In local search for the job shop scheduling problem, the design and evaluation of neighborhood solutions have a direct impact on algorithmic efficiency and scalability. To address the high cost and redundancy of traditional evaluations, this paper proposes a batch approximate evaluation method based on slack-time recursion, which uses the critical block as the basic unit to build a structured batch-evaluation framework and thereby reducing the evaluation complexity for all moves within a block. In addition, two new sufficient conditions for non-improving moves in the N7 neighborhood are derived to strengthen the neighborhood constraints. Computational results show that the proposed approach maintains estimation accuracy, significantly improves search efficiency, and achieves competitive scheduling performance within a tabu search framework.