The Emptying of Free Donbas

Saturday 15 August 2026

There is a particular kind of military victory that does not require the capture of a city. It consists instead in making the city progressively impossible to inhabit. The buildings may remain in Ukrainian hands, the blue-and-yellow flag may still fly above the municipal administration and Ukrainian soldiers may continue to occupy trenches and fortified positions around its approaches. Yet the schools have closed, the hospitals function intermittently, electricity and water have become uncertain, shops have disappeared, roads are travelled only at considerable risk and the population has dwindled to a fraction of what it once was. At some point a city ceases to be a city in the ordinary sense and becomes principally a military position surrounded by ruins.

This is increasingly the predicament of the remaining Ukrainian-controlled cities of the Donbas. Russia has not yet achieved its central strategic objective in Donetsk Oblast: the destruction or capture of Ukraine’s great defensive urban agglomeration running broadly through Kostiantynivka, Druzhkivka, Kramatorsk and Sloviansk. But it has achieved something else of potentially profound importance. Through years of bombardment, drone warfare, destruction of infrastructure and the steady contraction of the safe civilian hinterland, Moscow has progressively emptied much of free Donbas of the people whose presence gives Ukrainian sovereignty there its ordinary human meaning.

The figures are remarkable. At the beginning of January 2026, Ukrainian authorities reported approximately 192,600 civilians remaining in the Ukrainian-controlled part of Donetsk Oblast. By 22 July the figure had fallen to approximately 124,500. In July alone, according to the Donetsk Regional State Administration, some 7,800 people had evacuated or left the oblast under their own arrangements. More than 3,100 people left the Kramatorsk community during the first three weeks of July, more than 2,000 left Sloviansk and hundreds more departed Druzhkivka and Kostiantynivka.

This demographic retreat has accompanied the military one. Kostiantynivka illustrates the process in its most advanced form. A substantial industrial city before the full-scale invasion, it had been reduced to some 2,500 inhabitants by April 2026, with ordinary evacuation vehicles unable safely to reach the city and some residents leaving on foot or with military assistance. Municipal services have progressively disappeared. By summer Russian troops were fighting their way into the city itself, beginning what may prove to be the first great urban battle for the fortress belt proper.

This matters because the expression “fortress belt” can be misleading. It conjures images of concrete bunkers, minefields and elaborate defensive works. All these things exist. Yet the real fortress is geographical and urban. The belt consists of interconnected cities, industrial districts, railway lines, rivers, elevated terrain, roads, villages and prepared positions that together deny an attacker the freedom of movement required for rapid mechanised warfare. The urban areas themselves are enormous defensive obstacles. As recent reporting from the region has observed, although rivers, woodland and higher ground all assist Ukraine’s defence, it is the towns and cities themselves that constitute the greatest barrier to Russian advance.

The importance of this system explains why Russia has taken so long to reach its gates. The Russian armed forces have spent years consuming towns that on a map appear merely to be approaches to the principal Ukrainian conurbation. Bakhmut, Avdiivka, Chasiv Yar and the long struggle around Pokrovsk have each required extraordinary expenditure of men, ammunition and equipment. The apparent slowness of the Russian advance should therefore not obscure its strategic logic. Russia has been stripping away successive layers of Ukrainian defence until the principal cities of free Donbas themselves become exposed.

The battle for Kostiantynivka is accordingly more consequential than another incremental movement of the front line might suggest. Russian military and political sources have themselves presented its capture as a prerequisite for advancing upon Druzhkivka and ultimately the Kramatorsk-Sloviansk agglomeration. Independent assessments have been more cautious: Russia may continue to make tactical gains without possessing the forces or methods necessary for a rapid collapse of the entire fortress belt. Urban warfare favours the defender and the Russian Army has repeatedly demonstrated that it can spend months, or years, obtaining relatively modest territorial objectives.

Nevertheless there is another dimension to the Russian campaign that proceeds faster than the movement of infantry across the map. That is the expansion of what might be called the zone of uninhabitability.

Modern drone warfare has transformed the relationship between the front and the rear. In earlier wars a civilian might rationally remain twenty or thirty kilometres behind a stable front line because artillery fire was concentrated upon military positions closer to the battlefield. In contemporary Donbas, inexpensive reconnaissance and attack drones have blurred that distinction. Roads, petrol stations, civilian vehicles, electrical infrastructure, warehouses and municipal facilities may all lie within an expanding area of danger. Longer-range drones have extended severe access constraints far beyond the immediate trenches. Humanitarian organisations reported earlier this year that Druzhkivka, Kramatorsk and Sloviansk had suffered particularly serious deterioration in access to civilian services while Kostiantynivka had already lost much of its functioning municipal infrastructure.

This produces an insidious strategic effect. Russia does not need immediately to occupy Kramatorsk to alter the character of Kramatorsk. It needs only to make ordinary life progressively intolerable there.

The process has a grimly familiar sequence. Long-range attacks become more frequent. Families with children leave first. Schools move online or close. Businesses cease trading because customers and employees disappear. Medical personnel evacuate. Public transport deteriorates. Utilities become unreliable because repair crews cannot safely reach damaged infrastructure. Elderly inhabitants and those unwilling or unable to abandon their homes remain behind. Eventually the population consists disproportionately of soldiers, emergency workers, pensioners and a diminishing number of civilians who have resolved that they will leave only when remaining becomes physically impossible.

Then the drones come closer.

Eventually the principal roads themselves become dangerous. The city ceases to possess a conventional civilian economy. The last supermarkets close. Humanitarian deliveries become military-logistical operations. Finally the settlement acquires the character already familiar from Bakhmut, Avdiivka and other destroyed cities of eastern Ukraine: there may still be civilians hiding in basements but the city has effectively become a battlefield.

This is not necessarily evidence of a deliberate Russian master plan to depopulate every Ukrainian-held city before capturing it. Warfare produces displacement irrespective of intention and Ukrainian authorities themselves have ordered or encouraged evacuations precisely because leaving civilians under intense bombardment is unconscionable. Yet from Moscow’s perspective the strategic effect is advantageous irrespective of the precise mixture of intention and consequence. A depopulated Donbas is easier to fight over than an inhabited one.

It also changes the political meaning of territorial possession.

The struggle for Donbas has always contained a contradiction. Russia claims these territories principally through assertions about the political identity and wishes of their inhabitants. Yet Russian methods of warfare have repeatedly destroyed the communities in whose name Moscow professes to act. Mariupol was supposedly being “liberated” while being devastated. Bakhmut was supposedly Russian territory while Russian artillery was reducing it to ruins. The same paradox now hangs over the remaining Ukrainian cities farther west.

The more effectively Russia empties free Donbas, the more grotesque its original political justification for the war becomes. Yet simultaneously the demographic disappearance of Ukrainian civilian life creates a practical problem for Kyiv. Sovereignty is not merely the possession of territory. It is the maintenance of communities upon it. Ukraine can hold a line militarily while gradually losing the society behind the line.

This is why Kramatorsk and Sloviansk are so important.

They are not merely two further objectives on a Russian general staff map. Together with Druzhkivka and the surrounding settlements, they constitute the last great concentration of urban Ukrainian life in Donetsk Oblast under Kyiv’s control. Kramatorsk has served since 2014 as the administrative centre of Ukrainian-controlled Donetsk. Sloviansk possesses still greater symbolic significance: it was there in 2014 that the Russian-sponsored separatist insurrection first acquired unmistakably military form and there that Ukrainian forces achieved one of their earliest significant victories when separatist forces withdrew from the city.

Were Russia eventually to capture both cities, the result would therefore represent something qualitatively different from the loss of another collection of villages. Moscow could plausibly claim that the territorial struggle for Donetsk Oblast was approaching its conclusion. More importantly, Ukraine would have lost the principal urban centres from which the restoration of normal Ukrainian political and economic life in Donbas might someday proceed.

The consequences would extend beyond Donbas. The fortress belt is also a geographical shield. Beyond it the density of major urban obstacles changes and Russian forces would possess potential avenues towards the broader Ukrainian interior. This does not mean that the fall of Sloviansk would suddenly expose Kyiv to Russian tanks; modern armies do not advance hundreds of kilometres merely because a line on a map has been breached. But the disappearance of the fortress belt would unquestionably worsen Ukraine’s long-term strategic geography. Recent analysis of the defensive system has therefore emphasised that its loss could eventually open approaches towards major Ukrainian centres farther west and north.

Russia consequently faces a choice of its own. It can attempt to storm these cities, accepting the immense casualties that urban warfare against prepared Ukrainian positions would entail. Or it can continue doing what has already produced substantial results: constricting logistics, expanding drone kill-zones, destroying infrastructure and making civilian existence increasingly difficult while seeking tactical penetrations around and between the cities.

The latter strategy has considerable attractions. Time becomes a weapon.

Every month of bombardment removes another portion of the civilian population. Every destroyed apartment block makes eventual reconstruction more expensive. Every closed factory weakens the economic reason for displaced inhabitants to return. Every family that establishes itself permanently in Kyiv, Dnipro, Lviv, Warsaw, Berlin or elsewhere becomes less likely to reconstruct its former life in Donetsk Oblast. A temporary evacuation can gradually become permanent demographic relocation.

This is perhaps the most dangerous Russian success in free Donbas because it is substantially irreversible on military timescales.

A Ukrainian brigade can retake a village. Engineers can rebuild a bridge. Minefields can eventually be cleared. But recreating a city is another matter altogether. Cities are networks of memory, employment, family, education and habit. Once their inhabitants have spent four or five years elsewhere, their children attend different schools, their parents find different employment and their businesses acquire different customers. Even victory does not necessarily bring them home.

Hence the next struggle for Donbas will take place simultaneously on two battlefields. One is the obvious battlefield of trenches, drones, artillery and infantry assaults stretching around Kostiantynivka and towards Druzhkivka, Kramatorsk and Sloviansk. The other is demographic.

Ukraine must preserve enough civilian life in free Donbas that there remains something recognisably Ukrainian to defend.

This does not mean encouraging civilians to remain where doing so would expose them to unreasonable danger. Quite the opposite. Evacuation from threatened areas is frequently essential and the Ukrainian authorities have properly expanded it as Russian strikes have intensified. But Kyiv and its European partners must begin thinking about displacement from Donbas not merely as a humanitarian emergency but as a strategic problem of national survival.

The displaced population of Donetsk Oblast must retain reasons to imagine returning. Housing assistance, employment programmes, municipal institutions in exile, continuity of schools, property records, compensation mechanisms and credible reconstruction plans are therefore not peripheral social policies. They are components of territorial defence. If Russia’s strategy produces empty cities, Ukraine’s answer must be to preserve the invisible communities belonging to those cities until physical return becomes possible.

There is also a lesson here for Western governments tempted by superficially simple territorial compromises. On diplomatic maps the remaining Ukrainian-held portion of Donetsk Oblast can appear as an inconveniently shaped piece of disputed territory. It is sometimes discussed accordingly — as though kilometres of land might be exchanged for kilometres elsewhere or surrendered as the price of a ceasefire.

This misunderstands what is at stake.

Kramatorsk and Sloviansk are not empty squares upon a negotiating board. They are the remaining political and demographic anchors of free Donbas. To compel Ukraine to surrender them without a battle would achieve for Russia what years of extraordinarily costly warfare have thus far failed fully to accomplish. It would also place hundreds of thousands of displaced Ukrainians in the position of knowing that the homes to which they hoped eventually to return had been permanently assigned to the state whose invasion forced them to flee.

There is a deeper historical irony. For more than a decade Moscow has justified its intervention in eastern Ukraine by claiming that Donbas possesses a distinctive identity requiring protection from Kyiv. Yet the practical consequence of Russian policy has been the destruction of much of the Donbas that actually existed. Industrial cities have been shattered, populations dispersed and communities uprooted across Europe. The physical territory remains but the human geography upon it is being transformed.

Russia may therefore eventually discover that it has conquered what its own armies have emptied.

Ukraine’s task is to ensure that this does not happen to the remainder.

The battle for the fortress belt will probably not end quickly. Assessments earlier this summer continued to caution that tactical Russian gains around Kostiantynivka did not imply an imminent operational breakthrough across the entire defensive system. Russia still faces formidable urban terrain, prepared Ukrainian defences and the inherent difficulty of sustaining offensive operations through successive cities. But military endurance alone is no longer enough. The civilian evacuation figures demonstrate that another clock is running alongside the battlefield clock.

In January approximately 192,600 civilians remained in Ukrainian-controlled Donetsk Oblast. By late July there were about 124,500. Those numbers should be read not merely as humanitarian statistics but as one of the most important strategic indicators of the war.

The final contest for free Donbas may therefore be decided by two questions. The first is whether Russian soldiers can force their way through the fortress belt. The second is whether Russia can make the fortress belt cease to contain functioning cities before its soldiers have captured them.

Ukraine has so far proved remarkably successful at denying Russia the first victory. Russia has been considerably more successful at achieving the second.

That is why the coming battles for Druzhkivka, Kramatorsk and Sloviansk matter beyond their immediate military significance. They concern not merely where the front line will run but whether there will remain a viable Ukrainian Donbas behind it. The ruins of Bakhmut demonstrate what Russia can capture. The increasingly deserted streets farther west demonstrate something more disturbing — what Russia can destroy without yet capturing.

The future of Donbas may ultimately turn upon which of these processes happens first: the exhaustion of Russia’s capacity to advance or the exhaustion of civilian life in the cities standing in its way.

Upon that grim competition rests the future of free Donbas.