The longest federal government shutdown in American history appears finally to be approaching an end. After 40 days without normal government funding, a group of Senate Democrats has joined Republicans behind a compromise that would reopen most of the federal government temporarily, provide longer-term funding for several departments and guarantee back pay for furloughed federal employees. The central Democratic demand that prolonged the confrontation — an extension of enhanced Affordable Care Act health-insurance subsidies due to expire at the end of the year — has not been included directly in the agreement; Republicans have instead promised a separate Senate vote on the issue in December. The shutdown has meanwhile moved far beyond Washington’s political institutions. Hundreds of thousands of federal employees have worked without pay or been furloughed, airport staffing shortages have disrupted flights, and SNAP food assistance for roughly 42 million Americans has become entangled in an extraordinary legal and administrative battle after benefits lapsed at the beginning of November for the first time in the programme’s sixty-year history. The immediate political question will be which party “won”. The more important institutional question is why the United States repeatedly allows disagreements about future spending to disable services that both parties already agree the federal government should provide.
A government shutdown is unusual because it deliberately connects two separate questions: what should government spend, and should existing government functions continue while politicians disagree about that spending? In most parliamentary systems, failure to pass a budget can threaten the survival of the government itself, potentially producing elections or a change in administration. America’s separation of powers produces a different mechanism. Congress controls appropriations while the president executes the law, and when the institutions cannot agree on funding, substantial portions of government can legally lose authority to spend. The result gives political factions leverage because refusing agreement creates immediate costs that can be used to pressure the other side. But that leverage comes from imposing costs largely on people outside the negotiation. An air-traffic controller missing a salary, a family waiting for food assistance or a traveller facing a cancelled flight does not control the congressional bargaining process. Shutdown politics therefore works partly by turning public services into bargaining collateral.
The current confrontation began primarily because Democrats wanted Republicans to negotiate an extension of enhanced health-insurance tax credits created during the pandemic and later extended under the Inflation Reduction Act. Those subsidies reduce insurance premiums for millions of people purchasing coverage through Affordable Care Act marketplaces, but extending them costs the federal government substantial revenue. Republicans have argued that healthcare subsidies should be debated separately from the basic obligation to fund government, while Democrats feared that reopening government first would remove the leverage required to force serious negotiations. Both positions contain an internally coherent bargaining logic. Republicans ask why one policy dispute should determine whether unrelated federal agencies operate. Democrats respond that without leverage, a minority party has little ability to force the majority to negotiate at all. The shutdown therefore is not principally evidence that politicians cannot understand compromise; it is evidence that the institutional system can make imposing disruption strategically rational.
That distinction matters because political incentives can produce socially inefficient outcomes even when individual politicians behave rationally. Imagine a shutdown imposes $10 billion of total economic and administrative costs but gives each party some probability of obtaining a policy objective it values highly. Each side can rationally continue the confrontation because much of the $10 billion cost is paid by federal workers, businesses and citizens rather than directly by the politicians making the decision. Economists would describe part of this as an externality. The negotiating actors do not internalise the full cost of their bargaining strategy, so shutdowns can continue longer than they would if Congress itself paid the economic price directly.
The disruption to SNAP demonstrates the distributional problem particularly clearly. Food assistance normally supports tens of millions of low-income Americans, but November benefits became uncertain after the shutdown exhausted ordinary funding authority. Courts ordered the administration to use contingency resources, the administration appealed, states received changing instructions and the Supreme Court temporarily allowed the federal government to withhold enough money to prevent full payments while litigation continued. The constitutional questions may be complicated, but the economic incidence is straightforward: households with the least financial liquidity experience interruptions most severely. A wealthy household can absorb a delayed government payment from savings. A family already spending most income on rent, food and utilities cannot. The welfare cost of losing one dollar therefore depends heavily on who loses it.
This is why shutdown costs cannot be measured simply by adding delayed federal expenditure. Much government spending eventually occurs after agencies reopen and federal workers normally receive back pay, meaning part of the apparent economic loss is postponed rather than permanently destroyed. But some output cannot be recovered. A cancelled trip is not necessarily taken later. A delayed permit can postpone investment. Scientific experiments can be disrupted. Contractors may never recover lost revenue. Businesses around federal facilities lose customers. Administrative backlogs create additional labour costs after reopening. The economic effect therefore resembles a temporary negative productivity shock: the government eventually pays much of the money anyway while receiving fewer services during the period of disruption.
Air travel demonstrates how small disruptions inside one part of government can propagate through a much larger private economy. Air-traffic controllers and security personnel perform functions whose economic value extends far beyond their salaries. When staffing shortages produce delays, aircraft arrive late, crews exceed working-hour limits, passengers miss connections and companies lose productive time. The federal worker therefore occupies a network node. Removing one node can impose costs many times larger than the government’s direct payroll saving. This is why assessing public-sector productivity simply by comparing government salaries with private-sector wages can be misleading. Some public functions support enormous quantities of private activity.
The shutdown has also disrupted the collection and publication of economic statistics, potentially affecting data on employment and inflation. This sounds less dramatic than food benefits or airports but illustrates another hidden public good. Businesses, investors and the Federal Reserve rely on government statistics to make decisions. If reliable employment or inflation data disappear temporarily, uncertainty increases and monetary policymakers have less information when setting interest rates. A statistical agency therefore creates value partly by reducing uncertainty across millions of private decisions. The economic return is difficult to see precisely because everybody normally assumes the data will exist.
The political outcome is likely to frustrate much of the Democratic base because the emerging agreement reopens government without guaranteeing extension of the healthcare subsidies Democrats made central to the confrontation. A group of moderate Democratic senators concluded that the accumulating costs to federal employees, travellers and benefit recipients had become too large to justify continuing the shutdown merely for a promise that Republicans might eventually concede. Other Democrats argue that reopening government now sacrifices the leverage created by forty days of resistance without obtaining the central policy objective. That disagreement is essentially about when sunk costs should stop affecting strategy. After enduring forty days of shutdown, politicians naturally feel pressure to obtain something large enough to justify the suffering already incurred. Economically, however, costs already incurred should not determine whether additional costs are worthwhile. The correct question on day forty is whether the expected benefit of day forty-one exceeds the additional cost of continuing.
This is politically difficult because voters do not evaluate decisions like perfectly rational forward-looking models. If Democrats accept almost the same arrangement they could theoretically have accepted weeks earlier, critics can ask what the shutdown accomplished. Politicians therefore can become trapped by the need to justify previous sacrifice. Wars exhibit the same mechanism on a much more serious scale: governments continue fighting partly because accepting an agreement today makes earlier casualties appear wasted. Sunk costs do not disappear economically, but their political meaning can prolong conflict.
Republicans face their own incentive problem. If refusing negotiation eventually causes enough opposition senators to break ranks, future congressional majorities learn that waiting can defeat shutdown strategies. That may reduce the effectiveness of shutdown threats, which could be institutionally beneficial. Yet the opposite lesson is also possible: because the shutdown created significant political attention around healthcare subsidies, Democrats may conclude that disruption successfully elevated an issue even without obtaining immediate legislation. Reuters reported that the healthcare issue gained public attention during the confrontation even as Democrats ultimately fractured over whether to continue it. Political actors therefore can disagree about success because one measures legislative concessions while another measures agenda-setting.
The broader fiscal problem remains completely unresolved. The United States is running large structural deficits, public debt is extremely high and both parties continue protecting politically important spending or tax preferences. Republicans generally resist broad tax increases and have recently enacted major tax reductions. Democrats strongly resist cuts to major social programmes and want to preserve healthcare assistance. Both parties largely protect Social Security and Medicare, while defence and interest costs consume substantial resources. Government shutdowns therefore create the appearance of extreme fiscal conflict without necessarily confronting the programmes that drive the long-term budget trajectory. This is perhaps the greatest irony of the entire exercise: Washington can shut much of the government over comparatively narrow funding disagreements while leaving the much larger structural fiscal imbalance largely untouched.
A more rational system would separate the continuity of government from ordinary budget bargaining. Congress could adopt automatic continuing resolutions under which funding temporarily continues at the previous year’s level whenever legislators miss a deadline, perhaps with small automatic reductions creating incentives to reach agreement. Politicians could continue negotiating intensely without air-traffic controllers, food-assistance recipients or scientific agencies becoming bargaining instruments. Critics would argue that automatic funding reduces pressure on Congress to pass proper budgets. That is true. The question is whether the pressure created by shutdowns actually produces better budgeting or simply greater disruption. The repeated American experience suggests the latter often dominates.
There is a democratic argument for forcing visible fiscal choices. Legislators should not be able to avoid difficult decisions indefinitely through automatic spending. But accountability should attach costs primarily to the decision-makers rather than unrelated citizens. Elections already provide a mechanism for voters to punish legislators who fail to govern. Shutting government adds an additional mechanism whose economic incidence is highly uneven and only indirectly connected to responsibility. Institutional incentives should make political failure costly to politicians before making it costly to the public.
The healthcare dispute will return almost immediately because the proposed agreement only promises a later Senate vote on extending the Affordable Care Act subsidies. There is no guarantee that such legislation will pass the Senate, survive the House or receive Trump’s signature. The underlying policy question therefore remains legitimate: how much should government subsidise insurance premiums, who should qualify and how should the cost be financed? Those questions deserve an ordinary legislative debate with distributional estimates and fiscal analysis. They should not require airport disruption and uncertainty over food assistance merely to reach the agenda.
The shutdown has consequently demonstrated both the strength and weakness of American institutions. The government did not collapse. Courts continued functioning sufficiently to adjudicate disputes, states attempted to fill gaps, Congress continued negotiating and essential national-security functions persisted. The constitutional system can survive enormous partisan disagreement. But survival is an extremely low standard for institutional quality. A wealthy state should be evaluated by whether it can perform predictable routine functions without repeatedly manufacturing avoidable crises around them.
The final agreement, if enacted, will reopen agencies and eventually provide back pay. Government workers will return, benefits will resume more normally and airports will recover. Much of Washington will then move rapidly towards the next political confrontation. That would miss the institutional lesson. A forty-day shutdown is not simply evidence that one party behaved irresponsibly and the other eventually prevailed. It is evidence that the rules of the system allow both parties to believe temporarily disabling government can improve their bargaining position.
Until those incentives change, another shutdown will remain possible whenever political polarisation becomes sufficiently intense. The healthcare issue may be different next time, but the mechanism will be the same. The real reform is therefore not discovering politicians who dislike shutdowns. It is designing a budget process in which politicians can fight over policy without needing the government itself to stop functioning while they do it.
