coherence-based-assessment-of-multidimensional-capability-measures-recognizing-the-importance-of-incommensurability

Chamberlain, Travis Dean (2017). 'Coherence-Based Assessment of Multidimensional Capability Measures: Recognizing the Importance of Incommensurability' Paper presented at the annual conference of the HDCA, Cape Town 2017.

Abstract

Recent index-based measures of capability-based well-being face a significant problem: weighted aggregation of distinct features of capabilities disturbs the coherence of the measure and risks tracking a narrowed concept.  I use Norman Bradburn and Nancy Cartwright’s theory of measurement (ToM) to highlight three key ingredients of a multidimensional capabilities measure that must cohere.  I then argue that index representations of multidimensional capabilities measures are less coherent than table or matrix representations.  Key to the argument is the incommensurability of features characterizing capabilities: distinct features cannot be compared in value and cannot be measured with the same unit.

My argument proceeds as follows: I first show that capabilities-based well-being is a conglomerate or, following Otto Neurath, a ‘ballung’ concept: as an ordinary language term, well-being is constituted by shifting sets of context-relevant features.  I then consider in accord with ToM two possible responses to measuring ballung concepts: (a) provide more descriptively accurate but merely partial context-dependent representation, or (b) collapse partial representation into a more complete representation that substitutes a narrowed concept for the original concept.  Both Amartya Sen and Martha Nussbaum’s capabilities approaches emphasize a context-dependent, partial representation response, yet the index operationalization of Sen’s approach quantifies and compares features characterized as incommensurable.  For this reason, the index operationalization of Sen's approach measures a narrowed concept that no longer coheres with the original characterization.  I also argue that value-based selection of functionings within Sen’s approach implicitly ranks features characterized as incommensurable.  For this reason I claim that Sen's representation of capability-based well-being, and not merely the index-based operationalization, tracks a narrowed concept and is a type-(b) response to measurement of well-being.  Nussbaum’s objective list approach, grounded by philosophical tradition that supports the excellence of the particular over the universally commensurable good, remains descriptively accurate and context-sensitive toward the ballung nature of capabilities-based well-being.  I ultimately claim that Nussbaum's response follows a type-(a) response to measurement of well-being.

After arguing that Sen and Nussbaum’s approaches exhibit substantive differences in coherence, I analyze two recent measurement innovations: Jaya Krishakumar’s improved (2007) HDI index and Sabina Alkire and James Foster’s headcount measure of capability deprivation.  I consider Krishnakumar’s index as a refinement of Sen’s approach, while I consider Alkire and Foster’s headcount measure a refinement of an objective list approach similar to Nussbaum’s.  Further analysis of these innovations demonstrates how a table-based representation of capability-based well-being is more coherent than an index-based representation.

scroll to top