Rogers Brubaker, Grounds for Difference
What is the relations between difference and inequality? I want to approach this deceptively simple
yet formidably abstract question by way of a thought experiment. Consider a world characterized—like
our own—by both horizontal and vertical social divisions (Blau 1977: 8–9). On a horizontal plane,
people categorize themselves and others according to a logic of significant similarity and difference.
They identify with others whom they see as similar in some meaningful way, and they distinguish
themselves from others whom they see as significantly different—in ethnicity, nationality, citizenship,
language, religion, gender, sexuality, taste, temperament, or the like. On a vertical plane, people can be
ranked according to whether they have more or less of some generally desired good: more or less
wealth, income, education, respect, health, occupational prestige, legal rights, basic existential security,
or the like.
Now imagine—and here’s where the thought experiment comes in—that horizontal categories and
vertical rankings were entirely independent of one another. The horizontal categories into which
people sort themselves and others—groupings based on ethnicity, religion, or musical taste, for
example—would not differ systematically by income, wealth, education, and so on. Differences of
income, wealth, and education would be differences within social categories, not between them. Members
of different categories would have the same chances of being ranked high or low on any vertical
dimension.
In this hypothetical world, difference would have no bearing on inequality. People would be
different, and they would be unequal; but the mechanisms that generate inequalities would be
unconnected with the processes through which people sort themselves and others into categories
based on similarity and difference. The mechanisms that generate inequalities would be difference-
blind: who is what would be independent of who gets what.
This is evidently not the world we inhabit. In our world, differences of race, ethnicity, language,
religion, gender, sexuality, citizenship, and so on do have a systematic bearing on inequality. But
how? This is the question I address in this chapter, focusing on the ways categorical differences—
differences that are organized, experienced, and represented in terms of discrete, bounded, and
relatively stable categories (such as black and white, Sunni and Shiite, male and female, citizen and
foreigner)—are implicated in the production and reproduction of inequality.
These and other ascribed categorical differences are not intrinsically linked to inequality; different does
not necessarily imply unequal. The relation between difference and inequality is contingent, not
necessary; it is empirical, not conceptual. And the degree to which and manner in which inequality is
structured along categorical lines vary widely over time and context. Certain categorical differences that
were once pervasively implicated in regimes of inequality—such as distinctions among Protestants,
Catholics, and Jews and among certain ethnic categories in the United States—are no longer so
implicated today. And a wide range of legally mandated forms of categorically unequal treatment has
been delegitimized throughout the developed world in a remarkably short span of time. To study the
relation between difference and inequality is to study historically situated social processes; it is not
to identify timeless truths.
I begin by critically engaging Charles Tilly’s influential account of how categories of difference are
implicated in the generation and maintenance of inequality. Taking issue with Tilly’s claim that major
categories of difference work in fundamentally similar ways, I consider in subsequent sections how
citizenship, gender, and ethnicity—broadly understood as including race as well as ethnicity-like forms of
religion—contribute to the production and reproduction of inequality in quite differing ways. I return in
the penultimate section to a more general level of analysis and outline three general processes through
which categories of difference work to produce and sustain position-mediated inequalities: the allocation
of persons to reward-bearing positions; the social production of unequally equipped categories of
1
,persons; and the social definition of positions and their rewards. In the final section, I discuss ways in
which inequalities not only are mediated by reward-bearing positions but also—notably in the case of the
social distribution of honor—attach directly to categories of persons, independently of the positions they
occupy. I suggest in closing that even as inequality has increased dramatically in certain respects in
recent decades, it has assumed forms that are less strictly categorical.
Tilly on Categorical Inequality
The theory of categorical inequality Tilly developed in Durable Inequality (1998) focuses on
organizations—firms, hospitals, universities, and states, for example—as key sites of inequality.
Organizations are key because inequalities of wealth, income, prestige, and even health and basic
physical security are increasingly mediated by positions in formal organizations. Jobs are the obvious
example of such positions. Income inequality in the United States depends primarily on unequal rewards
from jobs rather than unequal holdings of capital assets. Today’s rich are not rentiers; they are the
“working rich” (Saez 2013; Godechot 2007): highly paid employees and entrepreneurs. Tilly’s account
focuses primarily on how inequality is generated through linked and bounded clusters of jobs to which
sharply differing rewards are attached. But positions in organizations structure inequality in other ways as
well. Citizenship, for example, is a position in an organization (the modern state); as I show below,
it profoundly shapes life chances on a global scale, structuring access to vastly different rewards
and opportunities.
Durable inequality, on this account, turns on the matching or pairing of internal organizational
categories with pervasively available external categories. Internal categories designate unequal
positions (or clusters of positions) within an organization, differentiated by some combination of
remuneration, authority, working conditions, and mobility opportunities. Examples include enlisted
soldier and officer, doctor and nurse, executive and secretary, and the like. External categories are those
that serve as major axes of distinction and inequality in the wider social environment, around which
cluster scripts and stories that explain and justify the inequalities. Examples include gender, race,
ethnicity, citizenship, religion, and education.
Tilly shows how external categories are “imported” into organizations along with scripts and local
knowledge—shared understandings (or stereotypes) about the incumbents of those categories. He gives
particular attention to the “matching” of internal and external categories: the processes through which
positions in organizations are allocated such that major internal categorical divisions (between
executive and secretary, for example) coincide with major external categorical divisions (between
men and women, for example).
This is an original and fertile way of thinking about the organizational dimension of durable inequality.
But while Tilly’s account of the mechanisms that sustain durable inequality is richly suggestive, it is also
elusive. Probing the ambiguities in Tilly’s account can bring into sharper focus the social processes
through which categorical differences are implicated in the production and reproduction of inequality.
Categorical inequality, for Tilly, is generated in the first instance by two mechanisms: exploitation
and opportunity hoarding. Exploitation “operates when powerful, connected people command
resources from which they draw significantly increased returns by coordinating the effort of outsiders
whom they exclude from the full value added by that effort” (1998: 10). As the last clause of the
definition suggests, this notion of exploitation—like the Marxist notion—would seem to depend on a
theory of value. But Tilly neither endorses the notoriously problematic Marxian labor theory of value nor
proposes an alternative. His notion of exploitation remains informal, resting on a commonsense
understanding of powerful people coordinating the labor of outsiders and reaping the benefits of
that labor.
2
, The reference to “outsiders” suggests that categories of difference are implicated in processes of
exploitation. Tilly illustrates this by analysing the exploitation of Africans in South Africa under
apartheid and of women in capitalist labor markets. While duly noting the evidently sharp differences, he
argues that exploitation works through analogous causal processes in the two cases (1998: 136). The key
in both cases is matching between major organizational divisions and external categorical pairs
(White/ African and male/female). Such matching is said to facilitate exploitation. The reasons for
this are not fully spelled out, but the argument seems to be that matching stabilizes regimes of inequality
and lowers the cost of maintaining them.
The matching processes that implicate race in South Africa under apartheid and gender in capitalist labor
markets may be analogous at a certain level of abstraction. But they differ sharply in both degree and
kind. Racial categories in South Africa under apartheid were constructed from above, legally defined,
formally administered, and coercively enforced. They are not easily subsumed under Tilly’s notion of
“external categories”—categories that are pervasively available in the wider environment and “imported”
into organizations along with scripts and stories. Racial categories were of course pervasively available in
South Africa prior to the construction of the system of apartheid. But the available categories were
radically reconstructed, codified, and formalized by the state in a gigantic top-down exercise in
authoritative categorization. The processes through which racial categories were matched with
economic position were directly political, legal, administrative, coercive, and formalized. The
processes through which gender is matched with positions in capitalist firms, by contrast, are loose,
informal, probabilistic, decentralized, and mediated through individual-level self-understandings,
occupational aspirations, and human capital endowments; and the degree of matching is also much lower.
Tilly identifies “categorical exclusion” as a key element of his general analytical model of exploitation
(1998: 128–132). This might seem to imply exclusion on the basis of categories of difference like race,
gender, or citizenship, as in the examples he discusses at length. But there is an equivocation here.
Categorical exclusion involves “boundaries between unequal and paired categories in which members of
one category benefit from control of sequestered resources and receive returns from the other’s output”
(1998: 131). But what are the “unequal and paired categories”? They may simply be internal categories,
defining unequally rewarded clusters of positions within an organization (manager and worker, doctor
and nurse, or officer and enlisted soldier). Or they may be external categories (such as race, gender, or
citizenship) that are matched (to differing degrees and through differing processes) to the internal
categories. Tilly highlights the latter configuration in his theoretical argument, but exploitation requires
only the former. And his most powerful and compelling empirical analyses of “unequal and paired
categories” that generate clearly categorical forms of exclusion in contemporary liberal democratic
capitalist contexts concern internal organizational categories, not external categories.
In Marx’s account, from which Tilly claims to draw inspiration, exploitation requires only what Tilly
would call internal categories: owners of the means of production, on the one hand, and workers who
have been separated from the means of production, on the other. It does not require the matching of
internal and external categories. And in Tilly’s own account, exploitation requires only that some—
those who control valuable yet labor-demanding resources—enlist and coordinate the labor of others,
while reaping for themselves (at least part of) the value added by that labor. These others need not
differ by race, gender, citizenship, or the like; they may simply occupy subordinate organizationally
defined positions (casual in relation to career employees; adjuncts in relation to tenured professors; or
nurses in relation to physicians). These organizational distinctions may—and of course often do—
map onto external categories (such that nurses are overwhelmingly women, and physicians, as was the
case not so very long ago, overwhelmingly men); and Tilly calls attention to such cases. But the
phenomenon of exploitation—and, more generally, the dynamics of capitalism—does not pivot or
depend on this mapping. And while the matching of internal and external categories may stabilize
regimes of categorical inequality (1998: 76, 78, 81), it may also have the opposite effect: in a world in
which formal categorical inequality has been powerfully delegitimized, the tight matching of
internal and external categories may destabilize regimes of inequality, while the loosening of
3