Causes and Consequences of the War in Eastern Ukraine: An Economic Geography Perspective

Abstract This essay provides an economic geography perspective on the causes and consequences of the war in eastern Ukraine. It focuses on the controversial proposition that the armed conflict in 2014 was triggered by domestic, economically determined factors. The essay argues that economic and material circumstances in the region had generated neither necessary nor sufficient conditions for a locally rooted, internally driven armed conflict. The role of the Kremlin’s military intervention was paramount for the commencement of hostilities. As the human and economic costs of the war continue to mount, Ukraine’s war-ravaged eastern regions face further depopulation, economic decline and erosion of development.

determinism, rooted in historical materialism. Consequently, the essay introduces human geography of the Donets Basin (Donetskyi basein or Donbas) and critically assesses the economic role which this region had played nationally prior to 2014. Subsequently, the prewar socioeconomic conditions of the region in question are compared and contrasted with disruption and destruction brought upon Donbas by the first five years of the undeclared Russo-Ukrainian war. In addition to the analysis of the post-2014 demographic, output, employment and income trajectories, the essay highlights the consequences of the forced displacement and de facto partitioning of Donbas's economy since 2014. Finally, the essay provides an evidence-based critique of historically materialist justifications of the war in Donbas, before suggesting a plausible alternative. It is contended that the region's economic and material circumstances in the early 2010s had generated neither necessary nor sufficient conditions for a locally rooted, internally driven armed conflict. The role of the Kremlin's covert military intervention, aided by its local proxies, remains paramount, necessary and sufficient for the continuation of hostilities. In the meantime, Ukraine's war-ravaged eastern provinces face further depopulation, economic decline and erosion of development.
From culture to structure: economic determinism and the causes of the war in eastern Ukraine The bulk of explanations accounting for the causes of the Russo-Ukrainian conflict are based on the foreign policy or geopolitics-centred considerations mentioned above. The rest of the literature focuses heavily on what Portnov (2015) describes as intellectually reductionist and essentialist 'cultural identity' and 'history' interpretations (Antoniuk & Prohnimak 2015;Kuzio 2015;Kulyk 2019). Despite some earlier studies highlighting Ukraine's remarkable regional stability (Sasse 2007), it is not hard to find academic publications today describing the country as the doomed 'land of irreconcilable differences', where 'identity is a given that cannot be changed' (Molchanov 2016, p. 204) and 'history is irreversible' (Sotiriou 2016). Some authors have gone as far as to follow the official Russian government line, depicting the 'Ukraine crisis' as a 'clash of civilisations' affair, a civil war between Russophone and Russian Orthodox Ukraine and Galicia-based 'Catholic nationalist extremists' (Loshkariov & Sushentsov 2016;Petro 2016). 7 Another-ostensibly more promising-strand of this literature claims that the armed conflict in eastern Ukraine was economically determined, one way or another. As summarised by Buckholz (2019), the economic determinist hypothesis suggests that the inhabitants of Donbas were particularly inclined towards separatism, either because of the region's economic exposure to Russia, or due to the region's economic significance and perceptions of discriminatory wealth redistribution within Ukraine. A common refrain running through economic determinism-driven arguments is that the conflict is a rational response based on the material interests of local industrial workers, vulnerable to economic shocks emanating from closer integration with the EU and fighting to 7 See also Kuzio (2019). preserve historical trading ties to the Russian market (Giuliano 2015;Zhukov 2016). At the same time, it has been argued that the war in Donbas was caused by a local push for fiscal autonomy or even self-determination from Ukraine, aimed at exploiting the industrial capabilities of its 'cash-generating cities' (Sotiriou 2016, pp. 61-4) and thereby establishing an economic base for the two breakaway 'people's republics' (Matveeva 2016, p. 40).
Indeed, as argued by Zimmer (2004Zimmer ( , 2007 and Osipian (2015), the region's political identity was based on the founding myth of Donbas as the most developed region of the country ('Donbas feeds the USSR'), and the associated notion that the nation's entire economy (first Soviet, later Ukrainian) rested on Donbas (Horbulin 2015, pp. 117, 273;Libanova 2015, pp. 7, 25, 33-4;Maiorova 2017, p. 28). Consequently, 'Donbas feeds Ukraine' was a slogan later used by the former President Yanukovych's Party of Regions (Partiia rehioniv) in the run up to the tumultuous 2004 presidential elections (Wilson 2016, p. 640). The 'affluent independent Donbas' storyline has also played a vital part in the Kremlin's Ukraine civil war narrative. The subsequent role of President Putin's aides, Glaz'ev and Vladislav Surkov,8 in directly instigating the 'popular uprisings' across Ukraine's eastern and southern regions has been well documented (Toler & Haring 2017). For example, two months into the conflict, Glaz'ev insisted: The Donets Basin, commonly called [the] Donbas, has all the possibilities to create its own autonomous financial system and Kiev understands this. … Financial capabilities of [the] Donbas will allow the region to avoid a sharp decline in living standards that the rest of Ukraine will see.
[The] Donbas can potentially increase its living standards by half, and become the most prosperous region of the new Ukraine. 9 Historical materialism and homo economicus: digging out the real basis of social conflict?
The emphasis given to rational human thinking based on real material interests is, certainly, a centuries-old idea that does not rest on an altered reality constructed by 'fake news'. The assertion that social phenomena have their roots in the economic structure of society-in 'social relations of production'-lies at the heart of Marxist historical materialism. In the preface of his 1859 treatise, A Contribution to the Critique of Political Economy, Karl Marx (1971) famously posited that all public life ('the superstructure') was determined by the prevailing mode of production ('the base'): 'The mode of production of material life conditions the general process of social, political and intellectual life. It is not the consciousness of men that determines their existence, but their social existence that determines their consciousness'. Further elaborations of the official communist dogma can be found in Joseph Stalin's pamphlet on Dialectical and Historical Materialism: Whatever is the being of a society, whatever are the conditions of material life of a society, such are the ideas, theories, political views and political institutions of that society. … Or, to put it more crudely, whatever is man's manner of life such is his manner of thought. … Hence, the clue to the study of the laws of history of society must not be sought in men's minds, in the views and ideas of society, but in the mode of production practiced by society in any given historical period; it must be sought in the economic life of society. (Stalin 1951, pp. 25, 36-7) Hence, the economic determinist arguments put forward by Giuliano (2015) and Zhukov (2016), seeking to explain the war in eastern Ukraine through the material interests of its industrial workers rationally uprising to preserve their jobs, have a long-standing theoretical pedigree, stretching back to the writings of Marx and Engels (Cornforth 1968;Ruben 1979). At first, it may appear that historical materialist explanations have so far permeated all social sciences and are very often invoked by non-Marxists or even anticommunists (Wood 2005). However, the vision of 'rational economic man' or homo economicus is not unique to Marxism. In its basic version, the rational economic man, in the words of Brown et al. (2018), is 'an imaginary person fought over by rival social scientists', ranging from laissez-faire economists and international relations theorists to child psychologists. For example, this is how homo economicus is defined in neo-classical economic theory: In its ideal type case the agent has complete, fully ordered preferences (defined over the domain of the consequences of his or her feasible actions), perfect information and all the necessary computing power. After deliberation, he or she chooses the action that satisfies their preferences better (or at least no worse) than any other. No questions are raised about the source or worth of preferences, reason focuses on the efficient selection of the means to given ends … . Such a social theory is individualist and contractarian, with a pedigree that includes Hobbes's Leviathan and Benthamite utilitarianism. The satisfaction of individual preference, aided by felicific calculation, is what makes the social world go round. Social relations become instrumental, in the sense that they embody exchanges in the service of individual preferences. (Hargreaves-Heap & Clark 2008, pp. 1-2) The 'rationality postulate' of neo-classical economics is more complex and sophisticated than that of the historical materialists, but the basic vision of self-interested individuals-who are rational in the sense that they are moved to maximise their most preferred outcome given the constraints upon choice-remains (Blaug 1992, pp. 229-34;Hargreaves-Heap & Clark 2008). In the following sections, the essay uncovers the real basis of social life in the prewar Donbas in order to establish whether the motives of the locals-their alleged preference for an armed rebellion-could have been economically rational and (pre)determined; that is, whether local economic agents could have been seeking to maximise their utility, in a broad sense, or their personal wealth, narrowly defined, through armed conflict.

Human geography of Donbas under post-communism
Donbas is an old mining region of eastern Ukraine, covering the area between the valleys of the Donets and lower Dnieper rivers. In terms of public administration, the Ukrainian Donbas consists of two provinces, centred on the cities of Donetsk, founded in 1869 by Welsh ironmaster John Hughes, and Luhansk, founded in 1795 by Scottish industrialist Charles Gascoigne. The Donets coal-field itself stretches from Ukraine's Dnipropetrovska Oblast' (province) in the west (western Donbas), through the central Donbas, and crosses the Ukraine-Russia border in the east into Russia's Rostovskaya Oblast' (see Figure 1). The Russian or eastern Donbas consists of four districts (raiony) of Rostov province-Kamenskii, Belokalitvenskii, Krasnosulinskii and Oktyabr'skii, with a total population of 304,784 in 2010. 10 The region has played a defining role in Ukraine's post-communist transformation (Swain 2007). Following the collapse of the Soviet Union, Donbas has suffered a devastating economic transition, plunging-alongside Ukraine-to the bottom of the world's human survival and development indicators. Overall, Ukraine's post-Soviet development was characterised by the deepest and longest economic depression experienced by any of the post-communist transition economies not affected by war, resulting in the loss of 60% of the gross domestic product (GDP) between 1990 and 1999 (Mykhnenko 2011). The impact of national economic decline on Donbas's economic fortunes was profound, accelerating the region's long-term demographic crisis. Under post-communism, Ukraine's crude death rate of 15.43 deaths per thousand people fell behind that of the war-ravaged sub-Saharan Africa (15.03), while the country's total fertility and population growth rates in the mid-2000s reached the historically lowest level ever recorded in the world. In terms of natural population change, during the past two decades Ukraine has been amongst the fastest shrinking countries on earth, well ahead of Bulgaria, Latvia, Russia and Hungary, the world's leaders in natural population decline (Mykhnenko et al. 2012a). Overall, between 1991 and 2014, Ukraine's total population fell by at least 6.8 million people. The population of Donbas on 1 January 2014 was officially estimated at 6,583,400, constituting 14.5% of Ukraine's total population, with Donetsk province accounting for 4,343,900 people (9.6% of Ukraine's total) and Luhansk province for 2,239,500 (or 4.9%; see Appendix Table A1). By the end of 2013, Donetsk province had lost 19% of its inhabitants from its peak in 1992, whilst population decline in Luhansk province was even steeper, over 22%. Thus, before the outbreak of hostilities in early 2014, the population of Donbas had already decreased by 1.7 million overall, shrinking almost twice as fast as Ukraine's on average. 11 Donbas's downward demographic and economic trajectories under post-communism have had a detrimental impact on the region's labour force. The Soviet economy of centralised planning was a full employment economy, based around the mobilisation of all labour recourses available. The continuous growth of the workforce in Donbas (and Ukraine as a whole) stopped in 1985, with the start of Gorbachev's perestroika reforms. Consequently, between 1985 and 2013, the workforce in Donbas shrank by one fifth, from 4,034,380 to 3,217,000 in total. By comparison, the total workforce in the country declined by 14%. In addition to natural population decline, fewer people were in work: under postcommunism, Luhansk province reported a nine percentage point (pp) fall in the number of   The quality of the regional workforce has historically been low, geared towards hard manual labour, with the relative share of knowledge workers in Donbas being just over half of Ukraine's average. Due to an ageing workforce and early retirement amongst coal miners, by the end of 2013, a third of the region's total population-2,122,694 peoplewere of pension age and retired. The total annual pension outlays in Donbas topped US $5.5 billion in current prices. These mandatory pension payments amounted to an unsustainable share of 20.0% of the aggregate regional GDP: ranging between 17.6% of GDP in Donetsk and 26.9% in Luhansk provinces (see Appendix Table A1). For comparison, public pension spending across high-income industrially advanced economies stands on average at 8% of GDP. 13 Donbas economy before the war: Ukraine's dynamo or rust belt?
Following the depression of the 1990s, the Ukrainian economy grew at 8.7% per annum between 1999 and 2014. Over that period, Donbas's economy-the nation's second largest after Kyiv-was expanding at a much slower rate (6.3% p.a.), with Donetsk province registering 6.9% of annual economic growth, and Luhansk province trailing at 4.8%. To situate Donbas's regional economic performance within the Ukrainian spatial economy, the real annual growth rates of provincial GDP in the 1999-2013 period, their actual sizes in 2013 (valued in Ukrainian hryvnia, UAH) and the starting point GDP per capita in 1999 are displayed in Figure 2. The data show that, in the 14 years that preceded the war, Donetsk province was losing ground to all the other affluent regional economies, except for Zaporizhzhia and Poltava. In turn, Luhansk province was growing at the fourth slowest pace amongst 27 regions, falling further behind all of Ukraine's poorest provincial economies except Sumy and Chernihiv. A national prosperity dynamo the post-Soviet Donbas clearly was not (Gentile & Marcińczak 2012).
The structure of the regional economy and its relative weight in national production have been changing over time, with the services sector driving most economic growth in the region. Agriculture, which accounted for 10.1% of the value of Donetsk provincial output in 1988, declined to 4.2% in 2013, making the territory firmly dependent on food imports. The share of mining and manufacturing in total regional output have sharply declined too, from 70.6% of Donetsk provincial output in 1988 to 34.2% in 2013. Deindustrialisation affected Luhansk as well, with its combined share of mining and manufacturing activities declining to 39.9% of the provincial economy by the end of 2013. The total level of economic activity in the region as a whole stood at UAH220.03 billion (US$27.52 billion at current prices), with Donetsk province accounting for three quarters of Donbas's aggregate GDP. The region's economic importance has continued in the traditional spheres of mining and quarrying, accounting for 29.4% of Ukraine's output, and also in manufacturing (21.0% of Ukraine's total output) and in energy supply (20.3%; see Appendix  Table A1). Nevertheless, the overall significance of the Donbas economy for post-Soviet Ukraine has steadily declined, from a 17.3% share of Ukraine's GDP in 1996 to 14.5% in 2013. Luhansk province's relative decline during this period was particularly steep, from a 5% share to just 3.6% in 2013 (see Figure 3).
The real economy of the prewar Donbas: the firm base for a prosperous future?
Following the post-communist transition depression, economic recovery in the region from the late 1990s was largely driven by exports of goods and services, which peaked in 2011 at US$24.66 billion, accounting for a remarkable 30% share of Ukraine's total exports, twice the region's population share. However, with the slowdown of the global economy in the 2010s, Donbas's exports dropped sharply to US$16.75 billion in 2013, shrinking to 21.4% of Ukraine's total, though remaining the nation's top exporting region overall. That year, Donetsk province exported US$12.41 billion worth of goods and US$640 million worth of services; Luhansk province's merchandise exports stood at US$3.55 billion, with the export of services accounting for US$157 million. In terms of trade composition, ferrous metals (steel), fuel (coal) and railway locomotives comprised three-quarters of the region's commodity exports, whilst transport (rail and sea) made up the bulk of the exported services. Russia and the other Eurasian Economic Union countries accounted for 32% of Donbas's merchandise exports in total, being the most important market for Luhansk province (41.4% of exports), and its heavy engineering and railway transport firms. The EU took 28.2% of Donbas's exports, whilst the largest share (39.8%) went elsewhere. Indeed, the bulk of Donetsk province's exports (44.8%) was destined for the markets outside of Europe and the former USSR (see Appendix Table A1). Despite the region's foreign trade success, Donbas has failed to attract any significant amount of inward foreign direct investment (FDI). By the end of 2013, Donbas's share of national FDI stock had not even reached 8%, and the per capita FDI stock amounted to just US$808 in Donetsk and US$394 in Luhansk provinces, respectively. In terms of domestic capital formation, the region was more successful: between 2000 and 2013, Donetsk province reported a 37.4pp increase in capital investment, compared to Ukraine's 14.6pp increase. Since 2000, Luhansk province has more than doubled its capital stock, investing 8% per annum into domestic capital, on average. The export boom has turned Donetsk into a major business and asset management hub, with two dozen of Ukraine's top 200 companies maintaining their headquarters in the provincial capital, including Metinvest, DTEK, ISD, DonetskSteel and Donetsk Railways. The surplus of local company profits from steel exports and other semi-processed manufactures was routinely channelled to offshore tax havens: by 2014, 83% of Ukraine's total outwards FDI originated from Donbas (see Appendix Table A1). Some of the foreign cash windfall enjoyed by the largest Donbas firms has also found its way into the pockets of ordinary people. Donbas's annual gross personal income in 2013 reached UAH35,642 per capita (US$4,459 or €3,359 in current market prices); Donbas's income level was 6% higher than the national average of UAH33,668. That was despite the fact that the region's output per head in 2013 was Ukraine's average: 14.5% of the country's total population generated 14.5% of its GDP (see Figure 3). The residents of Donetsk province were 12% (US$500/€400 a year) better off than the average Ukrainian, with gross annual personal income in the province reaching UAH37,680 per head (US$4,714/€3,551). The partial explanation for this was that salaries and wages in Donbas have been amongst the highest in the country: in 2013, an average employee in Donetsk province earned the country's second highest wage of UAH3,755 per month (US$470/ €366), whilst his or her counterpart in Luhansk province was paid UAH3,337 a month (US$417/€325), the fourth highest wage in Ukraine. However, primarily as the result of scarce employment opportunities, the standard of living in Luhansk province was rather mediocre, standing in 2013 at UAH31,692 per capita a year (US$3,965/€2,986), that is, 6% lower than the national average. In terms of gross disposable income, Donetsk province remained Ukraine's second most prosperous province throughout the 2000s (after Kyiv), whilst Luhansk province chronically lagged behind.
The war in eastern Ukraine: a donor region's revolt?
One of the two economic determinism-based hypotheses tested in this study postulates the war in Donbas to be the result of a separatist uprising by a wealthy region, trying to further its material interests and living standards by abandoning the 'net contributor' role and seceding from its poor, 'subsidised' neighbours. Yet in the simplest terms, by January 2014, Donbas's position overall was fiscally neutral. Given the region's output per head equalled Ukraine's average, Donbas-as a whole-was neither a net contributor to national inter-budgetary transfers nor a net recipient of equalisation subsidies from Kyiv. Taking into account the intra-regional development gap discussed above, the more affluent Donetsk province had probably continued to play its 'donor region' role in the Ukraine state budget in the run up to 2014 (Novikova 2016, Tables 4.4-4.5). Nevertheless, even there, the budgetary donor-recipient positions were not entirely straightforward. The fiscal transfer analysis of the Greater Donetsk conurbation reveals that, in 2010, Donetsk City Council transferred 39.4% of its own revenue 'surplus' to Ukraine's consolidated state budget, whilst the neighbouring municipality of Makiivka received up to 27.1% of its budget expenditure in the form of equalisation subsidies from Kyiv (Mykhnenko et al. 2012b). The politically explosive issue of inter-budgetary fiscal transfers was further complicated by Donbas's ever rising and nationally funded pension burden, and by the amount of explicit and hidden subsidies received by the local coal mines, steel mills, coking plants, chemical factories and coal-fired power generators in the form of generous state orders, output price controls and paltry environmental pollution penalties (van Zon 2007;Iwański 2015). Indeed, Donetsk province governor Serhii Taruta, addressing the administration staff on 27 March 2014, claimed the province was subsidised by the centre, living well beyond its means: In the past there was a myth that Ukraine fed the entire Soviet Union. But when the Soviet Union collapsed, we learnt that was untrue. Today there is a myth that Donbas feeds Ukraine. Really, if one is to ask, I think that probably each and every Donbas inhabitant would say that he feeds the whole country. And yet, if we look at the statistics, we see that Donetsk province is a subsidised region and the state budget supplements our income. 14 Furthermore, Luhansk province has been a constant drain on the public purse since the records began in 1996 (Mykhnenko & Swain 2010, pp. 154-55). Thus, the 'donorregion's revolt' argument does not provide a very compelling reading, if one is concerned with a rational self-interested (or selfish) response based on perfectly understood material interests. Yet the accounting peculiarities of Ukraine's public finance formation did not feature prominently amid the political intensity of early 2014. As a local witness observed, it was impossible to calm the pro-Russian separatist agitation: These people are in a state of altered reality and painful consciousness and aggressively defend their image of the world from any attempt at intervention. [The] 'image' is established by the Russian [television] channels, local press and the Party of Regions, which has long worked on churning out the stereotypes 'we'll die without Russia', 'decaying Europe', [the] 'Donbas feeds Ukraine' and the like. The most aggressive part of the population operates on pseudo facts that cannot be challenged, as they are unprovable. Professional psychiatrists characterise such behaviour as mass psychopathology. (Todorov 2015, p. 10) The war in eastern Ukraine: a 'rust belt' rising?
The second economic determinism-based hypothesis investigated here suggests that the war in Donbas is a local rebellion rooted in the old industrial interests of Russia-bound exporters. To pursue the 'rust belt rising' hypothesis, Zhukov (2016) used micro-level data on violence during the first 12 months of the Russo-Ukrainian war across 3,037 population settlements in Donbas and its correlation (through Bayesian Model Averaging) with the following set of explanatory variables: ethno-linguistic population ('proportion Russian-speaking'), economic structure ('proportion of the local labour force employed in machine-building/mining/metals industries'), characteristics of the local military geography ('population density', 'elevation', 'forest cover', 'distance to the nearest road', 'distance to the Russian border') and the prewar political loyalties of the local population ('proportion of the popular vote received by Viktor Yanukovych in the 2010 presidential election'). Zhukov (2016) claims a municipality's prewar employment mix to be the strongest and most robust predictor of militant activity. His study concludes that where the 'opportunity costs of insurgency' were low, as in machine- building company towns particularly exposed to trade shocks with Russia, the risk of armed conflict was greater overall and separatist violence was more frequent; the Ukrainian government had lost control of these areas earlier, and militants were able to hold territory gained for longer. By contrast, where the opportunity cost of rebellion was higher, as in cities dominated by Ukraine's relatively competitive and globally oriented steel industry, militants had difficulty establishing and maintaining control. Noticeably, neither local differences in ethnic-linguistic composition nor Donbas's prewar electoral allegiance to ex-president Yanukovych have turned out to be explanatory factors (Mykhnenko 2009).
Historically, Shulman (1999) highlighted an extremely high degree of alienation and foreign policy dissent between eastern and western Ukrainian business elites, fuelled by an imbalance in the sources-perceived or actual-of beneficial foreign economic ties. Shulman's conclusion, reached five years prior to the 2004 Orange revolution, was that Ukraine's asymmetrical international integration, with western regions moving closer to the EU and eastern regions clinging to their ties with Russia, had significantly weakened the country's unity and cohesion. A decade later, Mykhnenko and Swain (2010) portrayed the contemporary Ukrainian spatial economy as comprising three historically rooted and relatively geographically bounded regional economies that were functionally integrated yet externally orientated: each exhibited a degree of sector and/or commodity chain specialisation and was either dependent on exporting commodities and manufactures or on importing financial capital.
Whilst acknowledging the dangers of Ukraine's diverging spatial economy, one ought to distinguish between what Wilson (2016, p. 641) called 'baseline separatism', existing in many regions across the world, and a much more explicitly anti-Ukrainian economic interests-based 'master cleavage'. Most importantly, as previous studies testify (Mykhnenko & Swain 2010;Mykhnenko 2011), the areas over which the Ukrainian government lost control in 2014-2015 had never possessed a distinctively different industrial structure, which could have made them especially susceptible to an armed separatist rebellion in comparison with the neighbouring areas of eastern Ukraine. For some reason, Zhukov (2016) did not extend his municipality-level analysis to other heavily industrialised, coal-mining, steelmaking and machine-building east Ukrainian towns in neighbouring Kharkiv, Zaporizhzhia and Dnipropetrovsk provinces. Others did offer such a comparison: working independently, Portnov (2015), Buckholz (2019), Stebelsky (2018) and Matsuzato (2018) compared the success or failure of pro-Russian militant movements across eastern Ukraine, contrasting the experiences in Kharkiv, Dnipro, Kramatorsk and Mariupol, with that of Donetsk and Luhansk. They concluded the difference was best explained through the preferences and actions of local political elites, in addition to the obvious factors of military geography. This last factor was foregrounded by Zhukov (2016): he found that armed insurgents in Donbas were able to establish control more quickly in densely populated localities that were accessible by road, situated at relatively high elevation and close to the Russian border, thus concluding that military geography was strongly and consistently predictive of armed conflict.

Socially constructing the domestic causes of the war in Donbas
A historical materialist reading of the secessionist 'Donbas feeds ungrateful Ukraine' argument would render it false, as this hypothesis is not based on a real economic conflict between Donbas's social forces of industrial production and the pan-Ukrainian social relations of inter-budgetary fiscal transfers, geared towards spatial equalisation of income and support for poorer municipalities, including many shrinking cities across Donbas itself. It is contended that the economic explanation of the events in eastern Ukraine in early 2014 based on the rationality postulate cannot advance one's understanding of the war. The analysis provided so far questions the existence of necessary local conditions for an economically determined cause of the war in Donbas, be it of the 'donor region's revolt' type, or of the 'rust belt rising' variety. Furthermore, the local economic determinants were also not sufficient for an armed conflict to occur. It is important to highlight that even in peacetime decision-making circumstances, the application of the economic rationality concept is very problematic. As argued by a leading neo-classical economic theorist of the post-World War II era, rationality not only 'seems capable of leading to conclusions flatly contrary to observation' but also imposes 'the extremely severe strain on information-gathering and computing abilities. Behaviour of this kind is incompatible with the limits of the human being, even augmented with artificial aids (which, so far, seem to have had a trivial effect on productivity and the efficiency of decision making)' (Arrow 1986, pp. S397-98).
The war in Donbas was principally the end result of foreign instigation, intervention and invasion, aided and abetted by collaborating elements of the Yanukovych regime (Kudelia 2014; Malyarenko & Wolff 2018). Whilst 'economic rational man' was conspicuously absent from the streets of Donetsk and Luhansk in the tumultuous spring of 2014, pro-Russian agitators were not. As mentioned above, the deep-rooted myth that 'Donbas feeds the whole country' played a substantial part in pro-Russian secessionist propaganda at the time, convincing an increasingly militant minority. Myths often provide the framing for collective action. To investigate this possibility in the case of the Donbas conflict, classical Marxism or neo-classical economics should be put aside in favour of social constructionism (constructivism). This epistemological position holds that meaning-making, knowledge and understandings of the nature of reality are not the result of direct observation or perception but are constructed through social discourse and based on experience generated within human relationships (Berger & Luckmann 1967;Merz 2005). According to the basic tenet of social constructionism, 'what one takes to be true as opposed to false, objective as opposed to subjective, scientific as opposed to mythological, rational as opposed to irrational, moral as opposed to immoral is brought into being through historically and culturally situated social processes' rather than through the individual mind of 'the rational, self-directing, morally centred, and knowledgeable agent of action' (Gergen & Gergen 2008, p. 817).

Consequences of the war: human cost and economic losses for the Ukrainian Donbas
In the current context, the quest for causality has already moved beyond pure academic research and entered the sphere of the international law, with the aim of investigating crimes against humanity, establishing personal responsibility and delivering justice, including war reparations (Åslund 2018). Between 14 April 2014 and 15 February 2019, the war between the combined forces of pro-Russian separatists and Russian regular army units operating in Donbas against their Ukrainian adversaries resulted in at least 40,000-43,000 conflict-related casualties, as reported by the Office of the United Nations High Commissioner for Human Rights (OHCHR 2019, p. 6). According to this conservative estimate, the war has claimed 12,800-13,000 lives, with at least 3,321 civilians and over 9,500 combatants killed, and 27,500-30,000 people injured, including 7,000-9,000 civilians and 21,000-24,000 combatants. Around 3.3 million people have fled their homes, with forced displacement emptying cities, towns and villages of Donbas. According to the United Nations Office for the Coordination of Humanitarian Affairs (UNOCHA 2015), as of 2015 only 180,000 people remained in the occupied city of Horlivka, compared to 276,000 before the war. Only 8,000 people remained in Dokuchaievsk, compared to 30,000 residing in the town before the outbreak of hostilities. In Debaltseve, the resident population declined from over 25,000 to fewer than 7,000 inhabitants. The frontline town of Krasnohorivka was estimated to have a population of under 5,000 people from 16,000 residents before the war.
On 4 July 2016, the number of officially registered internally displaced people in Ukraine topped 1,790,267 (see Figure 4); the number of Ukrainian refugees in Russia and Belarus had reached more than 1.5 million earlier that year. 15 As a result, the situation in Ukraine has turned into Europe's largest humanitarian catastrophe since the Yugoslav wars of the 1990s, with Ukrainians accounting for 85% of all internally displaced people in Europe in 2,000,000 million remained in the non-government-controlled areas, officially known as the 'separate districts of Donetsk and Luhansk provinces' (okremi raiony Donetskoi i Luhanskoi oblastei-ORDLO); around 600,000 were caught in the so-called grey zone, a five kilometre strip on each side of the 457-kilometre frontline, living mostly underground in basements and improvised bunkers, with no water, heating or electricity, and no access to healthcare, shops, jobs or social services. There were constant reports of severe malnutrition and lack of basic foodstuffs (UNOCHA 2017).
The Russo-Ukrainian war has been particularly lethal on account of the range of weapons used, including old Soviet (T-64s) and new Russian (modernised T-72s and T-90s) models of main battle tanks, mechanised infantry, self-propelled and towed artillery, multiple-launch rocket systems, mortars, and an array of anti-aircraft and electronic warfare weapons (Echevarria 2016). The Ukrainian Air Force also deployed attack aircraft and fighter jets, whilst their opponents used very sophisticated anti-aircraft weaponry. 16 Ground combat formations on both sides have been vulnerable to surveillance conducted by remotely piloted aerial vehicles (drones), helping direct devastating artillery and rocket fire onto them within a few minutes. Reportedly, artillery has accounted for 85% of all casualties in this conflict, well above the World War II record (Echevarria 2016, p. 28). As most Russian and Ukrainian artillery systems deliver area-fires rather than precision-fires, with many engagements occurring at long ranges (10-15 kilometres), in densely populated urban areas, the damage caused by indiscriminate shelling has been immense (Case & Anders 2016;Echevarria 2016).
Since the outbreak of hostilities, a number of research centres and think-tanks have been keeping track of the physical and material damage and economic losses (Horbulin 2015;Libanova 2015;Maiorova 2017;Åslund 2018). According to these accounts, in addition to tens of thousands of deaths and injuries, and millions of people seeking refuge away from the theatre of war, the impact of the armed conflict on Donbas has been truly devastating (Buckley et al. 2018). Tens of thousands of enterprises have been closed, looted, dismantled and smuggled to Russia for scrap; countless homes, hospitals, roads, bridges, railways and pipelines have been destroyed; coalmines have been flooded; farmland has been littered with unexploded ordnance and land mines. 17 The Ukrainian Donbas has become the third most mine-contaminated stretch of land in the world, covering around 16,000 square kilometres (UNOCHA 2017 2015, before gradually recovering to US$7.39 billion in 2017. 18 (For the provincial breakdown, see Figures 5-8.) The war has resulted in rapid and severe deindustrialisation in the region, with the industrial output in Luhansk province dropping by 84.3% from its 2013 level and halving in Donetsk province. The previously lucrative foreign trade has collapsed, with Donetsk province's exports declining by 61.7% and Luhansk's exports virtually disappearing. Using Åslund's (2018) suggested formula for evaluating permanent economic losses suffered by Ukraine as the result of the war, the total permanent physical asset value loss in Donbas approximates US$84.649 billion (in 2013 prices), calculated as US$21.16 billion of lost gross domestic product multiplied by four (for details, see Åslund 2018).

The Ukrainian Donbas: de facto partitioning
Nominally, Donbas's non-government-controlled areas (ORDLO) cover around 38% of the region's total territory and up to 45% of its prewar population, including four out of the region's five largest cities: Donetsk, Luhansk, Makiivka and Horlivka. Hence, it is vital to assess the level of economic activity that has been maintained or restored there. Until March 2017, Ukrainian official statistics had partially registered the economic activity taking place in ORDLO, as dozens of commercial establishments located there re-registered on Ukraine government-controlled territory. Some of those continued operations, paying taxes, custom and excise duties to the Ukrainian government, in addition to bribery and extortion money demanded by ORDLO militants. 19 A significant blow to the intra-regional economic ties was dealt on 1 March 2017, when ORDLO militants began illegally seizing the remaining functioning economic assets owned by non-Russian firms and interests, including 43 large industrial enterprises, steelworks and collieries mostly belonging to Donetsk billionaire Rinat Akhmetov: SCM, 20 MetInvest, 21 Ukrtelecom 22 and DTEK. 23 In the meantime, a group of aggrieved Ukrainian Donbas war veterans imposed a full blockade on trade between the government-and non-government-controlled areas, an action formally legalised by the Ukrainian National Security and Defence Council resolution enacted by Presidential Decree No. 62/2017 on 15 March 2017. 24 The eventual confiscation of Ukrainian industrial assets has accelerated deindustrialisation in the region: industrial output in Luhansk province fell further by 31% in 2017, followed by a 17pp decline in 2018 (see Figure 6). After a sharp decline in 2014-2016, foreign trade exports from Luhansk province declined further in 2017 by 44.4%, followed by another 11.5pp drop in 2018. In Donetsk province, industrial output declined in 2017 by 10.9%, recovering by a modest 2.6% in 2018 and reaching 43.6% of the 2013 level (see Figure  6). During the period 2014-2018, aggregate exports from the region as a whole had declined by 68.8% (see Figure 7). Even if the external trade figures publicised by the self-proclaimed authorities in ORDLO are taken at face value, the aggregate Donbas total exports in 2018 still amount to just one-third of the 2013 level (see Figure 7).
Unemployment in Donbas has reached unprecedented levels: 15.6% of the labour force (working age) in the government-controlled areas of Donetsk province and 18.3% in Luhansk province were seeking employment in early 2017. 25  'Ukrtelecom is Forced to Cut Services to its Subscribers in the Non-government Controlled Areas of Donetsk Oblast due to Illegal Actions Undertaken at its Facilities in the Area', System Capital Management, 1 March 2017, available at: https://www.scm.com.ua/en/news/ukrtelecom-is-forced-to-cutservices-to-its-subscribers-in-the-non-government-controlled-areas-of-donetsk-oblast-due-to-illegal-actions-u ndertaken-at-its-facilities-in-the-area, accessed 14 June 2019.   highest levels of unemployment since 2014. 26 Massive unemployment has become prevalent in ORDLO, though no estimates have ever been generated by the de facto authorities in the nongovernment-controlled areas. What is known is that in 2017 there were 351,500 people formally employed in various enterprises, institutions and organisations across Donetsk's section of ORDLO, 27 with 49,600 people officially registered there as unemployed, 28 thus, comprising a formal labour force of 401,100 people. These figures suggest that the official registered unemployment rate in ORDLO in 2017 stood at 12.4%, as a share of jobseekers amongst the economically active. For comparison, in 2017, there were 13,500 formally registered jobseekers in the government-controlled areas of Donetsk province, 29 with 466,200 people officially employed there in enterprises, institutions and organisations, thus, comprising a formal labour force of 480,700. Therefore, the official registered rate of unemployment in Donetsk province that year amounted to 2.8%-a figure 4.4 times lower than ORDLO's. Extrapolating the ratio between the labour-force survey-based and formally registered unemployment rates in Donetsk province in 2017 (15.6% compared to 2.8% respectively) onto ORDLO, the level of real unemployment in non-government-controlled areas can be estimated to be as high as 69.1% of the working-age workforce. The economic situation in the non-government-controlled areas of Donbas, thus, appears especially dire. Reportedly, the disruption of business in ORDLO has had the most severe impact on the local economy, with senior management fleeing, capital and trade flows abruptly halted, and all existing production and supply chain logistics dismantled. 30 The majority of the companies seized by the militants in 2017 stopped production, putting workers on indefinite 'unpaid leave'. With the loss of the lucrative Ukrainian market, the ORDLO-controlled coal mines have faced closures, followed by abandonment of flooded underground excavations, worsening the humanitarian and environmental situation even further (Sokolova 2019). Food insecurity has doubled in ORDLO between 2016 and 2017, with 1.2 million people seeking food aid from charitable organisations (UNOCHA 2017). Unsurprisingly, ORDLO authorities do not openly publish any consistent or reliable statistical data, which would allow some monitoring of their performance. 31  to assess wartime economic circumstances in Donbas objectively, Coupé (2016) and Coupé and Obrizan (2016) used changes in night-time light intensity, measured by satellites. On the basis of the data for half a dozen cities in the region, they estimated the extent to which the war had destroyed the economy in Donbas, and whether there had been any recovery after the Minsk II truce agreement, reached in February 2015. Consequently, they observed a decline in economic activity in Donbas from March 2014 to March 2016 of 50-70% of the prewar level across the larger cities (Donetsk and Luhansk), with an even more staggering 90% fall across some smaller cities (Debaltseve, Ilovaisk).
The data gathered during this research, including ad hoc press releases by ORDLO militants, 32 provide a more comprehensive and up-to-date account: the aggregate industrial output of Donetsk province, including both government-controlled and occupied areas, in 2017 stood at 50.3% of the prewar level. In Luhansk province, in 2016, the last year for which the ORDLO data are available, the aggregate industrial output was no higher than 35.6% of the prewar level (see Figure 6). In 2017, even after the confiscation of all productive and commercial assets by the ORDLO militants, industrial output in the non-government-controlled areas of Donetsk province reportedly reached R146.4 billion (UAH66.81 billion/US$2.51 billion), whilst in the governmentcontrolled areas of the province, industrial production stood at UAH258.65 billion (US $9.72 billion). Thus, in relative terms, 79.5% of the aggregate industrial output of Donetsk province (US$12.23 billion in total, within the prewar borders) was generated on government-controlled territory, and only 20.5% was produced outwith. Notably, before the war, the currently occupied area of Donetsk province accounted for around 47% of total production. In Luhansk province, before the war, the currently occupied territory accounted for around about 60% of the total provincial industrial output. By contrast, in 2016, ORDLO territory accounted for only 29.8% of the aggregate provincial industrial output (R40.984 billion in absolute terms, or UAH15.74 billion/US $615 million), compared with 70.2% produced in the government-controlled area (UAH37.09 billion/US$1.45 billion); the total industrial output for both parts of the province amounted to UAH52.84 billion (US$2.06 billion).
The collapse in living standards in Donbas is best seen through a comparison of average wages before and after the war, on both sides of the frontline. Just before the outbreak of hostilities, in January 2014, the average wages in Donetsk and Luhansk provinces were UAH3,674 (US$460) and UAH3,295 (US$412), respectively. In 2017, the average wage in the government-controlled areas of Donetsk province was UAH7,764 per month (US$292, or 63% of the prewar level). At the same time, in non-government-controlled areas of the province, the average monthly wage was just over half (59.5%); this figure reportedly amounting to R10,130 (UAH4,619/US$174), and being 38% of the prewar level. In Luhansk province, the average monthly wage in ORDLO in December 2016 was R8,704 (UAH3,676/ $140, and 34% of the prewar level), or 61.2% of the average wage across the frontline 32 'Infografika-Infografika', Ministerstvo ekonomicheskogo razvitiya Donetskoi Narodnoi Respubliki, 2015-2019, available at: http://mer.govdnr.ru/index.php?option=com_phocagallery&view=category&id= 103:infografika&Itemid=693, accessed 14 June 2019.
In reality, the 'republics' are beginning to acquire the features of a military-bureaucratic regime in which military personnel and officials dominate society through coercion and the monopolization of the distribution of wealth. Meanwhile, the impoverished and intimidated population has no means of societal control and is in complete economic dependence on the authorities. (Kudelia 2016, p. 231) In turn, the separatist authorities are completely subordinate to Moscow. Officially, the Moscow-run civilian ORDLO governance structure was set up on 15 December 2014 via the Government of Russia's rasporyazhenie [Instruction] No. 2537-p, 35 amended on 8 October 2015 (Instruction No. 2008-pc) 36  The Inter-Ministerial Commission is located in Rostov-on-Don, Rostov province, bordering Donbas. Noticeably, the Russian Ministry of Defence is the only core federal ministry not taking part in the work of the civilian commission. According to the Ukraine military intelligence, a parallel quasi-governance structure has been set up to command and control the armed formations operating in ORDLO. The military command is also run from Rostovon-Don, from the headquarters of the Southern Military District, one of the four operational strategic commands of the Russian Armed Forces. 38 All the evidence suggests that ORDLO is entirely dependent on Russia for its survival, not just militarily but economically (Rácz & Moshes 2014). Well over 90% of all public spending administered by the two ORDLO administrations, including wages and salaries paid to teachers, nurses, doctors and municipal services personnel, and social assistance payments and payments to the elderly, regularly arrive from Moscow in cash by armoured trains (Röpcke 2016a(Röpcke , 2016b. Russia has also financed gas, fuel and oil, and distributed food and humanitarian assistance, on top of war-related spending. The ORDLO budget information published by the self-proclaimed 'Council of Ministers' only contains percentage growth figures against the previous year or percentage shares of various expenditures in the total budget (Polovyan & Lepa 2018, pp. 76-87); this does not enable an assessment of the actual amount of public expenditure in total or per head, or an evaluation of the share of internally generated revenue vis-à-vis total spending. Nevertheless, it is possible to estimate that in 2017, approximately 1,038,020 pensioners living in ORDLO were in receipt of a monthly 'local pension' payment, with the total annual bill running to approximately R52.6 billion (US$902.17 billion) (Polovyan & Lepa 2018, pp. 111-17). These pension outlays alone represent twice as heavy a burden on the local economy in comparison to 2013, when they already accounted for the unsustainable externally subsidised 20% share of the aggregate Donbas GDP (see Appendix Table A1).
The work of the ORDLO civilian shadow government is supplemented by final executive decisions made by Russia's deputy prime minister, Dmitrii Kozak and, ultimately, by Prime Minister Medvedev himself. For example, after the seizure of Ukrainian industrial enterprises in the spring of 2017, the confiscated assets were assembled by the Moscow-controlled militants into ZAO VneshTorgServis, a 'closed joint stock company' registered in the Russia-occupied South Ossetia and reportedly controlled by fugitive Ukrainian billionaire Serhii Kurchenko, who was closely linked to ex-president Yanukovych (Goncharova 2017;Dikhtiarenko 2018). Due to the shortage of raw materials (iron ore and coke), previously supplied from the government-controlled territories, the newly expropriated steel works in Yenakiieve, Alchevsk and Donetsk halted production immediately (Kushch 2018). On 4 May 2017, Medvedev instructed the Federal Agency for State Reserves to be provided with R10 billion from the National Reserve Fund for 'replenishing raw materials necessary for the ferrous metals industry' (Burmistrova & Tkachev 2017). According to Russian Business Consulting, this confidential Government of Russia Instruction No. 559-р of 28 March 2017 provided ORDLO with enough funds to purchase about two million tonnes of iron ore and coke, sufficient for the production of 1.5 million tonnes of steel (Burmistrova & Tkachev 2017). Notably, by the end of 2017, annual steel production in the non-government-controlled areas of Donetsk province reached 1.1 million tonnes (Polovyan & Lepa 2018, p. 14). The Russian government's handson management of the ORDLO economy is set to continue, accompanied by further deindustrialisation and de-development. In May 2017, Aleksandr Khodakovskiy, a deposed Donetsk warlord, accused the ORDLO militants of running the region down and turning it into another 'Abkhazia, herding goats and sheep'. 39 Indeed, 19 months later, Komsomol'skaya Pravda, a Russian tabloid daily, published a curious journalistic investigation into 'guest workers' from Donetsk and Luhansk, reduced to harvesting mandarin oranges in backbreaking conditions in Abkhazia. Numerous gastarbaitery from the occupied Donbas captured on tape and video-filmed by the Moscow-based journalist complained that life back home was even less rewarding than 'a night's stay and a bottle of lager a day' they were earning on the mandarin plantations (Tikhomirov 2018).
(see Figure 4). Around 88% of Ukraine's IDPs have found temporary residence in five neighbouring east Ukrainian provinces, including the government-controlled areas of Donetsk (42%) and Luhansk (13%), Kharkiv (10.9%), Zaporizhzhia (7.4%) and Dnipropetrovsk (4.3%), and in the capital city region (the city of Kyiv, hosting 8%, and Kyiv province, with 2.9%). Overall, the impact of Donbas's IDPs on the national economy has been fairly neutral, with the occasional success story of an IDP-run business start-up in Kyiv or Lviv running alongside news reports of pension payment offices, hospitals, shops and ATMs in Donetsk and Luhansk provinces being overrun by an influx of new arrivals from ORDLO. With a few notable exceptions, this study has not found enough convincing evidence to indicate a measurably positive economic effect of IDPs on their host communities, either in terms of an increase in consumer demand and retail turnover, or an accelerated rate of growth. If the war-damaged Donetsk and Luhansk provinces are included in the statistical analysis, the relationship between IDP inflows and regional economic growth is negative and statistically significant (see Figure 9). The main reason may lie in the demographic composition of Ukraine IDPs themselves, which is heavily skewed towards old-age pensioners (52%) and economically-inactive people with disabilities (4%). Working-age people account for just 29% of Ukraine's officially  registered IDP headcount. Although some IDP pensioners may have continued working fulltime, the absolute number of working-age IDPs has not risen beyond 500,000 (approximately 2% of Ukraine's labour force). 40 Faced with the challenge of postwar reconstruction, Ukraine IDPs would not be able to regenerate Donbas without a comprehensive large-scale public and international donorfunded reconstruction programme. The magnitude of this task looks particularly daunting, for the war, economic depression and chronic unemployment on the indistinct frontline zone and ORDLO areas have led to the deskilling of the remaining workforce. One remarkable exception, which needs further research, relates to the potentially transformational role played by IDPs in the local host economies of Kharkiv, Zaporizhzhia, Poltava and Odesa provinces, as well as in the Kyiv capital city region and Dnipropetrovsk province (see Figure 10, which excludes Donetsk and Luhansk). In particular, Kharkiv, Zaporizhzhia, Poltava and Odesa regions have been able to radically shift their economic gears, from being amongst Ukraine's growth laggards in the period 2011-2013 to becoming regional growth leaders, outperforming the national GDP growth rate in 2014-2017 (see Figures 2  and 10). The massive IDP influx has stimulated these regional economies and propelled further the urban and regional economies around Kyiv and Dnipropetrovsk (see Figure 10). More research is needed to investigate the relationship between high concentrations of IDPs, the local business environment, and the potential for entrepreneurship and business development in communities of protracted displacement.

Conclusion
This essay has shown the Ukrainian Donbas to be a typical shrinking old industrial region, suffering in the post-Soviet period from rapid deindustrialisation and persistent population loss. The Russo-Ukrainian war caused tremendous damage to the region in the period 2014-2018, with the total amount of permanent asset loss amounting to over US$84.6 billion in 2013 prices. The human cost of 13,000 lives lost and of many millions permanently scarred by the war has also been significant. The economic standing of Donbas as a whole within the Ukrainian spatial economy has more than halved, even considering the remaining economic activity in non-government-controlled territories. During the first four years of conflict, the officially recorded aggregate regional GDP plunged by 61.1% in constant local prices and by US$20.1 billion, or 73.1%, in foreign exchange terms. In terms of industrial output, the weight of the occupied territories of Donetsk province shrank to just 20.5% of the aggregate provincial production in 2017. In Luhansk province, ORDLO militants control 29.8% of aggregate provincial production. After the first four years of war, workers in ORDLO were being paid about 60% of their counterparts' earnings in the government-controlled areas, just over a third of the prewar wage, a loss of 62-66% in purchasing power terms. As the economic situation in ORDLO continues to deteriorate, the population outflow from the region has inevitably increased. The evidence presented in this essay shows clearly that ORDLO as a whole, let alone its two separate Donetsk and Luhansk sections, has an insufficient economic and demographic basis for anything resembling a financially and materially sustainable future as an independent statelet. The occupied Donbas is heavily dependent on Russia and formally governed by Moscow's Inter-Ministerial Commission for the Affected Districts, which acts as a shadow civilian government. The annual cost to Russia of sustaining ORDLO and maintaining and resupplying the armed formations operating there is in the region of US$5 billion (Goujon 2015). It is a costly but entirely affordable affair in the medium term, given that the total stock of financial assets controlled by the Russian state amounts to R30.7 trillion (US$507 billion). 41 The growing international pressure on Moscow, including further escalation of the Western economic sanctions regime against the Kremlin, its administrative apparatus and associated business interests, may provide an incentive to resolve the current Russian-Ukrainian stalemate in Donbas. 42 Nevertheless, Moscow's prolonged involvement in other frozen conflicts across the post-Soviet space, including its extraordinarily deep patronage of Transnistria, Abkhazia and South Ossetia, does not bode well for a peaceful settlement in the Ukrainian Donbas (Gerrits & Bader 2016). Indeed, President Putin's Executive Order No. 183 of 24 April 2019, 43 providing ORDLO residents with a fast-track procedure for obtaining Russian citizenship, has followed similar moves towards 'passportisation' in Russian protectorates, and is clearly aimed at deterring Ukraine from regaining its control over these territories in the future (Artman 2013;Nagashima 2019).
This essay has demonstrated that the ongoing armed conflict in the Ukrainian Donbas was not economically determined and cannot be explained in terms of economic rationality. Despite the existence of some long-term baseline separatism, Donbas was neither outstandingly prosperous nor excessively economically depressed-relative to the rest of Ukraine-to warrant an armed uprising of its own volition. Even if one assents to the historical materialist perspective that the social structures of production condition social relations in the final reading, the evidence presented here points to the paramount role of exogenous political agency and of endogenous military geography in explaining the outbreak of the armed conflict in eastern Ukraine in April 2014. Furthermore, the rationality postulate of neo-classical economic theory fundamentally depends on a set of working assumptions, which were clearly absent in Donetsk and Luhansk in the early 2014. For those interested in investigating further the local auxiliary causes of protracted geopolitical conflicts, this essay's recommendation is to employ the theoretical framework of social constructionism to uncover causal drivers at work.  3.3 Accommodation and food services activities (I) 9.9 6.9 3.0 Information and communication (J) 3.9 2.9 1.0 Financial and insurance activities (K) 8.7 6.9 1.8 Real estate activities (L) 13.2 9.7 3.5 Professional scientific and technical activities (M) 6.1 4.6 1.5 Administrative and support services activities (N) 11.9 9.6 2.3 Public administration and defence; compulsory social security (O) 10.2 6.4 3.8 Education (P) 11.2 7.3 3.9 Human health and social work activities (Q) 12.9 8.9 4.0 Arts, entertainment and recreation (R) 11.2 8.7 2.5 Other services activities. Activities of households as employers; activities of households for own use. Activities of extraterritorial organisations and bodies (S-U)