The inequality spiral is the mechanism that distinguishes Stiglitz's framework from the standard economic story of technology and distribution. Standard accounts treat inequality as a side effect of technological change that institutions can, in principle, correct. Stiglitz's demonstration is darker: the wealth generated by the technology funds the political activity that prevents the institutional corrections, producing a feedback loop in which the very success of the technology makes its governance more difficult. The AI transition accelerates the spiral because the technology concentrates wealth at unprecedented speed and, simultaneously, amplifies the political resources available to the concentrators. The broligarchs of the AI industry are not coincidentally advocating for smaller government; they are advocating for the structural conditions that preserve their capture of the surplus.
The spiral operates through four linked stages, each feeding the next. First, concentration: AI generates extraordinary productivity gains that, under current institutional arrangements, flow disproportionately to the owners of capital and the controllers of platforms. Second, political capture: the concentrated wealth funds lobbying, campaign contributions, think tanks, and media operations that shape the regulatory environment. Third, institutional design: the captured political process produces tax codes, labor laws, and regulatory frameworks that preserve and extend the concentration. Fourth, further concentration: the designed institutions produce the next cycle of disproportionate gains, and the spiral tightens.
Stiglitz has documented the spiral across multiple contexts. The financial deregulation of the 1980s and 1990s was funded in significant part by the financial industry's lobbying, and it produced the conditions that generated the 2008 crisis, which was resolved through bailouts that further concentrated the financial industry. The tax cuts of the same period were lobbied for by their beneficiaries, who used the resulting savings to fund further political activity. The pattern is not unique to finance. It is the generic operation of political economy under conditions where wealth translates into political power and political power translates into institutional design.
The AI version of the spiral has specific features. The technology concentrates wealth faster than previous innovations because its productivity gains are captured almost immediately through equity appreciation rather than distributed through wage increases. The political power of the technology industry is amplified by its control of information infrastructure — the platforms through which political communication flows — giving it influence over political processes that previous industries did not enjoy. And the ideology of the industry, with its emphasis on technological solutionism and skepticism of government, provides the rhetorical framework for opposing the institutional interventions that would break the spiral.
Stiglitz's observation about the self-contradiction is the key analytical move. The technology industry simultaneously (1) generates the productivity surplus that requires institutional intervention to distribute broadly, (2) uses that surplus to fund political opposition to institutional intervention, and (3) requires a competent government to manage the transition it is creating — a government whose competence the industry's political activity systematically undermines. The industry is, in Stiglitz's formulation, creating the conditions that make successful transition impossible while claiming to be the agent of successful transition.
Stiglitz developed the inequality spiral framework most explicitly in The Price of Inequality (2012), building on earlier work with Andrew Weiss on credit rationing and with George Akerlof on information economics. The framework draws on Albert Hirschman's earlier work on The Rhetoric of Reaction and synthesizes insights from political economy, institutional economics, and the comparative study of democratic capture by concentrated interests.
Wealth translates into political power. Not automatically, but reliably — through lobbying, campaign finance, media ownership, think tank funding, and regulatory capture.
Political power shapes institutional design. Tax codes, labor laws, regulatory frameworks, and trade agreements are written by or for those with political access, systematically favoring their interests.
Institutional design produces further concentration. The designed institutions structure markets in ways that advantage those who designed them, closing the loop.
The technology industry's structural contradiction. AI companies require competent government to manage the transition while systematically opposing the expansion of government capacity required.
Speed amplifies the problem. Previous inequality spirals operated over decades, giving democratic processes time to mobilize counter-power; the AI spiral operates over quarters.
Defenders of the technology industry argue that its political activity reflects legitimate policy preferences, not capture, and that the wealth concentration reflects genuine value creation rather than rent extraction. Stiglitz's response: the test is structural, not intentional. When the policy preferences of an industry systematically align with its short-term financial interests and systematically conflict with the broader welfare, the distinction between advocacy and capture becomes a distinction without a difference.