伴随着信息技术的勃兴，数字政府被广泛应用于我国各级政府的治理实践中。在顶层设计中，中央政府强调推动数字政府在多元政策子领域的协同推进。然而，各地方政府在数字政府议程中的注意力分配却表现出较大的府际差异。尤其在2013年以后，有些省份甚至偏离中央政策意图，不断地在强调数字政府在支撑经济发展上的作用，并逐渐弱化对其在政府建设支撑上的关注。基于此，本文的研究问题是：地方政府在数字政府议程中对经济发展与政府建设子议题差异化的注意力分配背后的驱动因素与生成逻辑是什么？在经济增速相对政府建设的重要性逐渐降低的考核环境下，为什么有些省份的数字政府议程反而越来越明显地表现出与经济发展子议题不断捆绑、与政府建设子议题不断疏离的变化趋势？既有文献从议题偏好、政治周期、委托代理以及信号压力等四个经典理论视角讨论政府注意力的驱动因素。然而它们在本文的实证情境中存在解释局限性。基于此，作者提出“议题捆绑”的理论视角来讨论政府的政策议程结构均衡性对其微观数字政府议程的影响。“议题捆绑”指的是某项政策的履职部门将该政策与某些相对优势议题联系在一起，从而获得组织权威和政策资源，并与某些相对劣势议题划清界限，从而回避政策敏感地带的策略性行为。其动力来源在于履职部门希望借此化解因自身治理资源不足而带来的治理困境。在该视角下，政府政策议程的结构均衡性的提升会增加数字政府履职部门的“协调难”困境，迫使其将数字政府建设与经济发展议题捆绑在一起，并与政府建设议题相剥离。基于整合式的混合研究设计，通过对各省份53841篇党委机关报的数字政府报道以及23439条政府常务会议讨论事项的整理分析，作者构建了一个覆盖31个省份2007-2018年共48个季度的时间序列-横截面数据库，并使用交叉多层模型进行实证检验。另外，通过对A省“互联网+政务服务”改革的案例分析，作者进一步挖掘了其背后的作用机制。研究发现：第一，政策议程结构均衡性的上升会对数字政府议程中相对优势子议题（如，经济发展子议题）的关注度产生正向影响效应，并对相对劣势子议题（如，政府建设子议题）的关注度产生负向影响效应。第二，与经典理论视角相比，“议题捆绑”的视角为我们同时解释经济发展与政府建设子议题在数字政府议程中的关注度变化提供了一致的分析框架。本研究扩展了注意力分配理论在中国制度情境下的适用性，丰富了政府注意力的解释维度，有利于深化对我国数字政府议程设置中策略性非正式行为的理解。

With the populairty of information communication technology, digital government emerges in the Chinese modern governing systems at different levels. In the national policy guaidence of digital government, the central state emphasized the imlementation of digital government projects in multiple policy areas. However, local governments still exibits differences in the patterns of attention allocation in digital government agendas. Particularly, there is a large expansion of attention to the sub-issue of Economic Development and a large cut of attention to the sub-issue of Government Construction after 2013, which violating the policy signals of central government. With regard to this, I propose the following research questions: What affects the pattern of attention allocation across different sub-issues in local governments’ digital government agendas? Considering that the relative importance of economic growth rates decreases in cardre evaluation systems, how do we explain the recent increases in the proportion of attention to the sub-issue of Economic Development and the decreases in the proportion of attention to the sub-issue of Government Construction in digital government agendas after 2013?The author develops a theoretical framework with the help of “Issue-Bundling” and explores the impact of the degree of agenda equilibrium of provincial governments on the salience of each sub-issue in the digital government agendas. The concept of Issue-Bundling refers to that policymakers usually strategically associate the policy issues with some favorable issues in order to get their authorities and resources, and dissociate the policy issues from certain unfavorable issues in order to stay away from controversies. The motivation of Issue-Bundling comes from the desire of policymakers to get rid of the governance dilemma. Follwing this perspective, the increase in local governments’ agenda equilibrium may increase the difficulty of policy coordination in polical domain of digital government, thus fostering relevant policymakers to allocate more attention to the sub-issue of Economic Development, and at the same time, decrease the attention to the sub-issue of Government Construction. This study employs “Integrative Multi-Method Research” to conduct empirical test. The author develops a unique dataset which covers 31 provinces 48 quarters from 2007 to 2018 (1488 observations) by analyzing 53841 digital government news reports from party newspapers and 23439 governments executive meeting notes. The cross-classified multilevel regression was employed to estimate the models. Besides, through the case study of the “Internet + Government Service” reform in Province A, the author further identifies two possible mechanisms by which the agenda equilibrium may affect the attention allocation. The empirical results reveal that: First, the increases in agenda equilibrium probably have a positive impact on the proportition of attention to the favorable sub-issues (eg, the sub-issue of Economic Development), and have a negative impact on the proportition of attention to the unfavorable sub-issues (eg, the sub issue of Government Construction). Second, in comparasion with the four classic perspectives derived from the existing literature, the perspective of “Issue-Bundling” could explain the variantions in attention paid to the both two sub-issues.This study extends the generalization of existing knowledge regarding the theory of attention allocation to the institutional context of China, and develops a new framework to exmine governments’ attention allocation. It may also deepen our understandings of the strategic informal behaviors in the practice of China’s digital government development.