Abstract
Public policy performance is increasingly being evaluated using multidimensional frameworks such as the Human Development Index, the Multidimensional Poverty Index, and the Sustainable Development Goals. Despite their appeal, these frameworks rely on fixed, exogenously defined aspirations and a scalar-based composite index approach, which masks persistent underperformance across dimensions. Instead, this paper argues for a vector evaluation approach consistent with Thaler’s mental accounting principle: (a) Individuals do not collapse all dimensions into a single aggregate measure; b) They track performance separately; c) They react strongly to shortfalls in specific accounts. Thus, this paper considers each dimension as a separate psychological account with its own reference point, where shortfalls in some dimensions are not fully compensated by gains in others, justifying non-compensatory, vector-based policy evaluation. It extends the reference dependence of Prospect Theory beyond individual choice to institutional processes of aspiration formation, and from single-dimensional to multidimensional policy evaluation, where reference points evolve endogenously in response to peer performance and institutional priorities. It formally embeds a “Leave No Dimension Behind” principle in the proposed framework. It delivers a transparent, policy-relevant diagnostic tool that identifies which dimensions are systematically left behind in an adaptive shortfall analysis, as performance shortfalls trigger problemistic search, i.e., selective attention, policy revision, and resource reallocation. Finally, it illustrates the methodology with an application to the evaluation of Sustainable Development Goal (SDG) 1 across Indian states and Union territories, based on empirical data from the SDG India Index 2023-24 report of NITI Aayog, Government of India.