Developing maqasid index for Islamic CSR: the case of Ummah’s Endowment Fund in Indonesia

Purpose This study aims to develop the maqasid index (MI) for Islamic corporate social responsibility (CSR), namely, Dana Abadi Umat (DAU) (Ummah’s Endowment Fund) or MI-DAU in Indonesia. Design/methodology/approach Modeling and weighting are based on Delphi and analytic network process (ANP) methods, called Delphi–ANP additive weighting. The Delphi method was applied to design and validate the factors of the MI-DAU model, and the ANP method was applied to generate and validate weights for these factors. Finally, the MI-DAU is calculated, based on the planned budget and actual allocation of DAU returns, called the maslahah fund, using additive weighting. Findings Delphi and ANP show significant and robust results. The priority order and weights of maqasid Shariah are safeguard the faith (0.32), safeguard the intellect (0.219), safeguard the life (0.204), safeguard the wealth (0.171) and safeguard the lineage (0.104). Meanwhile, the priority order and weights of the main activities are education (0.190), Ummah’s economy (0.167), Hajj service (0.155), Da’wah (0.124), health care (0.118), social-religious (0.097), worship facilities (0.085) and disaster emergency response (0.065). Finally, the results of MI-DAU show a high index in 2019 and 2020 of 71.89 and 69.51, respectively, generated from allocation ratio of 90.63% and 85.98%, respectively. Research limitations/implications Maqasid Shariah used in this study follows Al-Ghazali, where it could also follow maqasid Shariah of Abu Zahrah or Al-Najjar. Moreover, the MI-DAU score uses additive calculations, where it can also use Pentagon calculation. Practical implications The improved framework and method used to design MI-DAU in this study could be applied to design more scientific MI for other Islamic financial institutions. Originality/value The novelty of this study is in the improved method used to design the MI model, including its factors, using Delphi, and to assign weights of all factors using ANP, where both provide validation for more robust MI model.