The War Leader's Half-Life: Zelensky, Churchill, and the Physics of Wartime Approval

Nine months. That is roughly how long the average wartime approval rally lasts before political gravity drags it back to earth, according to decades of political science research. Volodymyr Zelensky has been fighting that gravity for four years. His approval stood at 62% in early March 2026, measured by the Kyiv International Institute of Sociology between March 1 and 8. Four years into a war that has killed tens of thousands of his citizens and displaced millions, the Ukrainian president commands public confidence that most wartime leaders lose within their first year of conflict. The number is not magic. It follows patterns, and those patterns have a history worth measuring.

The Rally Effect: A Short-Lived Drug

Political scientists have a name for what happens when a nation faces sudden external threat: the rally-around-the-flag effect. The concept entered academic literature through John Mueller's 1973 study of American wartime opinion, and it behaves with near-pharmaceutical predictability. A dramatic external event, typically involving military action or a direct threat to national sovereignty, produces a sharp spike in presidential approval. The spike is intense and brief.

Mueller found that rally events typically generate a 5 to 10 percentage point increase in approval, occasionally more for events of sufficient magnitude. The effect requires three conditions: the event must be international, it must directly involve the national leadership, and it must be specific, dramatic, and sharply focused. What Mueller also found, and what subsequent researchers confirmed, is that the boost decays. It does not sustain.

Matthew Baum refined the model in 2002, arguing that the information environment and the varying susceptibility of different voter groups shape how long a rally persists. When media coverage is diverse and opposition voices have access to the public, the decay accelerates. When the information space narrows, as it does in the early phase of a genuine security crisis, the rally lasts longer. Across the broader rally literature, from Oneal and Bryan's 1995 study of US foreign policy crises to Parker's work on rally dynamics, the consensus holds that the typical rally returns to pre-crisis approval levels within six to nine months. The political half-life, so to speak, is short.

This baseline matters because it provides the measuring stick. Every wartime leader in the modern polling era has been subject to rally physics. The question is not whether the decay happens, but how fast, and what determines the rate.

Churchill: The Man Who Won the War and Lost the Peace

Winston Churchill's approval during the Second World War represents the foundational dataset for wartime leadership polling. British Gallup began tracking satisfaction with Churchill as Prime Minister from 1940 onward, and the numbers tell a story more complex than the national myth.

Churchill entered office in May 1940 at a moment of genuine existential crisis. France was falling, the British Expeditionary Force was trapped at Dunkirk, and invasion appeared imminent. His early approval ratings reflected the severity of the moment: satisfaction with his leadership hovered between 78% and 89% throughout the war years, a remarkably high and sustained band. But the numbers were not static. Setbacks produced visible dips. The fall of Singapore in February 1942, Britain's worst military defeat in terms of prisoners taken, corresponded with a measurable decline in confidence. Reverses in North Africa created similar ripples. The public was not blindly loyal. It was conditionally supportive, with the condition being the perception of competent crisis management.

What makes Churchill's case instructive is the ending. On July 26, 1945, fewer than three months after VE Day, the British electorate handed Labour a landslide victory with 393 seats to the Conservatives' 197. Churchill, the man who had held 80%-plus approval through five years of existential war, was turned out of office the moment the crisis ended. The approval had not been for Churchill the governor. It had been for Churchill the war leader. The public distinguished between the two with surgical precision, and when the war function was no longer needed, the wartime approval evaporated.

This distinction sits at the center of every wartime leader's trajectory. Approval during crisis measures something specific: confidence in the leader's ability to navigate the present danger. It does not measure satisfaction with healthcare, the economy, or postwar vision. Churchill learned this at the ballot box.

Bush After 9/11: The Fastest Decay in Modern History

George W. Bush's post-September 11 approval arc is the textbook demonstration of rally mechanics in a democracy with a free press and functioning opposition. The numbers are stark enough to serve as a controlled experiment.

In the Gallup survey conducted September 7 through 10, 2001, four days before the attacks, Bush's approval sat at 51%. By the September 21-22 survey, ten days after the World Trade Center fell, it had reached 90%. A 39-point surge in less than two weeks, the largest rally spike in the history of American polling. The entire country, including overwhelming majorities of Democrats, expressed approval of the president.

The decay was visible almost immediately, though its pace varied. By January 2002, four months after the attacks, Bush still held roughly 84%. The invasion of Afghanistan had broad public support and the rally sustained. But the half-life was already measurable. By the first anniversary of September 11, approval had fallen to the high 60s. The invasion of Iraq in March 2003 produced a secondary rally, briefly pushing numbers back to 71%, but this second spike decayed faster than the first. Bush first dipped below 50% in late 2003, recovered intermittently, and then crossed permanently below that threshold during 2005, eventually bottoming at 25% in October 2008.

Two variables explain the speed of Bush's decay. First, the nature of the threat. September 11 was an acute shock, not a sustained existential danger to the American homeland. Once the immediate terror receded and daily life resumed, normal political evaluations returned. Voters started grading the president on the economy, on Iraq, on Katrina, on the things they had temporarily set aside during the rally. Second, the shift from defensive to discretionary war. Afghanistan enjoyed broad support as a direct response to an attack. Iraq did not. The moment the public perceived the war as a choice rather than a necessity, the protective shield of the rally dissolved.

Bush's curve demonstrates a principle: the rally effect's duration is proportional to the persistence and proximity of the threat. Remove the threat, or transform it from defensive to elective, and the half-life shortens dramatically.

Lincoln 1864: The War Leader Who Almost Lost

Modern polling did not exist during the American Civil War. There are no monthly approval ratings for Abraham Lincoln, no trend lines to chart. But the 1864 presidential election provides a hard data point that speaks as loudly as any survey, and the months leading up to it reveal a wartime leader whose survival in office hung on a single military event.

By the summer of 1864, the Union war effort had ground into a blood-soaked stalemate. Ulysses Grant's Overland Campaign had produced catastrophic casualties at the Wilderness, Spotsylvania, and Cold Harbor without capturing Richmond. William Sherman's advance toward Atlanta had stalled. Newspaper editorials across the North called for negotiated peace. The Democratic Party nominated George McClellan on a platform whose peace plank, drafted largely by Copperhead Clement Vallandigham, labeled the war a failure and called for an immediate armistice. McClellan himself repudiated the peace plank in his acceptance letter, insisting the Union must be preserved, but the platform signaled to voters that a Democratic victory could mean a negotiated end to the conflict.

Lincoln believed he would lose. On August 23, 1864, he wrote a private memorandum and asked his cabinet members to sign it without reading: "This morning, as for some days past, it seems exceedingly probable that this Administration will not be re-elected." He was preparing for a transfer of power to a president who might end the war on terms that preserved slavery and potentially allowed Confederate independence.

Then Sherman captured Atlanta on September 2. The military reality changed in a single day, and public opinion followed. Lincoln won the November election with 55% of the popular vote and 212 of 233 electoral votes. Where soldier votes were separately tallied, Lincoln carried approximately 78% of the military vote. The army, which had borne the cost of the stalemate, voted overwhelmingly for continuation once victory seemed achievable.

Lincoln's case introduces the variable that matters most for Zelensky: battlefield reality as the ultimate driver of approval. No amount of rhetorical skill, no communication strategy, no rally effect could overcome the public's perception of military failure. Only a tangible, visible success on the ground reversed the trajectory. The approval curve, had one existed, would have shown a crater in August 1864 and a sharp recovery in September.

Netanyahu's Cycles: The Approval Oscillator

Benjamin Netanyahu offers a different model from the single-arc trajectories of Churchill, Bush, and Lincoln. His career presents not one rally-and-decay sequence but many, stacked across decades. Each security crisis produces a spike, each postwar reckoning produces a correction, and the pattern reveals what happens when a leader's political identity becomes structurally intertwined with the security narrative.

Netanyahu's peacetime approval has historically oscillated between the low 30s and low 50s, depending on coalition dynamics and economic conditions. Each major security operation produced a measurable rally: the 2014 Gaza conflict, the 2021 Guardian of the Walls campaign, smaller operations in between. The rallies followed the standard pattern, spiking with the onset of military action and decaying as the operation ended and its costs became visible.

October 7, 2023 broke the pattern. The Hamas attack produced the kind of existential shock that would normally generate a massive rally effect. Trust in the government did briefly surge as the military response began. But the nature of the crisis, a catastrophic intelligence and security failure on Netanyahu's watch, meant that the rally carried within it the seeds of its own destruction. As the scale of the failure became clear and the hostage crisis persisted, the post-attack solidarity fractured faster than any precedent suggested it should.

Netanyahu demonstrates the limits of repeated rally-seeking and the asymmetry between different types of rallies. A rally caused by perceived competence, Churchill in 1940 or Zelensky in 2022, where the leader is seen as rising to meet a crisis, produces a different kind of loyalty than a rally caused by a failure that triggers solidarity. The former builds on trust. The latter builds on fear, and fear is a less durable foundation.

The Leaders Who Fell: When War Destroys Approval

The rally effect is not universal. Some leaders lose public support during conflict, and the pattern of failure is as instructive as the pattern of success.

Lyndon Johnson entered the Vietnam escalation with approval near 70% in early-to-mid 1965. The Gulf of Tonkin resolution the previous August had produced a modest rally, and the broader context of Cold War consensus gave Johnson room. By March 1968, his approval had fallen to 36%. The erosion was relentless and nearly linear: roughly 10 points per year as casualties mounted, as the credibility gap widened between official optimism and battlefield reality, and as the draft brought the war's cost into middle-class homes. Johnson chose not to seek reelection. The war had consumed his presidency.

The Argentine junta under Leopoldo Galtieri provides the inverse case: a manufactured rally that produced destruction. Galtieri's invasion of the Falkland Islands in April 1982 generated enormous nationalist enthusiasm. Crowds filled the Plaza de Mayo in support. But the rally was built on a military gamble, and when British forces retook the islands in June, the approval collapsed along with the regime. Within days of the military defeat, Galtieri was removed and the junta began its terminal disintegration. The Argentine case is the starkest warning in the dataset: a security crisis manufactured for domestic political purposes carries maximum risk because the leader owns the outcome completely.

The common thread in failed wartime approval is the nature of the conflict. Discretionary wars, wars of choice, erode faster than defensive ones. When the public perceives that the leader chose the fight, the leader bears personal responsibility for every cost. When the public perceives that the fight came to them, the leader receives the benefit of being the one who stands between the nation and the threat. Vietnam was Johnson's choice. The Falklands were Galtieri's gamble. This distinction explains more about approval trajectories than any other single variable.

Spain's experience with the Iraq War offers a parallel from a different angle. José María Aznar's decision to join the US-led coalition in 2003 faced overwhelming public opposition, and the March 2004 Madrid train bombings, coming three days before a general election, produced not a rally but a reckoning. The Spanish public punished the governing party for a war it had never supported. The rally effect requires the public to perceive the threat as genuine and the leader's response as legitimate. Without both conditions, the mechanism fails.

The Physics of Zelensky's Curve

With the historical comparisons mapped, Zelensky's four-year trajectory becomes legible not as a unique phenomenon but as a data point within a measurable system.

The baseline: in late 2021, months before the Russian invasion, Zelensky's approval had fallen to approximately 31%. He was a peacetime president struggling with unfulfilled reform promises, oligarch-related scandals, and the general attrition that affects most Ukrainian leaders by their third year in office. He was, by the numbers, a declining political figure.

The invasion produced one of the largest rally effects in the history of modern polling. By March 2022, KIIS measured his approval at approximately 90%, a roughly 60-point surge. For context, Bush's 39-point post-9/11 spike was previously considered the upper bound of rally magnitude. Zelensky's spike was roughly 50% larger. The difference lies in the nature of the threat: a full-scale ground invasion of the national territory, visible in every city, producing refugees in every household, is a qualitatively different experience from a terrorist attack, however devastating.

The decay followed, but at a rate that confounds the standard models. Where Bush lost 40 points in four years and Johnson lost 34 points in three, Zelensky's decline from 90% through the mid-70s by late 2023 and then to 53% in February 2026 represents roughly a 37-point decline over four years. The rate is comparable in magnitude but the trajectory is different: the floor is higher. At 53%, Zelensky's lowest point was still 22 points above his pre-invasion baseline. Even at his worst wartime moment, he commanded more public confidence than he had before the war began.

The February-March 2026 sequence is particularly illuminating. The dip from 61% in January to 53% in February coincided with the most intense period of US-Ukraine diplomatic friction, the Oval Office confrontation and its aftermath. The rebound to 62% in March, measured by KIIS between March 1 and 8, tracks the resolution of that crisis and what other analyses in this cluster identify as an externally triggered secondary rally. The pattern resembles Bush's mini-rally at the onset of the Iraq invasion, a secondary spike within the broader decay curve, but with a key difference: the trigger was not offensive military action but the perception of external pressure resisted.

Three variables explain why Zelensky's curve sits between Churchill's sustained wartime high and Bush's rapid decay. First, the threat persists. Unlike the discrete shock of September 11, the Russian invasion is ongoing, daily, and existential. The conditions that produce rally effects have not abated, which sustains a higher baseline. Second, the war is unambiguously defensive. No significant portion of the Ukrainian public views the conflict as a war of choice, which eliminates the primary mechanism through which wartime approval erodes. Third, the information environment, while not perfectly free, maintains enough opposition and independent media for the decay mechanism to function. The 32% disapproval is real and measurable, indicating that the rally is not artificially sustained by information control.

What the Curves Predict

The historical data does not predict Zelensky's future approval. It defines the boundary conditions within which his trajectory will move.

The Churchill boundary: sustained existential threat can maintain high wartime approval for years, but that approval measures crisis confidence, not governance satisfaction. When the threat ends, the wartime premium disappears. Churchill's 80%-plus approval did not survive VE Day by three months. Zelensky's wartime numbers, whatever they are when the conflict concludes, should not be expected to transfer to peacetime governance.

The Lincoln boundary: approval during protracted conflict depends ultimately on perceived battlefield trajectory. Lincoln nearly lost an election during a war that most of the public supported in principle, because the battlefield reality suggested the war could not be won. A negotiated settlement or frozen conflict that the Ukrainian public perceives as unfavorable could produce a Lincoln-style collapse in Zelensky's numbers regardless of how the rally effect has sustained until that point.

The Bush boundary: the gap between existential and discretionary perception is the single most important variable. As long as the Ukrainian public perceives the war as an existential defense rather than a choice, the decay rate remains slower than the historical norm. Any shift in that perception, through war fatigue, through unpopular mobilization policies, through a sense that the leadership is prolonging the conflict unnecessarily, would accelerate the decay toward the Johnson trajectory.

The Argentina boundary: at the extreme, leaders who are perceived as having manufactured or unnecessarily extended a crisis face not gradual decay but collapse. This boundary applies less to Zelensky, whose war was clearly imposed from outside, but it is worth noting as the outer limit of the model.

What 62% represents, measured against these historical curves, is something genuinely unusual. It is four years into a devastating conflict. It is roughly 30 points below the initial rally peak. It is roughly 30 points above the pre-war baseline. And it is higher than most political scientists would have predicted based on the decay constants observed in previous wartime cases. The threat persists, the war remains perceived as defensive, and the most recent external pressure event demonstrated that secondary rallies remain available.

Approval ratings during wartime are not report cards. They are seismographs, measuring the tremors of a society under sustained pressure. The 62% does not grade Zelensky's governance. It records Ukrainian society's stress tolerance in year four of an existential conflict, and it suggests that the irreducible core of wartime confidence, the element that remains after the half-life has done its work, has not yet been reached.