Despite the rising popularity of indicators of women's empowerment in global development programmes, little work has been done on the validity of existing measures of such a complex concept. We present a mixed methods validation of the use of the Relative Autonomy Index for measuring Amartya Sen's notion of agency freedom in rural Nepal. Analysis of think-aloud interviews ( = 7) indicated adequate respondent understanding of questionnaire items, but multiple problems of interpretation including difficulties with the four-point Likert scale, questionnaire item ambiguity and difficulties with translation. Exploratory Factor Analysis of a calibration sample ( = 511) suggested two positively correlated factors ( = 0.64) loading on internally and externally motivated behaviour. Both factors increased with decreasing education and decision-making power on large expenditures and food preparation. Confirmatory Factor Analysis on a validation sample ( = 509) revealed good fit (Root Mean Square Error of Approximation 0.05-0.08, Comparative Fit Index 0.91-0.99). In conclusion, we caution against uncritical use of agency-based quantification of women's empowerment. While qualitative and quantitative analysis revealed overall satisfactory construct and content validity, the positive correlation between external and internal motivations suggests the existence of adaptive preferences. High scores on internally motivated behaviour may reflect internalized oppression rather than agency freedom.

This paper discusses the relevance of recent research on the economics of human development to the work of the Human Development and Capability Association. The recent economics of human development brings insights about the dynamics of skill accumulation to an otherwise static literature on capabilities. Skills embodied in agents empower people. Enhanced skills enhance opportunities and hence promote capabilities. We address measurement problems common to both the economics of human development and the capability approach. The economics of human development analyzes the dynamics of preference formation, but is silent about which preferences should be used to evaluate alternative policies. This is both a strength and a limitation of the approach.

We report the results of a series of validity and reliability tests performed during the development of the Women's Capabilities Index (WCI) in Malawi. The WCI is a multidimensional measure based on Sen's capability framework for assessing women's quality of life. Construct validity was assessed by investigating the expected relationships of the dimensions with key socioeconomic characteristics. The majority of hypothesized associations were found to be statistically significant in the expected direction. This provides evidence that the index is measuring quality of life as intended in the conceptual model. Further evidence in support of the index's validity was given by the high degree of correlation between the WCI and another scale measuring comparable (but not identical) domains of quality of life. The results from the internal consistency and the test-retest repeatability also offered encouraging evidence on the reliability of the instrument. This is the first study to rigorously and comprehensively test for validity and reliability a capabilities index for a low-income setting. The results of the validity and reliability tests provide supportive evidence that a locally developed measure of capabilities can be used as a robust tool for the assessment of women's quality of life.