The Blackwell Companion to Sociology (80 page)

legitimacy and access to resources'' (Ibarra, 1993a, p. 27).

A fundamental hypothesis of network analysis is that social actors that occupy

more central network positions thereby control more resources and exercise

greater political influence. Although intraorganizational centrality empowers

employees, the reverse causal process may also operate: people seek connections

to the most powerful players, in hopes of enhancing their own power through

these connections. Occupants in central positions can derive several advantages

over employees in peripheral locations: acquire more timely and useful informa-

tion through their communication networks; mediate and broker deals between

interested parties who lack direct ties to one another; better monitor and control resource exchanges and mobilize support for workplace initiatives; and, through

ties to external organizations, participate in their organization's strategic activities. In sum, ``network centrality increases an actor's knowledge of a system's power distribution, or the accuracy of his or her assessment of the political

landscape. . . . Those who understand how a system really works can get things

done or exercise power within that system'' (Ibarra, 1993b, p. 494).

Conclusions

Theorizing and researching about organizational networks have largely pro-

ceeded at conceptually disjoint levels of analysis by investigators from differing disciplinary perspectives applying various methodologies and measurements to

explain divergent substantive interests. A major challenge for the new millen-

nium is to bring a closer integration among these disparate elements to achieve

more comprehensive explanations of how network structures and processes

form simultaneous constraints and opportunities for organizational populations,

fields, firms, subunits, and employees. Examining the connections among

Networks and Organizations

341

cross-level network phenomena is crucial to unraveling their complex dynamics.

The consequences of structural changes at one level ramify across other dimen-

sions. Thus, the outcomes of strategic alliances, lobbying coalitions, and other interorganizational relations among firms transform the daily work routines and

career opportunities available to top executives, middle managers, and produc-

tion workers of the collaborating enterprises (Kanter and Myers, 1991). Simi-

larly, aggregated webs of micro-level interactions among corporate departments

and their employees change the configurations of larger organizational networks

within which they are embedded. For example, an inability to develop trustful

relations among a firm's managers and workers could obstruct successful imple-

mentation of interorganizational agreements. As employees occupying key

boundary-spanning roles try to deal with uncertain company environments,

they socially construct mutual trust relations with peers in other organizations, which tend to obscure the distinctions between persons and their roles as

corporate agents. This confounding of interpersonal and interorganizational

trust poses potential problems about social capital ownership and control.

Tracking the co-evolution of network relations and their constituent actors

across multiple organizational levels of analysis requires patient longitudinal

data collection, analysis, and interpretation.

Network analysts have largely neglected to investigate how organizational

networks affect employee political activities, such as their participation in work decisions, interpersonal disputes, conflicts with management, and social movements to change company policies. Most previous studies examined personal

networks within only a single organization. We know relatively little about how

alternative network structures empower entire workforces, in the sense of raising collective organizational capacities to get things done. How do different intraorganizational network structures affect corporate power to mobilize human

and technical resources, perform productive tasks more efficiently, and achieve

profitable outcomes? Answering these questions requires comparative research

designs to measure multiple networks within many organizations.

On the applied side, an important implication of network analysis is that firms

and employees must recognize the ubiquity of inter- and intraorganizational

relations throughout corporate life. Identifying and building strong relationships can be beneficial strategies for both individual careers and the collective

performances of work teams, departments, firms, and strategic alliances. As

theoretical understanding and empirical findings about networking dynamics

accumulate and diffuse, management schools and organizations should encou-

rage explicit exposure to applied network management ideas as integral features

of their formal curricula and training programs (Baker, 1994). At the same time, attention must be paid to the dark side of network practices ± the potential for manipulation and abuse of power ± that threaten to undermine the norms of

reciprocity and trust crucial to sustaining fragile social ties in every organization.

Part VII

Individuals and Their

Well-Being

24

Social Inequality, Stress, and Health

Joseph E. Schwartz

Research into the determinants of disease morbidity and mortality during most

of the twentiethth century was dominated by a biological model of disease, as

was the practice of medicine itself. Illnesses and other physical disorders were thought to be caused by harmful agents (bacteria or viruses causing infectious

diseases, toxic chemicals, tobacco smoke, cholesterol, etc.) and/or a breakdown

in one or more of the body's organ systems. Consistent with this biomedical

model, primary prevention of disease has focused on reducing the population's

exposure to toxic substances and procedures to control the spread of infections ±

for example, ensuring a clean water supply, quarantining infected individuals,

and vaccinating individuals to make them immune to specific diseases.

It was not until 1977 that Engel published his seminal article setting forth the now widely accepted biopsychosocial model of disease. This article urged physicians and researchers to acknowledge and investigate the role of social, psy-

chological, and behavioral factors in the prevention, etiology, and treatment of both physical and mental illnesses. Despite its wide acceptance, at least in

principle, the vast majority of the medical literature ignores the role of psychological and sociological factors. A review of medical school curricula or the

contents of leading journals (for example, Journal of the American Medical

Association or Lancet) suggests that the medical establishment remains skeptical about the relevance of psychosocial factors in the etiology of disease.

The above not withstanding, the subfield of ``psychosomatic medicine,'' his-

torically dominated by psychiatrists, has a long tradition of emphasizing mind±

body connections. Increasingly, a broader array of researchers have been invest-

igating the impact of personality, behavior, and a variety of social factors on

morbidity and mortality. While the majority are psychologists, there are also

physicians, social epidemiologists, sociologists, anthropologists, and others.

Many of these researchers identify themselves with one or more of the

346

Joseph E. Schwartz

interdisciplinary fields of psychosomatic medicine, ``behavioral medicine,'' and

``health psychology.''

In this chapter I review selected empirical findings and issues from behavioral

medicine that are likely to interest sociologists. The first section is primarily descriptive, reviewing the fact that the risk of many diseases, and death itself, is socially patterned. Much of the emphasis is on mortality and cardiovascular

disease, but the general point applies to other causes of death and many non-

fatal diseases. Since, in my opinion, the subfield of social stratification lies at the core of sociology, this section emphasizes the link between stratification and

health. While many mechanisms surely contribute to this relationship, I am

particularly interested in the effect that stress may have on health and the

possibility that differential exposure to stress in the social environment partially accounts for social class differences in health. The latter part of the chapter

presents select findings from studies of animals and humans pertaining to the

impact of social stress on health.

The Social Patterning of Disease and Mortality

Demographic Factors

One of the goals of descriptive epidemiology is to document the extent to which

mortality and disease prevalence varies by age, sex, and race/ethnicity. Many

diseases are more frequent at older ages (heart disease, cancer, hip fracture), but others, such as AIDS and violence-related injuries, are more common in younger

individuals. Similarly, some diseases are gender-specific for obvious biological reasons (uterine/cervical cancer, prostate cancer), others are more common in

one gender (stroke, breast cancer, and fibromyalgia in women; AIDS and vio-

lence-related injuries in men), and yet others are relatively gender-neutral (heart disease, diabetes, pneumonia, and flu). With respect to race/ethnicity, many

genetic disorders are much more common in specific subgroups (for example,

sickle cell anemia in blacks, Tay Sach's disease in Jews of East European des-

cent). The same is true for several other disorders whose genetic basis is not

clear: renal disease is much more prevalent in black Americans than in white

Americans; stroke is more common in Asian Americans than white Americans;

and several cancers are less common in Hispanic Americans than white Amer-

icans. Overall, mortality rates increase with age (after age five), are higher in men than women, and are higher in black Americans than in white or Asian

Americans.

Of course it is one thing to document demographic differences in the rates of

diseases and mortality, and quite another to explain their existence. Take hypertension as an example. In the United States, increased age and being black are

two of the leading risk factors for hypertension. Prior to age 50, men are

substantially more likely to have hypertension than women, but following

menopause, the risk of hypertension increases substantially in women and

probably surpasses that of men (Subcommittee on Definition and Prevalence of

Social Inequality, Stress, and Health

347

the 1984 Joint National Committee on Detection, Evaluation, and Treatment of

High Blood Pressure, 1985). The causes of this social patterning are not well

understood, though it is widely believed that genetics, behavior, and social/

cultural factors play a role.

We tend to consider the positive association between age and hypertension in

the United States as reflecting thè`natural aging process,'' but there are data

suggesting that there is nothing innate about either the average level of blood

pressure (BP) or its increase with age. Waldron et al. (1982) analyzed BP data for 84 different adult samples from throughout the world, nearly all available data

sets satisfying a predefined set of criteria. Independent ratings of several cultural characteristics, based largely on anthropological data and reports, were also

obtained for each sample. The authors found that the average age-matched BP

of men and women were substantially lower in hunter-gatherer, herding, and

traditional agricultural societies than in morè`modern'' agricultural and indus-

trial societies in which most of the production was distributed through a market economy. These cross-cultural differences were greater in older adults (aged 50±

60) than younger adults (aged 20±30), reflecting an age gradient of blood

pressure that was flat or nearly flat in the more traditional societies but tended to increase by 0.5±1.0 mmHg with each year of age in market economies. The

result in market economies was a difference of 15±30 mmHg between those who

were 55 years old and those who were 25 years old.

In another intriguing study, Timio et al. (1988) followed 144 nuns from an

order in Umbria, Italy, and 138 women from surrounding communities for 20

years. The nuns were completely secluded from urban life, prayed and worked in

nearly absolute silence, and were not exposed to competition, insecurity about

their future, or the economic, familial, and other stressors of modern Western

society. The demographic and physiological characteristics, including BP, of the two groups were similar at recruitment into the study. While the nuns' BPs

increased only about 1 mmHg over the entire 20 years, systolic BP in the control sample increased by an average of about 1 mmHg per year and diastolic BP

increased about 1 mmHg every three years. The difference in age gradient

between the two groups could not be explained by differences in alcohol,

tobacco, coffee, or tea consumption, salt intake, cholesterol, or body weight

change. It also could not be explained by differences in childbirth, since the

age gradients were quite similar in the control women who had and had not

had children. While unable to specify a mechanism, the authors suggested

that the failure of nuns' BPs to increase with age must be due to the near

absence of ``conflict, aggression, and competition for power and money'' in