Millennium Development Goals: Measuring Progress towards their Achievement

Abstract

The Millennium Development Goals are time‐bound quantified targets for improving the human condition from different perspectives. Within each Goal several targets have been set, and to each target there corresponds one or more indicators. For each indicator we axiomatically characterize an index of perceived progress towards reaching the Goals such that it can be used for monitoring progress. We also present a composite index of progress, which allows the calculation of percentage contributions of progress made in different dimensions. This, in turn, enables us to identify the dimensions for which more progress is required, which is important from a policy perspective. We also provide an empirical illustration of the proposed indices using cross‐country data for different indicators.

Keywords:

• Millennium Development Goals
• Targets
• Indicators
• Axioms
• Indices
• Characterization
• Illustration

Acknowledgements

The authors wish to thank two anonymous reviewers for their comments on an earlier version of this paper.

Notes

1. Several contributions have stressed the need for viewing human well‐being as a multidimensional phenomenon. See, for example, Kolm (1977), Streeten (1981), Atkinson and Bourguignon (1982), Sen (1985, 1987, 1997), UNDP (1990–2006), Tsui (1995, 1999, 2002), Ravallion (1996), Chakravarty et al. (1998), Bourguignon and Chakravarty (1999, 2003), Atkinson (2003), Chakravarty (2003, 2006), Chakravarty and Majumder (2005), Weymark (2006), and Chakravarty and Silber (2007).

2. This procedure, which Dutta et al. (2003) referred to as procedure II, was adopted in Chakravarty (2003) and Chakravarty and Majumder (2005) for developing generalizations of the Human Development and Human Poverty Indices.

3. The term ‘global’ relates to the norms, as indicated in column two of the table in appendix A. The targets are, however, country specific.

4. For specification of these indicators and for the targets and Goals corresponding to all the indicators used, see appendix A.

5. An n x n permutation matrix is a square matrix of order n where each row and each column has exactly one entry having the value ‘1’ and the rest of the entries have the value ‘0’.

6. Since the numbers of indicators belonging to the different Goals are not the same, this simple averaging assigns different weights (determined by data) to indicators across MDGs in the aggregation.