Determining criterion weights in multi-criteria decision-making using Shannon entropy faces significant limitations, stemming from its sensitivity to outliers and ambiguities in probability estimation, which can compromise the validity of the results. This paper introduces a new model based on cumulative residual entropy, offering a more robust and transparent weighting method that is resilient to skewed data. Its application to performance-based budgeting in Iranian universities demonstrates that, when combined with subjective approaches like the Delphi method, this technique yields more accurate estimates of actual budgetary requirements. Furthermore, by mitigating bias caused by outliers, the model prevents the excessive concentration of resources in large universities; instead, through a more equitable redistribution of weights, it facilitates budget allocation that favors developing universities.