At the end of the nineteenth century, American corporate law changed into its modern, permissive form. Emerging first in \NJ[,] the new laws allowed corporations to own each other, and using them capitalists solved their collective action problem of ruinous competition, reorganizing American industry in a great merger wave. The most famous explanation for why the law changed, the efficiency argument, has been refuted by economic sociologists and political scientists who instead argue that the law changed because of the combination of powerful actors pursuing their interests and contingent conditions such as state fiscal crises. From this critical juncture, the law developed path-dependently. But these conditions did not occur in \NJ[,] and the actual development of the law there, with several anti-incumbent changes in a relatively short period, do not fit with a path-dependent model. Neither does the simultaneous adoption of both permissive corporate laws modeled on \NJ[']s, and their reaction, restrictive antitrust laws, across the United States. To address these problems, this dissertation develops a theoretical framework in which institutional change is determined by how the institution serves the interests of incumbents in adjoining fields. In its first part, this dissertation applies this framework to the development of corporate law in \NJ over time, comparing 28 failed and successful attempts to change the law, 1830-1913. In its second part, it tests this framework on the adoption of permissive and restrictive corporate laws across the United States, 1889-1915. Using both qualitative and quantitative methods on new datasets of bills, votes, laws, politicians, corporations, and taxes, this dissertation finds that how existing law served political and economic incumbents' interests, by limiting competition and providing tax revenues, explain its change and persistence. The power and interest relationships among classes are reflected back into the state through taxation, mediated by state capacity. Politicians sit like both spider and fly in a web of dependencies among actors and policies. When the web is tight, institutions persist, when it loosens, change is possible.