Donald Trump returned to the White House on 20 January and immediately began testing the boundaries of presidential power. Within days, he signed a large wave of executive orders and presidential actions covering immigration, energy, federal employment, diversity programmes, climate policy and foreign affairs. He declared a national emergency at the southern border, ordered additional military resources towards border enforcement, moved to restrict asylum, suspended the US refugee-admissions programme, began another withdrawal from the Paris climate agreement and ordered the United States to leave the World Health Organization. He also pardoned or commuted the sentences of roughly 1,500 people charged or convicted in connection with the 6 January 2021 Capitol attack, while directing the federal government to dismantle diversity, equity and inclusion programmes. The quantity is politically deliberate. Trump wants the contrast with the Biden administration to become visible immediately. Yet the more important question is institutional rather than ideological: how much can an American president actually transform through executive action when legislation, courts, federalism and administrative capacity still constrain what presidential signatures can accomplish?
Executive orders often create an illusion of instantaneous government. The president signs a document surrounded by cameras and the policy appears to have changed. In reality, executive orders operate inside authority already granted by the Constitution or Congress. A president can direct executive agencies, change enforcement priorities and reverse many policies established by previous presidents. He cannot simply create new statutory authority because Congress remains the legislature. This distinction will determine much of Trump’s second presidency because his political style rewards visible unilateral action while the American constitutional system deliberately fragments power. Signing an order and implementing an order are separate stages of government.
Immigration will provide the first major test. Trump has promised the largest deportation operation in American history, stronger border enforcement and much narrower access to asylum. Presidents possess substantial authority over immigration enforcement because the executive controls agencies such as Immigration and Customs Enforcement and Customs and Border Protection. Yet removing millions of people requires far more than presidential willingness. Officers need to identify individuals, make arrests, process cases, arrange detention, obtain travel documents and persuade destination countries to accept deportees. Immigration courts already face enormous backlogs.
This is the difference between legal authority and administrative capacity.
A government can possess the authority to deport someone and still lack the personnel, detention space or judicial capacity required to do it quickly.
Mass deportation therefore becomes a logistics problem before it becomes an ideological one.
Suppose the government wanted to remove one million people annually. That would require an average of roughly 2,700 removals every day, continuously, before considering arrests that fail, legal challenges, asylum claims or countries refusing repatriation. Scaling enforcement therefore requires substantial appropriations from Congress and cooperation across federal, state and local institutions.
This is why state capacity limits populist maximalism. Campaigns can promise outcomes directly. Governments need supply chains.
The economic consequences also depend on scale. Undocumented immigrants are concentrated disproportionately in industries including agriculture, construction, hospitality and food processing. Removing workers can reduce labour supply, potentially raising wages for competing domestic workers while increasing production costs and prices.
The distribution matters.
A construction worker may benefit from tighter labour supply.
A household purchasing a newly built home may pay more.
A farmer may face higher labour costs.
A worker replacing a migrant may receive higher wages.
Immigration therefore is neither simply “good for the economy” nor “bad for workers”. It creates distributional effects between labour, capital and consumers.
The correct policy debate should acknowledge them.
Trump’s attempt to reinterpret birthright citizenship creates a much clearer constitutional confrontation. The Fourteenth Amendment states that people born in the United States and subject to its jurisdiction are citizens. Trump has issued an executive order seeking to deny automatic citizenship to some children born to parents without permanent legal status.
Federal courts are already being asked to intervene.
This is precisely what judicial review is for.
The president has one interpretation of executive and constitutional authority.
Opponents have another.
Courts determine whether the executive’s interpretation fits existing law.
A judge blocking a presidential action therefore is not automatically “overruling democracy”.
Trump won an election.
Judges exercise authority under another part of the constitutional system.
Democratic legitimacy is distributed rather than concentrated.
This distinction will become increasingly important because Trump and his supporters often describe judicial or bureaucratic resistance as attempts by an unelected establishment to frustrate voters.
Sometimes that criticism can be legitimate.
Courts can overreach.
Bureaucracies can resist elected policy excessively.
But the opposite principle matters equally.
Winning an election gives a president authority over the executive branch within law.
It does not transfer legislative or judicial authority into the presidency.
This is why constitutional government intentionally feels inefficient.
Efficiency is not the only objective.
Preventing one institution from becoming capable of changing everything quickly is itself a design feature.
Trump’s withdrawal from the Paris climate agreement demonstrates an area where presidential discretion is much larger because international commitments frequently depend heavily on executive authority. Biden re-entered the agreement after Trump withdrew during his first term; Trump is withdrawing again.
This creates a serious credibility problem for American foreign policy.
Why should another government restructure long-term climate policy around an American commitment that can reverse every four years?
This is the policy volatility cost of polarisation.
Countries negotiate not only with the current US president but with expectations about future presidents.
If international partners believe American commitments are temporary, they discount their value.
The same problem can affect trade agreements, security arrangements and environmental policy.
Domestic polarisation therefore reduces international bargaining power because presidential promises contain a higher probability of reversal.
Ironically, stronger executive authority can worsen this problem.
Policies enacted through legislation are harder for successors to reverse.
Policies enacted primarily through executive orders can disappear through another executive order.
Unilateralism creates speed at the cost of durability.
This is the speed-versus-commitment trade-off of executive power.
Trump can reverse Biden quickly because Biden used executive authority.
A future Democratic president can reverse Trump for exactly the same reason.
If either party wants durable policy, eventually it needs legislation capable of surviving alternation.
The decision to leave the World Health Organization creates another example of sovereignty versus collective capacity. Trump argues that the WHO handled COVID badly, remains excessively influenced by China and demands unfair financial contributions from the United States. Those criticisms deserve examination; international organisations should not become immune from accountability simply because their objectives are valuable.
But withdrawal carries opportunity costs.
Pandemics are inherently cross-border externalities.
A virus emerging in one country can reach another before governments understand what it is.
International surveillance, data sharing and coordination therefore produce public goods.
The relevant question is not whether the WHO is imperfect.
It obviously is.
The question is whether the United States gains more from reforming it than leaving it.
This is the same institutional logic behind BRICS and the United Nations.
Exit increases autonomy while reducing influence over the institution left behind.
The country outside the organisation no longer pays the same costs.
It also possesses less formal authority over its decisions.
Trump’s broader worldview consistently favours the first side of this trade-off.
Sovereignty.
Bilateral bargaining.
Reduced institutional constraint.
His argument is that American power is large enough that multilateral institutions frequently constrain Washington more than they strengthen it.
This theory is not irrational.
A powerful country has stronger outside options than a weak one.
The United States can negotiate individually in ways smaller countries cannot.
The counterargument is that institutions allow America to convert raw power into rule-setting power.
A dollar spent supporting an institution can generate rules followed by dozens of states.
Leaving may save the dollar while surrendering some influence over the rules.
Again, sovereignty and power are not identical.
Trump’s 6 January pardons raise a different institutional question. Presidential pardon authority is extraordinarily broad under the Constitution for federal offences. Trump therefore possesses substantial legal power to issue them.
The more difficult issue is the norm created by their use.
If people commit offences while attempting to help a political leader remain in power and that leader later pardons them, future supporters can update their expectations about the cost of political violence.
This creates a moral-hazard problem.
Punishment deters partly because individuals expect consequences.
A credible expectation of future pardon can reduce those expected consequences.
The relevant question therefore is not simply whether Trump had constitutional authority.
He did.
It is whether using that authority in this way changes incentives surrounding future political mobilisation.
Again, legality is not the end of institutional analysis.
Constitutions often grant powers broad enough that restraint depends partly on norms.
Presidents can pardon.
Prosecutors can investigate.
Legislatures can impeach.
Judges can interpret.
Democratic systems work because actors usually avoid maximising every formal power continuously.
If every institution pushes authority to its absolute legal boundary, government becomes permanent constitutional confrontation.
Trump’s second presidency appears likely to test more of those boundaries than most.
Federal employment may become another major battlefield. Trump has long argued that presidents cannot control government effectively when career officials are protected strongly enough to resist elected policy. His administration is moving towards increasing presidential control over parts of the civil service.
There is a legitimate democratic argument here.
Elections should change policy.
A permanent bureaucracy should not be able to nullify that change simply because officials disagree with voters.
But civil-service protections exist for another reason.
Before professional bureaucracies, governments often operated through patronage systems where jobs went to political loyalists.
That produced corruption, weak expertise and enormous turnover after elections.
Modern civil services therefore separate political direction from administrative professionalism.
The optimal system needs both.
Presidents choose policy.
Civil servants implement it competently.
Officials can refuse unlawful instructions without treating ordinary political disagreement as illegality.
Again, the key distinction is loyalty to lawful government versus loyalty to an individual leader.
Trump believes the first administration contained too much bureaucratic resistance.
His critics fear the second will contain too little.
Both failure modes are possible.
The next four years will show which becomes more important.
Trump’s energy orders follow the same philosophy of rapid executive reversal. He has declared a national energy emergency, promised greater fossil-fuel production and moved to reverse Biden-era climate and energy regulations.
The economic effects may be less immediate than rhetoric suggests because American oil production is already near record levels and private companies decide whether additional drilling is profitable.
Presidents influence incentives.
They do not command investment.
An oil company drills when expected future prices justify the cost.
If global supply is abundant and prices relatively low, regulatory permission alone does not guarantee a drilling boom.
This is another example of the distinction between policy authority and market response.
Governments control rules.
Firms respond according to expected returns.
The same applies to Trump’s broader promise of lowering prices.
Presidents possess surprisingly little direct control over most consumer prices.
They can influence energy regulation, tariffs, taxes and fiscal policy.
The Federal Reserve controls monetary policy independently.
Global commodity prices respond to worldwide supply and demand.
Housing prices depend heavily on local construction constraints.
A president therefore cannot simply order inflation away.
This makes Trump’s second term politically risky because he won partly by associating Biden and Harris with high prices.
The grievance now becomes his responsibility.
That is the incumbency transfer.
Opposition converts complexity into blame.
Government inherits complexity.
If grocery prices remain high, Trump cannot continue attributing them indefinitely to Biden.
If tariffs increase some prices, voters will evaluate outcomes rather than theoretical intentions.
If deportations create labour shortages, those become government consequences.
This is why the second Trump presidency will be analytically more informative than the first.
In 2017, Trump inherited a strong economy and an administration his own movement had not expected fully to control.
In 2025, he returns with a party transformed around him, a clearer programme, experienced advisers and much greater knowledge of executive government.
There are fewer excuses available.
If Trumpism is a coherent governing philosophy rather than principally an opposition movement, this is the environment in which it should demonstrate that.
The first week therefore should not be judged primarily by the number of executive orders signed.
Quantity measures activity.
Not effectiveness.
The useful questions come later.
How many deportations actually occur?
Do border crossings fall sustainably?
Do courts uphold the birthright-citizenship order?
Do tariffs increase domestic investment enough to compensate for higher costs?
Does leaving international institutions increase American bargaining power?
Does restructuring the bureaucracy improve implementation or politicise administration?
Do energy policies lower household costs materially?
Those are outcome measures.
Politics prefers announcements because announcements are immediate.
Governance should be evaluated through implementation.
Trump clearly understands the political value of speed.
He wants voters to see that changing presidents changed government instantly.
In one sense, that is healthy.
Elections are supposed to matter.
A government that ignores the programme voters selected would weaken democratic accountability.
But America’s constitutional system intentionally ensures that elections do not give presidents control over every institution.
Congress still legislates.
Courts still review.
States still govern.
Markets still choose investments.
The Federal Reserve still sets interest rates.
The bureaucracy still needs to transform directives into programmes.
The central political argument of Trump’s second presidency may therefore be about whether those constraints represent obstacles to democracy or components of democracy.
Trump’s instinct is often that elected executive authority should dominate more strongly.
The constitutional tradition says power should remain fragmented even when fragmentation frustrates the winner.
The first week has already placed those theories in conflict.
That conflict is not evidence that the American system has failed.
It is the system operating under stress.
The real test is whether every actor accepts the eventual boundaries.
A president can challenge a law.
A court can reject his interpretation.
Congress can change the law.
Voters can replace Congress or the president later.
That cycle is democratic precisely because no single election ends it.
Trump has returned to office with a genuine mandate to change American policy.
The next four years will determine whether he interprets that mandate as authority to govern within institutions or authority to subordinate the institutions that limit him.
Those are very different projects.
His first week suggests the distinction will define the second term.
