The precise targeting of low-income assistance recipients has long been plagued by information asymmetry between the state and farmers, leading to targeting errors and institutional distortion. Based on multi-site fieldwork experiences, this study analyzes how the integration of ″virtual income″ intelligent calculation and multi-departmental data comparison mechanisms through big data technology transforms the ambiguous ″actual income″ in rural society into clear and quantifiable ″virtual income″, significantly enhancing the states ability to verify eligibility for low-income assistance and, to a certain extent, resolving the predicament of targeting errors. The research finds that the application of big data technology not only reverses the information advantage structure within the bureaucratic organization (superiors over subordinates) and between the state and society (state over farmers), but also drives the governance mechanism to shift from the ″contractual delegation″ model relying on grassroots agents to a ″co-governance″ model led by the state. On this basis, the study further proposes a ″information dimension″ theory of state governance, providing a new perspective for understanding the governance transformation in the process of Chinas modernization.