the Zeigarnik
Effect (we remember interrupted tasks better than completed ones).
Two other clusters of limitations are central enough to the concerns of this
book that they deserve their own brief summaries:
Implicit bias: A large body of experiments shows that people hold implicit
negative attitudes toward various categories of human beings—attitudes that
influence their explicit decisions even when they sincerely deny holding them.
Dozens of experiments have tested hiring managers’ reactions to fictitious
résumés that are identical except that the applicants’ names are common for ei-
ther Whites, Blacks, or Latinos. Identical résumés with White-sounding names
have received 36 percent more callbacks than those with Black-sounding names,
and 24 percent more than those with Latino-sounding names, with no evident
change in these disparities since the first study in 1989.52 This is a fairly pure and
unambiguous example of bias, although an unknown proportion of it may be ex-
plicit and intentional rather than implicit. Similar forms of bias also arise when
periment with nonprofit job openings in Chicago and Boston, colleagues and I found that Latino-
sounding names drew more callbacks than whites, followed by Blacks. Levine, Kawashima-Ginsberg,
Benenson & Hayat 2018.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
32 What Should We Do?
human beings impose other categories on each other: gender, religion, sexual
orientation, and many others.
Motivated reasoning means that we are good at selecting and emphasizing
facts that support our predetermined ideas, and equally good at marginalizing or
debunking facts that complicate or challenge those ideas. For example, as people
obtain more education, their opinions of climate change correlate more with
their political ideology. Conservatives become less likely to believe in climate
change, and liberals more so, the more education they have.53 Well-educated
conservatives are sophisticated enough to recognize that accepting evidence of
climate change would challenge their economic views, so they use mental tech-
niques (also exhibited by liberals on other topics54) to reject or marginalize the
evidence. Deliberate efforts to debunk myths actually reinforce the same myths
because people hear the information selectively. That “backfire effect” is a strong
illustration of the more general tendency of human beings to assimilate new in-
formation with bias.55
These phenomena are held to be deeply rooted in the cognitive limitations of
human beings.56 Not only has the burgeoning literature on cognitive biases chal-
lenged market models, but it undermines the “folk theory” of democracy taught
in civics textbooks and widely believed by citizens and pundits, which holds that
“Ordinary people have preferences about what their government should do.
They choose leaders who will do these things, or they enact their preferences
directly in referendums. In either case, what the majority wants becomes gov-
ernment policy.”57 Citing the research on human cognitive limitations as well as
other evidence, Achen and Bartels argue that this folk theory is not only false as a
description of actual politics in the United States; it is impossible.
All the problems briefly summarized here raise serious doubts about whether
citizens can reason together about what they should do and then act collabo-
ratively. However, groups of people do solve human problems. They do build
institutions and norms that improve life. They build tools—daily newspapers,
adversarial courtrooms, secret ballots—that mitigate the limitations of human
reason and allow us to discuss, coordinate, and form functional groups. Every
decent government, association, neighborhood, and network is a triumph of rea-
sonable hope over chastened experience. Some citizens who ask, “What should
we do?” have learned how to overcome these problems.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
3
Three Traditions in Search of Solutions
No single thinker can answer the question, “What should we do?” Armchair
theorizing cannot reveal satisfactory solutions to problems as thorny as free-
riding, motivated reasoning, or quiescence. Nor will it suffice to study the re-
cord of past and current solutions that have already been invented by practical
people such as activists and political leaders. Often, to motivate a solution,
a new idea is needed: theory must provide the inspiration. However, an idea
rarely succeeds when it is first tested. To make any social experiment work,
we must refine and improve it until the hypothesis becomes right. The original
idea combines—at least implicitly—empirical evidence, normative principle,
strategic insight, and design. It almost always incorporates previous theories
and experiments. But the new idea only becomes right if it leads to experimen-
tation and critical reflection that ultimately yield desirable outcomes. Because
this cycle requires the contributions of many people over decades, we should
seek guidance from relevant theory/practice traditions, not from individual
authors alone.
An everyday example is public schooling. Nineteenth-century reformers like
Horace Mann drew from previous thinkers—and from successful experiences
in countries like Prussia—to propose a new idea: every child should be educated
at the government’s expense in a state-funded common school. Since then, not
only have educators and policymakers refined and revised most aspects of public
schooling, but scholars have critically evaluated actual schools from a wide va-
riety of perspectives. A few observers have concluded that Mann’s core idea was
misplaced,1 but most see their role as helping to make his vision become suc-
cessful. Public schools did not arise like a new species in Darwinian evolution
to survive or fail on its own. Nor did Mann propose a hypothesis that could be
tested with a single experiment (e.g., “Common schools will work”). Rather,
1 E.g., the libertarians Milton and Rose Friedman argue that a mandate is unnecessary and the
public school monopoly poorly serves children: “We believe that the growing role that government
has played in financing and administering schooling has led not only to enormous waste of taxpayers’
money but also to a far poorer educational system than would have developed had voluntary coop-
eration continued to play a larger role.” (Friedman & Friedman 1980, 187) From a very different per-
spective, Illich 1971 argues that schools exist to indoctrinate children into the assumption that they
must obey institutions.
What Should We Do?. Peter Levine, Oxford University Press. © Oxford University Press 2022.
DOI: 10.1093/oso/9780197570494.003.0003
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
34 What Should We Do?
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
Three Traditions in Search of Solutions 35
between good and bad systems? Compared to the Bloomington School, they
were more explicit and self-conscious about normative questions and more in-
terested in discourse and public reasoning, although much less helpful about
problems of collective action.
In my reading, the first generation of the Frankfurt School offered scant so-
lutions to the problems they insightfully mapped. After the Second World War,
Habermas refocused the School on making distinctions between good and bad
discourse and understanding the social conditions of desirable deliberation. In
the process, he has moved even farther from Marxism, although he has retained
a critical edge. One could say that for Habermas, a good citizen is one who delib-
erates well with others and makes deliberation possible.
Habermas is certainly an academic theorist, but practitioners have attempted
many practical experiments with public deliberation in the past four decades.
The tradition that I will explore in chapter 5 consists of the thought of Habermas,
responses from his key critics, and this array of practical experimentation.
The third tradition involves mass social movements that are nonviolent or that
sharply limit their participants’ use of violence. A pivotal figure in this tradition
is M. K. Gandhi. He drew on theoretical works, including the late writings of
Tolstoy, the Bhagavad Gita, and Henry David Thoreau, as well as practical initia-
tives such as the civil disobedience of British suffragists. But he also innovated,
both theoretically and practically, calling his life My Experiments with Truth.3
The Gandhian tradition has radiated to many other countries, and I will focus
on its return to the United States via the civil rights movement and its echoes in a
set of self-limiting popular uprisings since 1989 known as the Color Revolutions.
One could say that for Gandhi, the ideal citizen is the satyagrahi, the bearer of
truth-force.
I will argue in c hapter 6 that the Gandhian tradition is more useful than the
other two when clashes of identity are combined with great disparities of power,
as was the case when the British ruled India and wherever Whites have domi-
nated African Americans. This tradition is about movement, while the other two
theories are more comfortable with stability. It also offers a distinctive set of nor-
mative ideals of wide application. It is less useful than the other two traditions,
however, for designing and refining durable institutions.
After discussing each of these three traditions in its own chapter, I will con-
clude with the outline of a synthesis that uses each tradition to compensate for
the weaknesses of the others.
3 Gandhi 1927.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
36 What Should We Do?
persons who take part in it, and where the realization of the good is such as to effect an energetic
desire and effort to sustain it in being just because it is a good shared by all, there is in so far a com-
munity. The clear consciousness of a communal life, in all its implications, constitutes the idea of
democracy.” Dewey 1927, 149.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
Three Traditions in Search of Solutions 37
beings had drawn the “well-attested conclusion from historic facts”—in short,
we had learned—that “government exists to serve its community, and that this
purpose cannot be achieved unless the community itself shares in selecting its
governors and determining their policies.”7 People had become democrats by re-
flecting on experience. Their future experiments would only deepen and enrich
democratic culture.
Even at that moment, a few wealthy and well-armed nations that were democrat-
ically governed by their own people were holding most of the rest of the world as
exploited colonies. Within two decades, millions had perished in senseless wars and
domestic slaughters that had popular support. The democratic current was hardly
steady. Here Dewey ignores the fact that human beings can “learn” all manner of
lessons, including the evil lessons that other people are subhuman and deserve ex-
ploitation, if not extermination. As Cornel West observes, Dewey drew from his op-
timistic American forebears Jefferson and Emerson, but he “failed to seriously meet
the challenge posed by Lincoln—namely, defining the relation of democratic ways
thought and life to a profound sense of evil.”8
Dewey’s optimism about learning and self-government led him to reject all limi-
tations on new institutions and procedures. He decried ideas like “the Constitution,
the Supreme Court, private property, free contract, and so on,” for which “the words
‘sacred’ and ‘sanctity’ come readily to our lips.”9 Everything should be open to ex-
perimentation. But if the problems I listed earlier are dangerous, then institutional
innovations can be harmful instead of helpful, and some of the forms that we have
built so far may be precious achievements. Perhaps we should block changes by cre-
ating rigid norms and rules, including written constitutions that we hold sacrosanct.
Because pragmatism is overly sanguine about progress, it fails to give much guid-
ance in a world rife with dangers.
During the 1930s and 1940s, Dewey grappled more explicitly with evil, partly as
a result of sharp critiques from the likes of Lewis Mumford (who thought pragma-
tism had weakened Americans’ will to resist tyranny) and Reinhold Niebuhr (who
charged Dewey with naive optimism).10 Like his younger self, the septuagenarian
Dewey was commendably open to learning from the experiences of those dark
years and from his critics. Still, I do not find in the late Dewey an adequate account
of how to prevent corruption, barbarism, or social collapse.
Unlike Dewey, Hannah Arendt was fully aware of evil. Writing in the shadow
of the Holocaust, from which she had herself escaped, she predicted in 1945 that
“the problem of evil will be the fundamental question of postwar intellectual life
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
38 What Should We Do?
in Europe.”11 Her thought is congenial to the spirit of this book because she un-
derstood that being an active citizen under conditions of injustice is part of living
a full life. She thought that Jefferson and other American revolutionaries had dis-
covered that “no one could be called happy without his share in public happi-
ness, that no one could be called free without his experience in public freedom,
and that no one could be called either happy or free without participating, and
having a share, in public business.”12 She saw that people who deliberate about
what to do and then act together in public are free, equal, and dignified in ways
that are impossible for those who are governed, even benignly. In common with
other civic republican thinkers, Arendt recognized that politics (the active en-
gagement of diverse people in public spaces) is a good in itself, not merely a way
of obtaining desirable outcomes, such as security and prosperity.
Arendt also contributed valuable ideas about the public domain as distinct from
the private sphere. She was especially helpful on this topic because the public/pri-
vate distinction had often been used to deprive women of public roles, but Arendt
was a tireless fighter for women’s liberation (and an important female public figure
herself) who believed that privacy should be reconsidered and protected rather
than abandoned.
Finally, Arendt offered insights about institutional design, derived from what
she considered her discovery of the “lost treasure” of the “revolutionary tradition.”13
During revolutions, she thought, people spontaneously construct small, self-gov-
erning units that run their own workplaces as well as their villages or neighbor-
hoods. Revolutionary leaders are threatened by these rival sources of power and
usually shut them down, but they permit dignified politics while they last.
There is, indeed, much to be learned from spontaneous revolutionary organ-
izations, but Arendt does not have enough to say about how they could be made
durable and expanded to a large scale. Although her words echo through this
book, I don’t treat her thought as a separate tradition on par with the three I that
explore more deeply.
A final tradition that deserves consideration derives from two ideas in
Aristotle. In chapter 6 of his Nichomachean Ethics, Aristotle uses the word
phronesis (practical wisdom) to refer to a trait of character that yields wise all-
things-considered decisions under conditions of moral and empirical uncer-
tainty. Phronesis is not rule-governed; there are no algorithms for it. It comes
from experience and from openness to the situation at hand. Phronesis is how we
choose the right means to pursue good ends.14
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
Three Traditions in Search of Solutions 39
Two chapters later in the same book, Aristotle develops the idea of friendship
(philia) in a particularly political or civic way, calling it the basis of the city. For
Aristotle, friends are people who value their relationships, not only the outcomes
or activities that they accomplish together. This sounds like a utopian ideal, con-
trary to the assumption that people are basically selfish or concerned only with
getting results that they desire. But we actually observe many situations in which
groups of people show as much concern for the quality of their relationships as
for the decisions that they make together or the outcomes that result.15 In turn,
citizenship is the interaction of civic friends who exercise phronesis together.
This account is broadly consistent with the Bloomington School, the
Frankfurt School, and the Gandhian tradition. Moreover, the phronetic aspect
of Aristotelianism has recently been revived by the influential Danish planning
professor Bent Flyvbjerg, who shook up social science when he argued that the
search for general, predictive rules was a “wasteful dead end.”16 Instead, he has
argued, social scientists should display phronesis in collaboration with laypeople.
They should ask of any specific situation: (1) “Where are we going?”; (2) “Who
gains and who loses and by which mechanisms of power?”; (3) “Is it desirable?”;
and (4) “What should be done?”17 Although Flyvbjerg states the culminating
question in a passive form, he is really asking social scientists and laypeople
to decide together what they should do. And although he asks whether a phe-
nomenon is “desirable,” rather than whether it is good or just, his third question
makes room for normative analysis.
I find this tradition valuable, but I have chosen not to devote extensive space to
it because of certain limitations. For one thing, Aristotle’s discussions of phronesis
and philia occupy just a few hundred suggestive words in the Nichomachean
Ethics. Even if we decide to use these words as slogans, we need a lot more detail
about what they should mean for us in practice. For instance, civic friends will
have to overcome free-riding; but how to do that is not explained in Aristotle.
Second, Flyvbjerg is relentlessly skeptical about general rules. “No predic-
tive theories have been arrived at in social science, despite centuries of trying.”18
Flyvbjerg evokes phronesis as an alternative to the vain search for general laws of
human behavior. But the Bloomington School demonstrates that we can usefully
generalize about human behavior and use the resulting principles and predic-
tions to improve our odds of success as citizens. The Ostroms avoided simplistic
generalizations, but they sought to generalize carefully as a way of improving de-
cisions. If that approach is consistent with Aristotle’s phronesis, then his position
15 Mansbridge 1980.
16 Flyvbjerg 2006b, 38.
17 Flyvbjerg 2010, 145ff; cf. Flyvbjerg 2006a, 374.
18 Flyvbjerg 2006b, 38.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
40 What Should We Do?
is moderate and unassailable, but not very challenging. By the same token,
Flyvbjerg’s more radical version seems implausible.
Finally, although Flyvbjerg is impressively attuned to imbalances of power,
I am not sure that either he or Aristotle has enough to say about large-scale in-
justices that arise from clashes of identities. Aristotle thought of civic friends as
well-born Greek men in small cities. Flyvbjerg’s best case is his own successful
campaign against an unjust municipal zoning campaign.19 Impressive as this ef-
fort was, it succeeded within a modest-sized Danish city that had a free press,
independent courts, substantial social equality, and a responsive national gov-
ernment. Gandhi could count on few of those advantages.
A Focus on Means
The three traditions that I emphasize in this book share a focus on getting the
means right. They are much vaguer about their ultimate objectives or ends.
This puts them at odds with much political philosophy in the English-speaking
world since the 1960s, which has been focused on defining justice, understood
as an end-state, a goal. Philosophy seeks to define the objective of political ac-
tion as sharply as possible. Political ethics then involves a set of questions about
whether various means (e.g., civil disobedience, propaganda, compromise, or
violence) are acceptable—or necessary—when pursuing justice under various
circumstances.
A century ago, as Karuna Mantena notes, there was a more vibrant debate
about political means.20 The central question was not what constituted justice
but whether and when to use electoral campaigns, strikes, boycotts, assassin-
ations, or revolutions, among other options. Mantena reads Gandhi as a par-
ticipant in that debate who developed and defended nonviolence as a cluster of
strategies. Moreover, Gandhi explicitly argued that the best way to think about
politics was to determine the right means or strategies, not to pretend to define
justice as an end.
“Means are after all everything,” Gandhi wrote, in response to a group of
Indian political leaders who had issued an “Appeal to the Nation” in 1924. These
leaders had proposed a concrete ideal of justice: the immediate creation of a new,
independent “Federated Republic of the United States of India.” They argued that
this end justified a wide range of strategies. They wanted to “delete the words ‘by
peaceful and legitimate means’ from the Congress creed, so that men holding
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
Three Traditions in Search of Solutions 41
21 Gandhi 1924/1999, vol. 28, 307–10 (May 22, 1924–August 15, 1924). I owe the reference to
Mantena 2012a, 457. See also Gandhi 1921, 81: “The means may be likened to a seed, the end to a tree;
and there is just the same inviolable connection between the means and the end as there is between
the seed and the tree.” It is ironic that Jawaharlal Nehru later wrote, “If anyone wants to know what
India wants, let him go to Gandhi.” And still later, Gandhi said, “Nehru . . . has got ability, know-
ledge, and close touch with the public here and can interpret India’s mind.” Guha 2018, 666, 985. Each
founder of India was eager to declaim knowledge of India’s goals and attribute that knowledge to his
colleague.
22 Gandhi 1924/1999, vol. 28, 310.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
42 What Should We Do?
2012b, 11.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
Three Traditions in Search of Solutions 43
different cultural context and political sensibility from Gandhi’s, but he also re-
jects instrumental, means/ends reasoning in favor of creating groups of people
who endlessly make justice by interacting. It’s just that Habermas’s interactive
groups are highly critical, explicit, and discursive, whereas Gandhi’s weavers may
be literally silent.
Likewise, the Bloomington School focuses on how people can construct insti-
tutions for managing common resources. Paul Dragos Aligica, a major represen-
tative of the School, cites the distinction between “end state social theories” and
“process theories,” which “focus on the way ends states emerge.” He observes,
“While a large part of theorizing in political economy is more or less explicitly
‘end state’ oriented, the Ostroms’ research agenda seems to be closer to the ‘pro-
cess’ perspective.”28 Another phrase that could describe all of these schools is
“open-ended”—action that one undertakes with intentional uncertainty about
where it will conclude.29
Even if we know what justice is, the wisest direction is not always straight to-
ward it. For example, Christopher Winship acknowledges that improving the
justice of American education would require raising the quality of the schooling
available to the least advantaged students. However, he argues, “the best way to
approach serving the interests of the least well off [may be] to avoid policies that
decisively pit the interests of the less advantaged families against those of the
more advantaged families.” He cites evidence that Scandinavian countries have
achieved the highest levels of shared prosperity and economic equality in the
world not by directly pursuing equality but by negotiating policies that are at-
tractive to business as well as labor.30 Winship’s advice may or may not apply in
any given situation; the general point is that the best path to any given end is not
always direct.
In the movie Lincoln, the president tells Rep. Thaddeus Stevens:
A compass, I learnt when I was surveying, it’ll—it’ll point you True North from
where you’re standing, but it’s got no advice about the swamps and deserts and
chasms that you’ll encounter along the way. If in pursuit of your destination you
plunge ahead, heedless of obstacles, and achieve nothing more than to sink in a
swamp, what’s the use of knowing True North?
These are the words of the screenwriter Tony Kushner, not (as far as I know)
of President Lincoln himself. But they make an important point. Knowing
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
44 What Should We Do?
where we ought to end as a society tells us very little about our best next move.
Sometimes a tactical retreat or a sidestep is well advised.
I would push the point further. There is no end, no True North. As we move
through time as a community, we keep deciding where we ought to go. Moving in
the right direction is important, but so is holding ourselves together as a group so
that we can keep deciding where to go. We didn’t know a century ago that same-
sex marriage was a human right, but we maintained a republican form of gov-
ernment long enough to learn that. Sometimes, the imperative of maintaining
our ability to govern ourselves is more important than forward motion. Stephen
Elkin introduces this metaphor:
Those who wish to constitute a republican regime are like shipbuilding sailors
on a partly uncharted sea who know the direction in which they sail, since
the kinds of ports they prefer lie that way. This much they can agree on. To at-
tempt to agree on anything more specific will defeat them, their opinions on the
matter differing significantly. They also know too little for substantive agree-
ment to be possible. . . . It is clear that the relations among the shipbuilders are
fundamental. Because they must build, rebuild, repair, and modify the vessel
as they sail and learn—and because they must alter their course . . .—it mat-
ters whether the shipbuilders’ modes of association are such as to facilitate this
learning and the decisions they must make. . . . These modes of association are
then at least as important as the ports toward which the shipbuilders sail.
Elkin adds: the “essential problem is one of creating a design that provides the
capabilities that are needed to keep the regime oriented in the right direction.”31
Although our focus should be broader than regimes—we should consider all
forms of groups—Elkin offers good advice. Our main task is to join, sustain, and
(if necessary) design and find a diverse array of organizations that encourage
diverse people to work and learn together, oriented generally toward justice but
open to exploring what justice may actually mean.
A focus on means and a reluctance to specify ends does pose a risk. Whether
intentionally or inadvertently, a person might select and defend means that de-
termine a foreseeable outcome or that foreclose an outcome that others prefer.
This could be a back-door strategy for getting the ends that the person wanted
in the first place. To claim that you are too humble and aware of your own limits
to know the best ends is disingenuous if it’s clear where your favored means will
take you.
Muhammad Ali Jinnah, later the founder of Pakistan, made essentially this
complaint against Gandhi. Gandhi insisted that the social movement for Indian
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
Three Traditions in Search of Solutions 45
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
46 What Should We Do?
without forthrightly defending it? Do critics of those values have any reason to
accept the arguments of this book?
My response is that democracy, liberalism, cosmopolitanism, polycentricity,
peacefulness, and openness to change are vague terms, subject to many formu-
lations and degrees, and sometimes in mutual tension. (For example, the dem-
ocratic value of responsiveness to public opinion can easily conflict with the
liberal value of individual rights.) Human cognitive and motivational limitations
suggest that we should not try to define these values in advance or permanently.
Instead, we should adopt means that are decentralized (in the Ostroms’ sense),
deliberative (in Habermas’s sense), and peaceful (in Gandhi’s sense) and leave
the ends undefined. The actual outcome will predictably be somewhat demo-
cratic, liberal, and so on, but we cannot and should not know what that will mean
in detail.36
36 What does this conclusion imply about the argument between Gandhi and Jinnah? I would say
that Gandhi joined and then led an interfaith party for swaraj that encouraged debates about both
means and goals. Jinnah was a member of that party, albeit mostly before Gandhi’s arrival from South
Africa. Jinnah and others had the right to quit the party and movement. Exit is a legitimate choice in
movements and party politics. As a result of Jinnah’s exit, Gandhi’s means failed: Congress ceased to
be a forum for dialogue and cooperation that included the kinds of people who preferred the Muslim
League. But Gandhi’s failure doesn’t invalidate his general advice to focus on means rather than spe-
cific ends.
37 Vincent Ostrom uses the same analogy in V. Ostrom 1997, 282. See Fotos 2015.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
Three Traditions in Search of Solutions 47
biases uncertainty
However, academic disciplines are often taught and practiced in ways that
marginalize the citizen’s concerns. We can begin with a standard model of how
the social sciences ought to work (fig. 3.1).38 According to this model, a social
scientist aims at truth about how the world works. The chief questions are “What
is happening?” and “why?” From the very first day of a class on social science,
any responsible teacher will note that social science is beset by biases. Also, our
understanding of the world is inevitably uncertain. Therefore, “truth” deserves
quotation marks; it is always problematic. Still, social scientists use a large array
of sophisticated tools to reduce both bias and uncertainty and cumulatively ap-
proximate truth.39
The truth about what is happening and why does not tell anyone what to do;
facts are not judgments or prescriptions. To move from description to policy ad-
vice, we need something else: opinions, preferences, or values about what should
be. Those cannot come from social science, viewed as a science.40 They may
come from public opinion, or from policymakers, or perhaps from a different
discipline, such as philosophy. Wherever they come from, they are added to facts
to produce policy recommendations.
But policy recommendations do not jump off the shelf to implement them-
selves. Someone must take strategic action to implement any recommendation.
38 Cf. what Diem, Young, Welton, Mansfield & Lee (2014) call the “traditional model” of policy
analysis.
39 Portions of this section of text appeared as Levine 2017 and are reprinted here with permission.
40 “The majority of sociologists consider it illegitimate to move from explanation to evaluation.
In their view, sociology should strive to be value-free, objective, or at least to avoid making explicit
value-judgements.” J. Scott & Marshall 2009, 520.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
48 What Should We Do?
That someone is usually quite separate from the social scientist who originally
studied the truth about the social world. Policy recommendations plus strategic
action equals change in this standard model, which delineates a very specific role
for the analyst as an independent and value-neutral seeker after truth.
There are many problems with the model in fig. 3.1, and I will just sketch a few
major ones. To start with, norms or moral commitments are not mere biases.
If I say, “education is good,” I am not expressing an opinion that might bias my
analysis. I am proposing a truth, albeit one that needs more detail. (What kind
of education is good? For whom? Why?) How to ground or justify moral claims
is a complex question, but it is nihilistic to treat facts as objective and all moral
claims as subjective in the sense of arbitrary. There are better and worse moral
claims.
Meanwhile, data are— and should be— imbued with norms. When we
measure an educational system by calculating graduation rates or test scores, we
are claiming that these outcomes are valuable. A set of requirements for grad-
uation or a list of questions on an exam must reflect value judgments. There is
no such thing as a value-neutral measurement, nor should we aim for one.41
A thoughtful, humane assessment of students is much better than a measure-
ment system meant to uphold a social injustice. Social scientists should not only
study the values held by other people, and not only disclose the values that they
happen to hold as possible sources of bias, but also strive to identify and justify
the best values and employ those values in collecting and interpreting data.
Not only must values influence empirical information, but data also rightly
and inevitably influence values. After all, why do I think that education is good?
In part because we have more than a century’s experience with near-universal
schooling. It has not benefited every student, but it has been good enough for
enough children to support the ideal of making school available to all. Our expe-
rience influences this normative judgment.
Another reason for the norm in favor of education is that we have developed a
strategy for making schooling universal. We know how to pass laws that require
enrollment and how to fund a system of public education. “Everyone should be
educated” is not an empty slogan or a utopian ideal that might have terrible un-
anticipated consequences when put into action. It is a policy framework that
has had decent results in practice, although it needs constant review and im-
provement. In contrast, the ideal that every human being should have highly re-
warding work may be appealing in the abstract, but it is not clear what it means
for policy. In general, our norms are, and ought to be, influenced by what we
know about how to implement them.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
Three Traditions in Search of Solutions 49
And why do we have data on students’ performance and graduation, but not
on many other topics that might be equally important, such as their happiness?
Because people have organized to compel institutions to collect certain kinds of
data, but not other kinds. In that sense, strategy influences the empirical evi-
dence that we have.
In asking “What should we do?,” a citizen must consider facts, values, and
strategies in an integrated way, for they are deeply interpenetrated and practi-
cally inseparable. Ruth Anna Putnam writes, “facts and values emerge at the end
of inquiry on an equal footing: the facts are value-laden, and the values are fact-
laden.”42 It is consistent with her pragmatist background to add strategies as a
third term. But the division of labor within the academy—reflecting conceptual
distinctions between values and facts and between theory and practice—separ-
ates the different aspects of the citizen’s question in harmful ways. To be sure, no
individual is capable of putting all the aspects back together. To address values,
facts, and strategies together requires cumulative, collaborative work. That is
why I write of whole traditions, not individual authors, that unite these three
topics. Citizens need models that integrate strategy, empirical evidence, and nor-
mative argument much more thoroughly.
Not only is academic research compartmentalized, but it tends to be cynical
in a particular way. Because the social sciences focus on the explanation of so-
cial phenomena, they often conclude that intentional acts are not particularly
significant. Market outcomes and shifts in populations and technologies often
have greater causal impact than the deliberate choices made by members and
leaders of groups. A common distinction is structure versus agency, and often
structure emerges as the more significant explanation. For instance, the con-
dition of the economy in the last six months seems to affect the outcome of an
election more than any candidate’s choice of policies or strategies.43 However,
the factors that explain outcomes are not necessarily the factors that should
matter most to us as actors. For us, the important question is our own agency:
what we can accomplish by acting in various ways.44 Only if we have zero po-
tential impact should we give up, and even then, we should devote our atten-
tion to a different problem rather than ceasing to act at all. The social sciences
tend to ask, “What is happening and why?” or “What would happen if this
were done?” rather than what is best for any particular group of citizens to try
to do.45 To focus on what we should do as members of groups requires more
focused inquiry into how groups work.
or ‘What is good for us in the long run and on the whole?’) from everyday life in an unmediated way,
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use
50 What Should We Do?
treating these questions without the objectivating filter of social science.” We need a social science of
what we ought to do that objectively investigates the conditions under which groups work.
46Rawls 1971, 3, 7, 8.
47Rawls 1999, 90. Defending Rawls’s view, Swift 2008, 382, writes, “only by reference to philos-
ophy—abstract, pure, context-free philosophy—can we have an adequate basis for thinking about
how to promote justice in our current, radically nonideal circumstances.”
48 Pettit 2000, 207, 129.
EBSCOhost - printed on 5/3/2024 11:18 AM via UNIVERSITY OF CINCINNATI. All use subject to https://www.ebsco.com/terms-of-use