In Chapter 2, Tendler explores replicable lessons in the Cears Health Agent Program. The program began in 1987. Within five years of its inception, the program produced tremendous improvements in child health indicators. Tendler proposes several mechanisms to explain how a large number of low-paid, unskilled, public health workers are able to perform exceptionally well.  Tendler emphasizes the role of the central in the decentralized. She argues that the success of the program is not a triumph of local government, but rather that of the state government, which involved itself in the health sector by:

(1)  Quelling fears of clientelism and potential opposition.

The state retained control over certain aspects of the program.  During early years, it kept funds for health agent salaries from the municipal government and the Department of Health.  The state also retained responsibility for hiring in order to prevent hiring through informal patronage at the local level. This quelled potential opposition from medical professionals by turning nurses from potential resisters into ardent advocates. The Department of Health left the nurse-supervisors with substantial control over the way they ran the program. They then became champions of the program.

(2)  Creating a sense of mission, or calling, to the program and to its public servants.

Through massive publicity campaigns, state officials declared that all those involved in the program would be taking part in the noble mission of bringing the community into the 20th century by reducing infant mortality and disease (33). Health agents would not only contribute to this mission, but their involvement would come with added prestige, free training/education, and a decent wage. To be chosen for the job of health agent was like being awarded an important prize in public (29).

(3) Strengthening pressures for accountability.

The state government held massive publicity campaigns to instruct citizens on their rights to quality public services. They told citizens to pressure their mayors with the threat of voting non-participating or corrupt mayors out of office in subsequent elections. During the hiring process, they explained to those who were not chosen for the job that they were still motivated community leaders with the ability to monitor those who were chosen. The state created an informed group of public monitors through the hiring process and through publicity campaigns.

(4) Providing agents and nurse-supervisors, with more autonomy, discretion and control over the program and thus encouraging in them existing predilections to do good.

Workers engaged in a broader set of tasks than that formally proscribed. Some complemented curative with preventative health activities. Others engaged in community-wide activities to reduce public health hazards. Many helped overburdened mothers with housework and child care. Workers were allowed to customize their services to their clients/community.

Ultimately, the Health Agent Program decreased infant mortality and the prevalence of disease in rural communities by enabling an environment conducive to worker dedication. Workers, with a certain level of autonomy, built relationships of mutual respect and trust with their clients which created internal pressures for accountability.  Concomitantly, public messages to teach citizens that they had a right to demand better health care created external pressures. What emerged was an informed public with close relationships to its public servants based on ties of trust and respect.  These explanations, Tendler argues, are much more useful to the planner than any description/proscription for good performance based on a notion of good leadership.