Democracy & Governance
Deep political polarisation and declining institutional trust make long-term policy commitments increasingly difficult to sustain.
Where does the USA stand, and where does it break?
Political polarisation in the United States is among the deepest of any industrialised country (Pew Research). Trust in the federal government has stood at a historic low for years, at around one fifth of the population (Pew Research). Federalism shifts key decisions to the states: taxation, social policy, and rules on AI and data. As a result, states are diverging in fundamental ways. Texas, California, and New York pursue opposing approaches on regulation, social protection, and technology policy. Where one lives increasingly shapes the conditions of one's life. At the federal level, partisan gridlock makes long-term, cross-party reform difficult. Major legislative initiatives frequently fail due to narrow majorities and changing administrations. Institutional trust varies across Congress, the judiciary, and the executive, but is declining overall (Edelman Trust Barometer, Pew). At the same time, the United States is rich in democratic experimentation: ballot initiatives in many states, independent redistricting commissions, and new electoral methods such as ranked-choice voting in Alaska and Maine are all testing ways out of polarisation.
The AI transformation demands long-term, coherent decisions on taxation, participation, social protection, and the regulation of powerful technology companies. A deeply polarised, federally fragmented system finds precisely this kind of decision-making difficult. If AI continues to concentrate profits and power among a small number of corporations and inequality rises, political pressure will build. Whether that pressure flows constructively into reform or destructively into further division is the defining unknown of the American path. Federalism cuts both ways. It turns states into laboratories in which capital participation, social protection, and AI regulation can be tested. At the same time, it drives states apart and makes a coherent national response harder to achieve. If trust in institutions continues to fall while automation unsettles the middle class, the capacity of democracy to act risks eroding precisely when it is needed most. The governance question is therefore perhaps the most critical turning point in the US scenario.
Paths for the USA
The United States deliberately exploits its diversity of states: successful models — such as Baby Bonds, automatic enrolment in retirement savings, or short-time working schemes — are piloted in individual states and, if proven effective, adopted nationally. This fits the federal order and bypasses gridlock at the federal level, because progress begins at the local level. This is how the United States has historically scaled many reforms. The weakness: without a mechanism for transferring lessons, successes remain local, divergence grows, and citizens in states unwilling to reform do not benefit.
Trade-off: decentralised innovation versus national coherence
Feasibility: Effective only if a mechanism exists to scale successes nationally
The federal government establishes cross-party commissions or independent institutions. These remove long-term questions about the AI transformation — pension security, AI governance, fiscal trajectory — from day-to-day political battles and develop viable reform proposals. This addresses gridlock by enabling compromise beyond electoral cycles, in the manner of earlier reform commissions. The weakness: in a highly polarised environment, even cross-party bodies are quickly captured by partisan interests, and their recommendations remain non-binding for as long as Congress declines to act on them.
Trade-off: long-term orientation versus democratic accountability
Feasibility: Only effective with a credible implementation mechanism
The United States is expanding reforms such as ranked-choice voting, independent redistricting commissions, and open primaries. The aim is to reduce the incentives for polarisation and strengthen centrist majorities. This addresses a structural cause of gridlock by rewarding the capacity for compromise. Alaska and Maine provide functioning examples. The weakness: such reforms are themselves politically contested, their effects are disputed, and they take hold slowly — potentially too slowly for the urgent AI transformation.
Trade-off: structural detoxification versus pace of implementation
Feasibility: A long-term lever, not a short-term instrument
The United States is establishing capable institutions for the regulation of AI, with responsibility for safety, competition, liability, and the concentration of power among large models. This is central because, without competent state counterweights, the AI transformation will be left entirely to corporations. Approaches such as an AI safety institute and executive orders exist, but oscillate between regulation and deregulation depending on the administration in power. The weakness: without stable, cross-party institutional anchoring, such bodies remain vulnerable to every election cycle and difficult to connect with international frameworks.
Trade-off: regulatory capacity versus political stability
Feasibility: Only effective if anchored on a cross-party and permanent basis
Trigger-based phases
Trigger: Individual states are increasingly making opposing decisions on AI, taxation, and social protection, while the federal government remains deadlocked.
Federalism becomes a laboratory: pioneer states are piloting capital-sharing schemes, social protection, and AI rules. At the same time, inequality between states is growing, and place of residence increasingly determines living conditions.
Trigger: Rising inequality driven by AI concentration meets declining institutional trust, intensifying political conflict.
The pressure may feed constructively into reform or destructively into deeper division. Cross-party commissions, electoral reforms, and AI governance reforms are competing pathways to preserving the capacity to act.
Trigger: The system either arrives at viable national policy choices or remains locked in deadlock and deepening federal divergence.
The United States will determine whether its democracy shapes the AI transformation or is driven by it. Governance capacity is the prerequisite for whether any of the other decisions can be made at all.
Real-world pilot programmes
Status: In force (new electoral procedure aimed at reducing polarisation), narrowly retained in Alaska by referendum in 2024
Alaska and Maine use ranked-choice voting, designed to favour moderate, compromise-oriented candidates. The procedures are regarded as a real-world laboratory for testing whether structural electoral reforms can reduce the incentives for polarisation. The track record is mixed and closely watched.
Status: In Kraft in mehreren Bundesstaaten, etwa Kalifornien, Michigan
Several states have transferred the drawing of electoral districts to independent commissions rather than legislatures, in order to limit partisan manipulation (gerrymandering). They demonstrate an institutional path for addressing the structural causes of polarisation.
Status: Ongoing — including the Colorado AI Act (entry into force deferred to 2026)
In the absence of comprehensive federal rules, individual states are enacting their own AI legislation and establishing working groups — Colorado among them, with one of the first comprehensive AI laws. These initiatives serve as federal laboratories for AI governance and as illustrations of growing regulatory divergence.
What we are watching
This metric measures the legitimacy base for state action. It matters because declining trust undermines the capacity to enact and sustain deep reforms. If it tilts downward, democracy becomes incapable of acting precisely when it is needed most. It is a precondition for every other policy lever.
This measures how far the states are drifting apart. The threshold matters because growing divergence creates legal uncertainty, makes access to the economy ahead dependent on where one lives, and prevents a coherent national response. If it tips, federalism ceases to be a laboratory and becomes an obstacle. Uncertainty rises for businesses and citizens alike.
This metric shows whether the federal government is finding long-term answers to the AI transformation. It matters because, without national coherence, questions of taxation, participation and governance will remain unresolved. If it tips, democracy will not shape the transformation — it will suffer it. For investors and citizens alike, it is the indicator of political predictability.