The Benchmarks of Fairness are a generic matrix for assessing the fairness of health sector reform in developing countries. “Fairness” is taken to include equity (in risk factors, access to services, and financing), accountability, and efficiency. There are nine main benchmarks, each specifying a key component or goal of fair design of the health sector; each benchmark in turn contains various criteria and sub-criteria specifying key elements of the main goal or key means of achieving them. An evidence base is constructed by selecting locally useful indicators and scoring rules so that the degree to which reform improves (or worsens) the situation relative to each criterion can be made operationally specific. The approach can be adapted for use in evaluating comprehensive reforms at the national or sub-national level, or it can be used to focus on some key issues that are affected by reform, such as the impact or decentralizing reforms on the delivery of public health services or the impact of reforms on reproductive health.
The generic form of the matrix was developed in 1999-2000 through workshops involving collaborating teams in Pakistan, Thailand, Colombia and Mexico (see Daniels et al 2000; WHO Bulletin 78:6:740-50). A striking result was agreement on a matrix for assessing fairness despite the cultural differences in the collaborating countries. Starting in 2002, we began adaptation of the generic matrix for use in specific countries at the national and sub-national level. Experiments were undertaken in Yunnan (China), Cameroon, Mexico, Guatelmala and Ecuador, and Zambia. We report on some work in Daniels et al 2005; WHO Bulletin 83:534-40.
There is considerable interest in this approach because it allows an evidence-based evaluation of reforms underway; it can be used to assess reform proposals not yet implemented as well. It involves no cross-country indexing - and so is attractive to teams of policy makers, academics, civic advocates who want to improve things relative to where things are in their own country. The approach emphasizes practically useful information, not perfect measurement, and thus is adaptable to local conditions.