This paper examines how artificial intelligence reshapes subnational public finance through familiar channels posed by prior technological change. AI shifts income from labor toward capital and tax bases toward consumption and market-based allocation, raising issues like the recent discussion of the sales tax treatment of digital services. With respect to the use of AI, we argue that AI relaxes long-standing informational and administrative constraints in state and local taxation, enforcement, budgeting, and service delivery, while simultaneously strengthening scale economies. The ability to reduce costs at the depends critically on labor-intensive services such as K-12 education. At the same time, the use of AI may advantage larger jurisdictions because of the larger amounts of data that they have, but also potentially raising equity and transparency concerns and increasing the value of interstate cooperation to harness the scale advantages of having more data. AI amplifies— the classic trade-offs emphasized in fiscal federalism.

We investigate whether providing politicians with low-cost access to peer information influences fiscal behavior, exploiting an Italian program that allowed mayors to access detailed municipal expenditure data via a restricted website. By tracking digital activity, we construct a directed network of peer monitoring. We find that participating mayors are younger, more educated, and govern larger municipalities. Exploiting network intransitivity to address the reflection problem, we demonstrate that digital transparency fundamentally alters fiscal competition. Within the network, strategic interaction in property tax setting is conditional on electoral incentives, driven primarily by mayors eligible for re-election. Conversely, municipalities outside the network exhibit geographic correlation unrelated to term limits. These findings suggest that digital platforms facilitate sophisticated, reputation-based yardstick competition, demonstrating that transparency tools influence politician behavior even prior to public disclosure.

This paper models theoretically and tests empirically the hypothesis that the decision about the location of a public bad within a multi-tiered structure of government can be driven by strategic electoral considerations exploiting the heterogeneous migration responses to the location of the public bad by voters of different ideologies - a sort of mobility-based gerrymandering. As long as the average utility loss from living close to the public bad is larger for progressives than it is for conservatives, conservative and progressive central governments will pursue opposite strategies. The former locate the public bad in an electorally tight region to induce exit of progressive voters and gain the region to the conservative party, while the latter attempt to spread progressive voters out of safe and towards electorally tight regions. An application to waste treatment plant locations across Italian municipalities returns evidence in support of the model’s main hypotheses.