Policy Research Working Paper 10464 Impact of the Russian Invasion on Ukrainian Farmers’ Productivity, Rural Welfare, and Food Security Klaus Deininger Daniel Ayalew Ali Ming Fang Development Economics Development Research Group May 2023 Policy Research Working Paper 10464 Abstract Data from 2,251 small and medium-size farms for 2021 and formal and informal farms in the same size group. This 2022 show that area reductions in response to the Russian suggests that limited transferability of land use rights that invasion of Ukraine remained limited. However, worsening are disproportionately used by smaller farms may be one terms of trade reduced farm profitability, implying that 46 reason for low productivity. Improving transferability of percent of farms had a negative cash flow and 54 percent land, digital access to markets, and mortgage lending could (67 percent in the 50–120 hectare group) were credit con- thus trigger investment and growth in higher value products strained in 2022, implying that longer term effects may be by small and medium-size farms to solidify Ukraine’s com- more adverse. Total factor productivity varies significantly parative advantage in agriculture and improve rural living across size groups but is not significantly different between conditions in the context of reconstruction. This paper is a product of the Development Research Group, Development Economics. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at kdeininger@worldbank.org. or dali1@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Impact of the Russian Invasion on Ukrainian Farmers’ Productivity, Rural Welfare, and Food Security Klaus Deininger1 Daniel Ayalew Ali1 Ming Fang2 JEL Codes: Q10, O13, H56, R14 Keywords: Ukraine, land market, credit markets, conflict/war, agricultural production 1 The World Bank, 1818 H Street NW, Washington, DC 2 Michigan State University The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Funding support from the European Union (ENI/2017/387-093 and ENI/2020/418-654) is gratefully acknowledged. We particularly thank Victoria Zakhozha and the KIIS team for data collection, Nataliia Kussul and Andrii Shelestov for support with data processing, and Denis Bashlyk, Markiyan Dmytrasevych, Ben Hell, Thea Hilhorst, Roman Hrab, Sergyi Kubakh, Vasyl Kvartiuk, Andrii Martin, Roman Neyter, Oleg Nivievskyi, Taras Vysotskyi and Sergyi Zorya for helpful discussions and insightful comments and Daria Manzhura as well as Tania Khorzovskaya for outstanding coordination and administrative support. Impact of the Russian Invasion on Ukrainian Farmers’ Productivity, Rural Welfare, and Food Security 1. Background and motivation With a total agricultural area of close to 45 million ha of very fertile land, 1 Ukraine has long been a major breadbasket and exporter of agricultural commodities. Beyond causing immediate hardship and triggering a large exodus of displaced people, the Russian Federation’s military invasion of Ukraine and the economic blockade of its Black Sea export routes have also led to sharp increases in grain prices (von Cramon- Taubadel 2022) and raised concern about global food security. While fears of such consequences have eased somewhat with partial opening of maritime export routes in the context of the UN-brokered grain deal, there is little doubt that the war and associated political changes will profoundly affect Ukraine’s farm sector and rural areas. To better understand impacts on the welfare of Ukraine’s farm population as well as the profitability, food supply, and prospects for future development of the country’s agriculture sector, this paper reports results from a survey of small and medium sized farms for the 2021 and 2022 agricultural seasons. We aim to answer three questions: First, has the invasion created an imminent threat of decapitalization that might permanently undermine the rural economy’s capacity to invest including in human capital? Second, how have area cultivated, profitability, crop choice, and level of market participation been affected by the war overall, what are the underlying mechanisms, and how do war effects vary across farm size group and regions? Third, what is the extent of informality in Ukraine’s agriculture sector and how do productivity and war impact differ between farms in the formal and informal sectors? On the first issue, studies show that, if it forces households to liquidate assets to survive, exposure to conflict can have persistent effects, beyond short term impacts on food security (George et al. 2020; Muriuki et al. 2023) or agricultural output (George et al. 2021). In Colombia, violence led households to shift to activities with lower but less risky returns and, in the extreme, revert to subsistence (Arias et al. 2019). While direct effects may be of temporary nature only (Miguel & Roland 2011), violence can have long term effects if it is associated with refugee flows via forced displacement (Ibanez et al. 2022), reduces households’ investment in human capital (Büttner et al. 2022), leaves residues such as land mines (Arcand et al. 2015), or is used as cover for opportunistic land holding expansion by elites (Tellez 2022). Adverse long-term impacts of violence via human capital accumulation are documented for Nigeria’s 1967–1970 Civil War 1 Ukraine’s agricultural land endowment is larger than the agricultural area of France (18 million ha), Germany (12 million ha) and Poland (11 million ha) combined. 2 (Akresh et al. 2023) and Colombia’s mid-20th century ‘La Violencia’ that slowed structural transformation via occupational choices by individuals who experienced violence in school or pre-school age (Fergusson et al. 2020). 2 Conflict can affect those not directly exposed if households adjust their behavior (Tapsoba 2023) or asset portfolios (Rockmore 2020; Sinha et al. 2022) in anticipation. Our survey shows that, in Ukraine, the war triggered a marked drop in expectations for the future, but social safety nets and non-agricultural income sources held up. Ukraine’s comparative advantage in agriculture contributed to continued strong fundamentals in the sector; in fact, instead of divesting assets, most farmers expressed interest in expanding their holdings by leasing or purchasing land through markets. 3 Farmers’ perceived priorities focus on better access to output markets and protection against rapidly worsening terms of trade. Regarding the second question, estimates on the extent to which direct war-related actions have affected area cultivated vary widely partly because 2021 yielded a record harvest so that any direct comparisons that fail to adjust for weather or for potential substitution would tend to overestimate war effects. 4 Moreover, conflict can affect economic activity indirectly, e.g., by reducing trust (Korovkin & Makarin 2023), increasing mobility and trade cost to force input substitution (Amodio & Di Maio 2018), reducing interaction (Couttenier et al. 2022), or lowering project returns (de Roux & Martinez 2021). Impacts may thus be felt well beyond the areas directly affected by war and conflict (Federle et al. 2022). The farm survey data analyzed here shows that any direct war effect on area cultivated was eclipsed by large increases in prices for imported inputs that affect all farmers together with changes in labor cost that vary across regions in line with labor availability. Together with marked drops in output prices and reduced market participation, high input prices reduced farmers’ profitability and liquidity, an outcome most pronounced but not limited to areas affected directly by conflict. Finally, a distinguishing feature of Ukraine’s agriculture sector is that many producers remain informal (Nivievskyi et al. 2021). While there has been considerable progress on characterizing and quantifying informality (Ohnsorge & Yu 2021) and its correlates in terms of financial development (Capasso et al. 2022) and other development outcomes (Ulyssea 2020), the debate on the direction of causality remains largely open. If informal firms are less productive than formal ones (Nguimkeu 2022), they will be unable to compete (La Porta & Shleifer 2014) or afford fixed formalization fees. If this is the case, informal firms 2 Short-term health outcomes may also be affected (Akresh et al. 2011), especially close to the location of violent events (Akresh et al. 2022). 3 Due to egalitarian distribution of land when large holdings were privatized and a moratorium on land sales that was only lifted recently, most Ukrainian households own and rent out land. To the extent that rental contracts can adjust, better agriculture sector performance could thus provide broad-based benefit via higher land values. 4 USDA’s April 2023 USDA/WASDE report projects wheat production in Ukraine at 21.0 million tons, a drop of 12.1 million tons compared to previous years (see https://www.usda.gov/oce/commodity/wasde/wasde0423.pdf), much higher than what is estimated by the EU (Ben Aoun et al. 2022). As Deininger et al. (2023) show, some of the difference between estimates may be due to different geographical domains. A clarification of methodological assumptions will be desirable. 3 offered subsidized registration (Benhassine et al. 2018) would fall back into informality, a notion supported by a number of randomized trials from Latin America or South Asia (Bruhn & McKenzie 2014). If, on the other hand, entry barriers or other factors reducing the benefits of formality apply only to a subset of firms, eliminating these could set off large gains in terms of competition, output, and wages for skilled labor (Ulyssea 2018). We contribute to this debate in two ways. First, to measure the size of informality in the agriculture sector, we make use of the fact that informal farms operate unregistered land to combine our sample frame with estimates of agricultural crop cover based on satellite imagery. We find that 14.5 mn. ha, or about 32% of the 45.7 mn. ha of Ukraine’s cultivated area, is operated informally. Although informal farms are smaller, their total factor productivity (TFP) is indistinguishable from that of formal enterprises of comparable size, suggesting that, rather than productivity differences, either other barriers to farm entry (Deininger et al. 2018) or factors affecting the benefits from formalization for specific farms may have precluded formalization. One candidate for such barriers is restrictions on transferability and inheritance to rights of ‘permanent use’ of land that was given to emergent farms in the early 2000s. As use rights are concentrated with the smallest farms, such restrictions could partly explain variation in TFP across farm size groups. Moreover, although they can be transferred to a legal entity established by the original assignee, such personal rights would not survive the death of the original assignee, reducing the benefits from formalization and incentivizing farms with larger shares of land in use rights to remain informal. Exploring this in greater detail is an important area for follow-up research with potentially far-reaching implications for Ukraine’s future development. The paper is structured as follows. Section 2 provides conceptual and methodological background by discussing evidence of the aggregate impact of the invasion and details of constructing the sample frame and ways to infer the size of the informal sector. Section 3 provides evidence from our data at household and farm level for 2021 and 2022 to show how asset positions and factor market participation changed with the invasion. Section 4 provides evidence on how direct and indirect war effects changed land allocation, profitability, and productivity by farm size. Section 5 concludes and draws out policy implications. 2. Motivation and context This section discusses studies that assessed the impact of the Russian invasion of Ukraine on agricultural exports and global food security, the channels through which such impacts could come about, and the micro data underpinning such estimates. It highlights how representative farm level data can complement existing data sources and, by providing information on economic factors at a more granular level, informs estimates of negative conflict impacts and ways to try and minimize them. 4 2.1 Aggregate evidence on the effects of the invasion Given the historically strong integration of Ukraine as well as Russia into global commodity markets, 5 the Russian invasion of Ukraine, in February 2022, together with the economic blockade of its export routes through the Black Sea, resulted in a spike of grain prices and led to major concerns about effects on global food security. A key reason was the economic blockade of Ukraine’s main seaborne export routes via the Black Sea that limited Ukraine’s ability to generate export revenue and forced a costly shift to overland routes with much lower capacity but high physical and transaction costs. The ‘grain deal’ agreed under the auspices of the UN in July 2022 and extended twice since then, allows a limited quantity of grain to be exported from Ukraine through the port of Odessa, albeit at higher cost and risk than before the invasion. 6 Yet, while it may have avoided food insecurity in countries heavily dependent on Ukrainian imports (FAO 2022a), higher transaction cost of this export route may translate into lower prices for farmers in Ukraine. Trade models have been used to highlight likely impacts of different conflict scenarios on global wheat markets, especially for African countries (Balma et al. 2022), the resilience of the global food system (Ihle et al. 2022), or countries that have historically been characterized by heavy reliance on wheat imports from Ukraine (e.g., the Arab Republic of Egypt, Türkiye, Mongolia, Georgia, and Azerbaijan). A global model drawing on satellite imagery for different scenarios concludes that, even in the most optimistic case, reductions in welfare and shortages of food and energy are likely unless production gaps are compensated by increased production elsewhere (Lin et al. 2023). One concern has been that countries could use the crisis as a pretext for adopting protectionist policies (Ben Hassen & El Bilali 2022) that, while possibly yielding political benefits in the short term, could create negative externalities that would make achieving global food security more difficult (Glauben et al. 2022). The results of macro models depend significantly on the quality and granularity of the data used which are often derived from satellite imagery or household surveys. Remotely sensed imagery allows to identify changes in crop cover over time and separate direct conflict effects due to active fighting, mining, explosion of ordnance, heavy vehicle movement, and deliberate burning of agricultural fields from indirect ones at a granular level. Kussul et al. 2023 infer direct effects for 2-week intervals since the start of the war using satellite imagery and interpretation by means of machine learning as well as manual classification. 7 Using this information together with a 4-year panel (2019-2022) of cropped area for 10,125 village councils aggregated up from the field level, points to an area reduction of slightly less than 1 mn. ha (or 10%), well 5 Ukraine accounts for 10% of world trade in wheat, 42% trade in sunflower oil, and 16% of maize. 6 In March 2023, 3.5 million tons of grain and oilseed crops were exported through the Black Sea, adding to 2.8 million tons exported via rail, road, and river, bringing the total amount of crops exported in the 2022/23 marketing year to 44.7 million tons about 70% of the potential of 64 million tons (84% grain and 16% oilseeds), comprising the 2022 harvest as well as stocks that had been in storage in February 2022. 7 High resolution imagery is needed to detect craters from ordnance explosion, a measure correlated with presence of unexploded munitions (Duncan et al. 2023). 5 below those provided by aggregate models (Deininger et al. 2023). 8 Moreover, less than one-fifth of the total effects is attributable to direct conflict-induced field damage while the remainder comes about through indirect effects, including damages to the electricity grid and logistics infrastructure including road transport and grain elevators (Khoshnood et al. 2022); limited availability of key inputs including fuel, labor, and agro-chemicals; and reduced demand as a result of higher cost for transport and marketing. To link data on area cultivated and yield to welfare and prices or profitability, surveys at household or farm level are needed. Household surveys conducted by FAO highlight the importance of home production on garden plots dating from Soviet time as a safety net to increase resilience (FAO 2022b). In a nation-wide survey of 5,230 rural households, drops in income were reported by 55% of respondents, especially for IDPs. For the population who are not displaced three salient characteristics emerge, namely (i) 25% of the surveyed rural population reported reducing or stopping agricultural production due to the war; (ii) 72% of crop and 64% of livestock producers reported increased production costs; and (iii) more than half and around 20% of respondents reported spending more than 50% and 75% of their total expenditure on food, respectively (FAO 2022c). While these figures illustrate the breadth and depth of war impacts, disaggregated data on small and medium scale farm production is important to understand the extent to which the invasion may deplete the productive capacity of rural areas and to design responses on the continuum between humanitarian and productive support to maintain and expand productive capacity, diversification, and employment opportunities in rural areas. This is particularly relevant as experts agree that a significant part of Ukraine’s agriculture sector operates in informality, but little is known on either its size or the nature and profitability of its operations. 2.2 Sample construction and level of informality To obtain information on changes in welfare, production, and productivity in the small and medium farm sector between 2021 and 2022, a nation-wide phone survey of such farms in areas controlled by Ukraine, was conducted in cooperation with the Ministry of Agricultural Policy and Food (MAPF), from October to December of 2022.9 The original intent was to construct a sample frame using data from the State Statistics Service of Ukraine (SSSU), complemented with the company registry. However, these sources cover only registered legal entities. To capture informal farms that, based on expert estimates, cultivate 32% of Ukraine’s agricultural area (Nivievskyi et al. 2021), a decision was taken to use the State Agrarian Registry (SAR) as a sample frame instead. 8 https://downloads.usda.library.cornell.edu/usda-esmis/files/3t945q76s/cc08jp14g/cr56p755q/wasde0822.pdf 9 The survey was implemented by the Kyiv International Institute of Sociology (KIIS) with financial support from the European Commission. 6 The SAR is an electronic registry, established in August 2022, with the objective of transferring support to small and medium farmers in a transparent yet expeditious way. Farmers can sign up at the SAR website (https://www.dar.gov.ua/) using their electronic signature and provide a minimum of personal information including a bank account to which any resource transfers can be made, irrespectively of their legal status, i.e., registered legal entity, family-owned business (FOP), or individual. The system gathers information for all land parcels to which the farmer has registered rights from the registry of rights and the cadaster and adds information on the farm from several other official registries. 10 Information in SAR can be used by MAPF or any authorized entity to advertise or implement programs in support of the agriculture sector and to interact electronically with potential participants. Farmers can take any actions required digitally rather than by filling paper forms, including uploading scanned documents, photos, or providing authorization for providers of certain services to access specific types of personal information stored on the system. The SAR simplifies program implementation by providing reliable information from official registries and providing an incentive to ensure the currency of these; simplifying the process of program application by eliminating the need for repeated filling of paper forms; and leveling the playing field and increasing competition in factor markets by creating a basis of potential clients to which service providers or banks can market their services. Use of the SAR to implement an EU-supported producer support grant (PSG) as well as other programs, 11 facilitated rapid sign-up, especially by small and medium scale farms, many of which had been in the informal sector before. As the survey was launched in the second half of October, we set Oct. 15, 2022, as the cut-off date to construct the sample frame by dropping farms without any registered land or a valid phone number. Table 1 provides comparison of the resulting sample frame with the 2020 all registered farms survey (Form 29) of SSSU by farm size category (<50, 50-120, 120-500 and > 500 ha). Panels A and B of table 1 presents a simple comparison of number of establishments and total area cultivated of the two frames. Panels C and D draw out implications for informality in areas controlled by Ukraine (ACUs) and areas affected by conflict (AACs) by comparing, for each of these, the area cultivated by farms that are included in our and SSSU’s frame or in either of these frames separately to identify the area cultivated informally (i.e., by farms that are neither included in Form 29 nor the SAR). 10 For example, the SAR automatically gathers information on any outstanding debts to the state (which would legally disqualify them from receiving state support) and on farmers’ registered livestock from the animal registry and the government plans to add information from other registries, including the mortgage registry and the registry of court cases, in the near future. 11 The PSG that targeted to farmers who cultivated less than 120 ha provided a cash grant equivalent to US$100 for each hectare of land cultivated during the 2022 agricultural season in non-conflict affected areas. Information from the SAR was first used to establish whether the farmer was below the 120 ha threshold. After subtracting land parcels registered in the farmers’ name located in conflict affected areas, maps with registered parcel boundaries were used to cross-check cultivation status against a crop map elaborated based on remotely sensed imagery to compute the total grant amount. The € 50 mn. available under the program had been fully disbursed by Nov. 2022 and an evaluation that aims to compare differences over time in parcels by successful as compared to unsuccessful PSG applicants is currently underway. Beyond this, several donors are using the system to implement programs such as provision of short-term grain storage, delivery of seeds, fertilizer, or critical capital equipment (generators). 7 Panel A of table 1 shows that our frame includes 75,571 farmers, more than double the 36,184 in the Form 29 survey as used by SSSU. Of these, 21% are legal entities, 11% FOPs, and 68% individuals overall with the latter concentrated in the farm size class below 50 ha of which 82% are individuals and 9% each FOPs and legal entities. Legal entities dominate for larger farm size classes, making up 58%, 76%, and 97% of those in the 50-120, 120-500, and > 500 ha group, respectively. Coverage in terms of farm numbers by the SAR is better than that of Form 29 for the below 120 ha group (with 3.8 and 1.1 times the number of farms in the < 50 and 50-120 ha) but worse above 120 ha with SAR containing only 0.62 and 0.44 times the number of operations in the 120-500 and the >500 ha group, respectively. Table 1 panel B illustrates that the farms covered by Form 29 and SAR cultivate 27.85 and 11.6 million hectares, respectively with 25% of Form 29 area in ACUs and 25% in AACs. While they comprise less than a quarter of the number of registered farms, legal entities operate more than 90% of the area registered in SAR overall (33%, 64%, 83%, and 97% of area in the <50, 50-120, 120-500, and > 500 ha groups, respectively). The total area registered in SAR amounts to 42% of the area in Form 29 at the national level and 150%, 127%, 67%, and 37% of which in the <50, 50-120, 120-500, and > 500 ha groups, respectively. 12 To quantify the extent of informality in Ukraine’s agriculture sector, we use crop maps for 2020 elaborated using the methodology described in Kussul et al. (2017); Shelestov et al. (2017), and Shelestov et al. (2020) to identify the total area cultivated with crops in 2020 and then compare this area to what is covered by Form 29 as well as SAR. 13 Table 1 panels C and D highlight that, based on these maps, Ukraine’s total cultivated area amounts to 45.73 million ha, 33.87 mn ha in ACUs and 11.86 mn ha in AACs. With 27.8 mn. ha (21 mn ha in ACUs and 6.8 mn ha in AACs), SSSU’s Form 29 thus covers 61% of the country’s cultivated area in 2020, leaving an estimated 39% of cultivated area in informality. Using enterprise IDs to match farms between SAR and Form 29 suggests that 7.5 mn ha nationally (6.4 mn. ha in ACUs and 1.1 mn ha in AACs) is cultivated by farms included both in SAR and Form 29 while 3.36 mn. ha, almost all in ACUs, are registered in SAR only. This suggests that in the 3-month period from Aug. 12 to Oct. 15, 2022, registration in SAR reduced the amount of land cultivated informally by 3.36 mn. ha or about 7% of the total. Still, 14.5 mn. ha, more than the total agricultural area of Poland or Germany, are cultivated by farmers that are neither part of SSSU’s sample frame nor registered in SAR, implying that this land is not reflected in any official statistics. 12 As a plausibility check, we compare mean farm sizes between Form 29 and SAR. For the above 50 ha group, average farm size in SAR is 803 ha compared to 934 ha and 999 ha for ACU and AAC, respectively, in Form 29 and the corresponding figures for farms greater than 500 ha are 2,834, 2,255, and 2,137 ha, respectively, for SAR and Form 29 in ACUs and AACs. 13 As area cultivated in 2022 is likely to have been affected by the conflict and to avoid bias that might arise from the fact that 2021 was an exceptional year with very favorable growing conditions, we use both satellite imagery and Form 29 data from 2020. Area cultivated excludes forest, bare land, wet land, water, and areas with artificial (built up) cover. 8 3. Data and descriptive statistics Household data show that, although continued access to non-agricultural incomes, access to public social support and the scope to draw on informal safety nets provided short-term support, the war led to a dramatic drop in long-term perspectives. In the agriculture sector, continued high demand for renting in or buying land together with low willingness to sell land points to strong sectoral fundamentals. Size-biased access to credit that is reinforced by traditional agricultural subsidy policies may, however, limit access to working capital finance even for investments that would be advantageous in the short term and, together with a complete lack of long-term finance, may limit farmers’ ability to take advantage of opportunities for long term investment and diversification into higher value crops in due course. 3.1 Sample, household characteristics, coping strategies, and conflict effects To obtain the sample that could form the basis for a phone-based farm survey, we exclude conflict affected areas and define four farm size strata (< 50 ha, 50-120 ha, 120-500 ha, and > 500 ha) for a stratified random sample. 14 To assess representativeness, panel A of table 1 compares median and mean farm size between SAR and Form 29, the frame currently used by SSSU, in each stratum, pointing towards a near perfect match in the 50-500 ha category, no difference in the median for farms in the > 500 ha category (but better coverage of large farms in Form 29) and a much better coverage of small farms in SAR as compared to official statistics. A graphical representation of the sample’s geographic coverage in figure 1 shows it covers most of the rayons in areas not affected by conflict but coverage is scant in the East and the North. 15 At the farm level, the survey asked about total output and market participation as well as input costs for key inputs (hired labor; fertilizer and chemicals; machinery and fuel) by crop. If respondents could not provide information by crop, inputs were recorded at a more aggregate level. For owner-operated farms (82%), information on household composition, key income sources, and coping mechanisms is also included. As far as possible, questions were asked for before and after invasion or the 2022 and the 2021 cropping season. In tables 2-6, we present information by region (columns 2-6) and farm size group (columns 7-10). Key data on household composition in table 2 suggest that Ukraine’s small and medium farm owners are about 50 years old, well educated, and linked to banks as well as the internet. 16 The head almost always has a bank account, higher than secondary education in 89% of cases, and a Facebook account in 67% of farms. 14 To allow use of the sample as a basis for evaluating the PSG cash grant, farms in the below 120 ha category are split equally into PSG applicants and non-applicants. 15 Appendix table 1 shows that, with less than 21% of the sampled farms resulting in a successful interview (66% in the first, 22% in the second, and 12% in the third attempt), non-response (45%) and refusal rates (34%) were high. As indicated there, large part of the non-responses can be attributed to war conditions with cellphones out of range or power the most frequent reason. 16 The high level of digital literacy might be accelerated by the shift towards digital service provision associated after the invasion. As opening a bank account is one of the most expeditious ways of obtaining an electronic signature for individuals who do not yet have one, the share of heads with a bank account may be biased upwards though this does not apply to other household members. 9 Some 72% and 67% of farm households have multiple bank or Facebook accounts. War-related damage to land or structures were experienced by 8% or 4% of respondents overall, being most prevalent in the East (26% and 15%), followed by the North (21% and 11%) and the South (10% and 5%) in line with the incidence of hostilities. While outmigration of an entire household would not be captured in the survey but instead result in a non- response, family members migrated within or beyond the oblast in 8% or 9% of cases, respectively. The rate of beyond-oblast migration is highest in the East (18%), followed by the South (11%). With 46% indicating that they could get informal support up to a mean of US$7,311, an amount that ranges from $4,145 for the below 50 ha farm size group to $12,141 for farm households cultivating more than 500 ha, informal safety nets remain common. This is consistent with studies showing that exposure to war violence increases local cooperation (Bauer et al. 2016) and changes individual behavior to be more altruistic, risk- seeking, and with higher discount rates that affect savings and investments decisions (Voors et al. 2012). To assess if households would use additional cash for consumption, working capital, investment, or savings, respondents were asked on how they would use a hypothetical small (US$200) or large ($5,000) amount won in the lottery. Data in table 2 show that working capital was by far the most frequent response, accounting for 56% of valid responses for the small and 69% for the large amount, especially for farms below 500 ha. 17 It was followed by investment (20%) for the $5,000 amount and consumption (24%) for the small amount of $200. The propensity to invest a larger cash windfall was highest for farms greater than 500 ha, possibly suggesting that this farm size group was better able to deal with working capital shortages than smaller farms (Deininger et al. 2023). Table 3 provides summary statistics on income sources, attitudes, and coping mechanisms in 2021 (panel A) and 2022 (panel B) to assess changes over time. The number of adult household members (2.6) remains roughly the same in both years with close to 2 adult members working in agriculture either full-time (1.45) or part-time (0.62). About 54% of farms had permanent employees, a share that increased from 18% and 45% for those in the below 50 and the 50-120 ha class, respectively, to 82% and 98% in the 120-500 and the above 500 ha class with a concomitant increase in the number of employees from 1, 1.15, and 4.5 to almost 50 for the largest farms. There was little change in the share of farms with members obtaining resources from off-farm wage (94%) or self-employment (62%). In fact, average monthly income (or revenue) dropped only slightly by about 9% for the former (from $253 to $231) and 7% (from $288 to $269) for the latter. The drop in wage or self- employment income was most pronounced in the South (with 14% and 15%) and East (13% and 6%). 17 Unambiguous answers were provided by 94% of respondents for the large and 72% for the small amount. 10 To assess changes in overall expectations due to the invasion, respondents were asked to rank their personal situation as well as that of the country, respectively, on a scale from 1 to 10, both before and after the invasion. Figures in table 3 point towards a marked drop in both from 7.34 and 7.15 to 3.66 and 3.24, respectively and with a drop of more than 4 points in both categories, it was most dramatic in the East. Social assistance programs including pensions emerge as a key coping mechanism with 32% having a member who received on average US$69 per month before the invasion. The share of recipients of such assistance increased by 7 percentage points to 39% during the war; albeit, with a slight decrease in the average amount received, highlighting the continuity in government functioning. This expansion was particularly pronounced in the South where the share of households with at least one social assistance recipient increased by 14 percentage points to 47%. Compared to public support, remittances were received by only 4% in 2021 and 5% in 2022, an increase driven by the South where the share of remittance recipients increased from 3% to 6%. 3.2 Asset ownership and factor market participation In terms of farm characteristics, table 4 highlights that virtually all (84%) farmers lease land at an annual lease price of about US$103.7/ha.18 This is more than three times the lease price that prevailed in 2015, 19 suggesting that reforms that improved tradability of land and increased competitiveness of land markets significantly improved small landowners’ welfare. Across farm size groups, amounts of lease payments are the lowest (US$86/ha) in the smallest farm size group and the highest (US$117) in the more than 500 ha category. To gauge economic prospects in agriculture, respondents were asked if they would be interested to either lease in more land at the current lease price or buy or sell land at a price they could name. In contrast to general pessimism about the future as documented above, responses point towards strong fundamentals in the agriculture sector despite the war: 85% of farmers (from 80% in the South to 87% in the Center) indicated they would like to lease more land at the current lease price; for farms above 50 ha, this share is well above 90%. In addition to strong demand for leasing land, farmers are also eager to buy land, with 78% indicating they would like to buy land (56%, 86%, and 89% in the < 50 ha, 50-500 ha, and 500 ha group, respectively). Interestingly, the price farmers report to be willing to pay (US$1,432 per ha on average, ranging from $1,353 for those in the < 50 ha group to $1,548 in the > 500 ha group) are in a narrow band slightly above 18 Only farmers in the below 50 ha group pay, with US$87.7/ha, significantly lower leases. This may be attributable to greater prevalence of informal arrangements in this group or to respondents including land that has been transferred to them by the state for permanent use in the ‘leased’ category. 19 See https://kse.ua/wp-content/uploads/2022/06/eng_jan2022_land_monitoring.pdf; http://www.kse.org.ua/en/research-policy/land/governance- monitoring/yearbook-2016-2017/. 11 the post-war market price (of $1,369) for agricultural land (Deininger & Ali 2023), pointing towards broad awareness of market conditions after the 2021 opening of sales markets for agricultural land. Compared to the near-universal desire to buy more land, only 3% are willing to sell land, a share highest in the <50 ha farm size group (6%) and zero for farms in the > 500 ha category. The mean proposed sales price, $3,142 (ranging from $4,787 in the > 50 ha group to $1,206 in the 120-500 ha group), is mostly above what potential buyers are willing to pay, consistent with a strong endowment effect (Zhu et al. 2021) and highlighting that, with strong fundamentals, agricultural land continues to be a promising investment. As a proxy for farms’ capital stock, information was obtained about ownership and current value of tractors and mechanical equipment. With 84% (82%, 88%, and 72% in the < 50 ha; 50-500 ha, and > 500 ha groups, respectively) of sampled farms owning tractors, a significant share of large and small farmers rely on machine rental services. The per ha value of equipment ranges between US$610 and US$680 for all size groups, except the < 50 ha one ($1,213), possibly due to indivisibilities and frictions in markets for rental services. Data on credit market participation suggests that access to credit is limited and biased in favor of larger farms: 55% of farmers overall—34% in the < 50 ha group vs. 84% in the above 500 ha group ever accessed credit. While 2.3 years passed since the average credit recipient obtained her last loan, the average farm in the > 500 ha class received its last loan 9 months ago (0.78*12), i.e., in the 2022 production season after the invasion. This compares to a gap of 4.8 and 2.9 years for farms in the below 50 and the 50-120 ha class. With an average loan size of US$200,000 at a nominal interest rate of 5.4% that is likely to reflect interest rate subsidies provided by the government’s 5-7-9 program, 20 loan terms were more favorable for larger farms than for smaller ones. The latter, if they cold access credit at all, received amounts of US$7,028, US$17,597, and US$53,418 at rates of 11.5%, 8.3%, and 7% for the <50, 50-120, and 120-500 ha group, respectively. Such size bias is inherent to the way in which the interest rate subsidy program is currently administered. 21 The data also show that, except for loan recipients in the smallest farm size group, 21% of which borrowed for purposes other than working capital—most likely to smooth consumption in an emergency—virtually all the credit received went to short-term working capital loans. If the ability to use agricultural land as collateral for investment, backed by longer-term credit, to diversify into higher value-added agriculture is 20 For a description of the program, see https://www.kmu.gov.ua/en/news/uriad-rozshyryv-prohramu-5-7-9-na-pidpryiemstva-zruinovani-pid-chas- viiny; for the design https://www.kmu.gov.ua/en/news/oleksij-goncharuk-uryad-zapustiv-programu-dostupnih-kreditiv-5-7-9-pid-rekordno-nizkij- vidsotok#:~:text=The%20State%20Program%20%22Affordable%20Loans%205%2D7%2D9%25,Ukrainian%20micro%20and%20small%20ent erprises; https://ukraineinvest.gov.ua/news/22-11-22/ and https://www.kyivpost.com/post/8003 for figures for November 2022 and possible impacts. 21 As the current program structure requires producers to obtain a loan first and then provides a portfolio-based interest rate subsidy, it will do little to increase credit access beyond those who have been able to secure loans in the past. 12 to be realized, focusing subsidies on market imperfections and changes in the way government supports credit access by farmers are likely to be needed. 22 We define farmers who indicate that they would want to borrow but did not receive a loan in 2022 as credit constrained. Table 4 shows that 54% of farmers overall fall into this category but that there is significant variation across farm size groups: credit constraints are most binding for farms in the 50-120 ha category, more than two-thirds (67%) of whom are credit constrained by this measure. With 46% and 33% in the 120-500 and the > 500 ha group, the share of constrained producers is much lower at larger farm sizes due to better credit access. With only 62% interested in borrowing, lower demand is the main reason for lower incidence of credit constraints in the < 50 ha group. Improved access to credit for smaller farms could thus make a significant contribution to leveling the playing field. Although 91% indicate that they would welcome technical support, access to advisory services is, with 18% overall, of which about half is provided by government and the remainder by other farmers or non-profits (with little difference across farm size categories), more limited than credit access. To identify priorities for public sector engagement in the current situation, farmers were asked to name the most important areas for government action. Credit (15%) is in third place, after measures to reduce high input prices (29%) and improve access to output markets (25%) take first and second place. To identify reasons for this, we turn to changes in output and input prices as well as overall production and the agriculture sector’s terms of trade after the Russian invasion. 4. Production outcomes Although limited war impacts on area cultivated highlight the resilience of Ukraine’s agriculture sector, higher output and lower input prices worsened terms of trade that, together with declining participation in output markets, led 50% of farms to experience a negative cash flow in 2022. While direct damage reduced productivity by 12%, indirect war effects decreased it by between 30% in the South and 7.5% (marginally significant) in the West. TFP increased slightly with size but does not differ between formal and informal farms, making identification or removal of factors that keep farmers in informality an important area for future research and policy. 4.1 Changes in land use and economic outcomes Evidence on production outcomes, in terms of area devoted to main crops for the 2021 and 2022 cropping seasons, is displayed in table 5 using the same structure as table 3. For sampled farms (i.e., the farming population outside the AACs), between 2021 and 2022, cultivated area increased by 7%, from 346 to 371 22 Such changes will need to be complemented by regulation to reduce the transaction costs of posting or foreclosing on collateral pledged for loans in case of default. 13 ha per farm. This is in line with evidence from remote sensing that suggests that (i) cultivated area decreased less than implied by aggregate estimates; and (ii) in areas such as the North, farmers were able to plant summer crops on fields where 2022 winter crops were damaged or destroyed during the early phases of the war. The magnitude of the increase in area planted was similar across all farm size groups, though there are regional differences with area cultivated dropping by 9% in the East and increasing by 12% in the Center. Data for crops cultivated in the 2022 season point towards a shift from grains to oilseeds such as soybean and rapeseed (+42% and +22%, respectively, both from a modest base) or sunflower (+18%) compared to 2021. These crops either require less fertilizer or can sustain short-term lack of fertilization better due to a deep root system and often also are processed locally, lowering transport cost for output marketing. Maize and barley area decreased by 10% and 3%, respectively while wheat area increased by 4%. Contrary to what is reported from household surveys (FAO 2022b), area under ‘other’ crops (buckwheat, rye, beets, potatoes, etc.) decreased by 3% overall, mainly in the largest farm size group (-15%), a decrease balanced by increases of 17% and 30% for the 50-120 and the 120-500 ha class and in the South (+36%). Regional disaggregation also suggests the shift out of maize is driven by the North which, together with the Center, also saw a large increase in area cultivated with sunflower, soybean, and rapeseed. 23 While changes in crop mix across regions could possibly be observed using remote sensing (and linked to farm size categories if there is a complete cadastral map linked to the registry), only survey data can provide economic information on costs, prices received, input and output market participation, profit, and cash flow. To obtain such information, the survey collected data on the cost of hired labor, purchased inputs (fertilizer, pesticides, and other chemicals for crop protection), and machinery services by crop for the 2021 and the 2022 agricultural season. In several cases, respondents instead provided total input cost or total cost of machinery services and purchased inputs either by crop or for all crops grown by the farm.24 If inputs are reported at farm rather than crop level, we apportion them across crops using crop area as a weight. As respondents provided consistent information only on output prices, we rely on a survey of community leaders and local dealers’ adverts, complemented by completed auctions for machinery services on the Prozorro sales platform, to compute regional prices for fertilizer, fuel, machinery services, and labor in 2021 and 2022. 25 Appendix table 2 shows that at the national level, prices for fertilizer more than doubled, those for diesel increased by 90%, and those for machinery services by 31%. Wages increased by 25% on average, with differences across regions (no increase in the North vs. 75% in the South). 23 See the discussion of regression evidence in table 7 below. 24 To separate different categories of input costs in cases where costs were reported in total, we use median cost shares in the sample as follows: If non-labor inputs are reported jointly, purchased material inputs (i.e., fertilizer and other chemicals) and machinery services account for 64% and 36% of the total, respectively. If labor and non-labor inputs are reported jointly, they account for 15% and 85% of the total, respectively. 25 Prices are for ammonium nitrate and urea for fertilizer and the cost of harvesting and ploughing/soil preparation for wheat and maize, respectively, for machinery services. 14 Figures in table 6 illustrate that, valuing per hectare output for the 6 main crops at market prices implies a monetary output value of US$872 in 2021 but only US$571 in 2022, a 34% reduction (see appendix table 3 for corresponding information at crop level). Valuing 2022 output at 2021 prices suggests that this figure comprises a 19% decline in real output/ha and a 15% drop in real output prices that is likely to be linked to higher transport cost. Transport cost also contributed to an increase in the cost of purchased inputs (fertilizer and chemicals, machinery, fuel, and hired labor) from an equivalent of US$154/ha to US$184/ha, despite a drop of real input quantity by almost 40%. Together, this resulted in a drop of revenue net of purchased inputs by almost 50%, significantly reducing farmers’ ability to pay for fixed factors (family labor, capital depreciation, land rent, entrepreneurial risk). Across farm size groups, 2022 output value is, with US$471, lowest in the below-50 ha size class and highest, with $704, for those above 500 ha. As large farms also spent more on purchased inputs (despite reducing spending in real terms), net revenue is, with US$380/ha, marginally higher for farms in the 50- 120 ha group than for those above 500 ha (US$368/ha), followed by the 120-500 ha group (US$345/ha). Although these trends are similar across regions, output value and net revenue are least affected in the West, possibly due to better access to external markets, where output value decreased from US$962 to US$738 and net revenue from US$735 to US$496 which is more than double the value in the East (US$217), 59% above that of the North (US$311), and 36% above the level in the Center (US$364). The above figures assume all output produced is sold in the market, as was largely the case in the 2021 season when, except for 15% in the smallest farm size group, virtually all farms participated in output sales markets and sold on average 83% of their output. 26 Compared to 2021, the share of producers who sold (or planned to sell) in the market dropped by 17 percentage points from 94% to 77% and the share of the harvest sold from 83% to 57%, a 26 percentage point drop. To illustrate potential impact of worsening market access on farmers’ cash on hand, we subtract the cost of purchased variable inputs from actual sales revenue to obtain cash on hand. The challenges posed by worsening terms of trade to farmers’ liquidity are highlighted by the fact that this measure of cash dropped from US$432/ha in 2021 to -US$26/ha in 2022. Moreover, while 10% of farmers had negative cash on hand in 2021, this share increased to 46%% in 2022 (51% in the South and West and 56% for the largest farm group). To summarize patterns by region and farm-size group and account for exogenous shock of war damage, we regress variables of interest in both years on an indicator variable of whether the farm’s land or structures were damaged by war-related fighting in 2022, an interaction between region or farm size group, and a 2022 dummy. Results for area, yield, a dummy for output market participation, the share of total output 26 To the extent that farmers stored the 2021 harvest and were planning to sell it after February, they may have been affected by the intervention. 15 value sold, the amount spent on inputs, net revenue, cash flow, a dummy for having had a negative cash flow, and the share of cropped area devoted to oilseeds are summarized in table 7. Beyond reducing cropped area and monetary output by 12.4% and 11.1%, exposure to direct war-related damages did not affect any of the variables considered here. Other than for reductions due to war damage, area cultivated in 2022 was not significantly different from 2021. Output value per ha dropped by between 37.5% in the South and 14.4% in the West over the two periods while the year effect is not significantly different from zero across farm size groups once regional effects are controlled for. Market participation at the extensive margin decreased by about 15% in the East and the North and for the below-50 ha size category, and 12%, 10%, and 9% in the West, Center, and South, respectively. It also decreased by more than 30% at the intensive margin in all regions—from 37% in the East to 31% in the North—again no additional variation across farm size groups. Real input value dropped by between US$184 and US$154 across regions but increased by US$112 for the smallest farm size group and US$80 for those in the 50- 120 ha farm size category. Stark declines are also visible for farm net revenue per hectare which declined between US$356 in the West and US$452 in the North and cash flow, which dropped by between US$819 in the North and US$556 in the South, an effect that was less pronounced for farms in the below-50 ha group. The share of farms with negative cash flow increased by between 53% in the East, 46% in West and Center and 47% in the North. We also note that in line with descriptive statistics, the share of area cropped with oilseeds increased by 11% in the North. 4.2 Levels and changes in productivity While the above evidence points towards direct war effects on area cultivated and indirect effects war- induced changes in the intensity of input use, net revenue, and cash flow, inferences on productivity require a production function estimate in per-hectare terms. Indexing farms by i and time by t, we use a farm fixed effects regression of the form 𝑌𝑌𝑖𝑖𝑖𝑖 = 𝛼𝛼𝑖𝑖 + 𝑺𝑺𝑖𝑖𝑖𝑖 𝜷𝜷 + 𝑿𝑿𝑖𝑖𝑖𝑖 𝜸𝜸 + 𝑹𝑹𝑹𝑹𝑖𝑖 𝜹𝜹 + 𝝀𝝀𝒊𝒊 + 𝜀𝜀𝑖𝑖𝑖𝑖 , (1) where Yit is the value of output per ha, 𝛼𝛼𝑖𝑖 is a farm fixed effect, Sit an indicator of whether the farm’s land or structures sustained war-related damage in 2022, 𝑿𝑿𝑖𝑖𝑖𝑖 is a vector of the monetary value of inputs (labor, chemicals, machinery) used by farm i in period t; 27 RTi is an interaction between region and the 2022 war period; 𝜀𝜀𝑖𝑖𝑖𝑖 is a random error term; and 𝛼𝛼, 𝜷𝜷, 𝜸𝜸 and 𝜹𝜹 are parameters to be estimated. 27 To convert the stock of family and hired labor as well as machinery into a service flow, we assume that (i) full-time family and permanent farm workers spend on average 240 days per year on agricultural activities while part time family workers spend half of that time and their wages are equal to those of casual workers; (ii) tractors and equipment depreciate linearly over a 10-year lifespan with a 36% salvage value; and (iii) for the cases where inputs are aggregated across crops, imputed values are apportioned using crops’ area share. 16 Results for (1) are in cols. 7 and 8 of table 8, complemented by those from the OLS version from the 2021 and 2022 cross sections (in cols. 1-4) and the combined sample using OLS (cols. 5 and 6). In all cases, direct damage from the war is estimated to reduce productivity by about 12%. Indirect war effects differ across regions with point estimates of 29% in the South, 20%, 11%, and 9% in the Center, East and North and 7.5% (significant at 10%) in the West. Coefficients on land, labor, fertilizer, and machinery are highly significant in the OLS regressions, with point estimates of between 0.094 and 0.111 for land, 0.029 and 0.045 for labor, and 0.072 or 0.056 for fertilizer and machinery, respectively. Only fertilizer is significantly different from zero (with a coefficient of 0.051) in the fixed effects regression. Farm fixed effects 𝛼𝛼𝑖𝑖 net of rayon fixed effects provide a measure of farmers’ total factor productivity (TFP). Figure 2 which plots estimated TFP against cultivated area highlights great dispersion in productivity levels and modest increases in productivity with size. 28 Results from regressing TFP on farm size categories and an indicator variable for formal registration provide two insights. First, differences in TFP across farm size groups are statistically significant, most notably a large gap of 14-20 percentage points between the smallest farm size category and the next (50-120 ha). Differences between subsequent categories are much smaller with gaps of 4 and 5-7 points, respectively, between the 50-120 and the 120-500 ha as well as the 120-500 and > 500 ha category. 29 Second, registered farms’ TFP is not different from that of informal ones overall (table 9, col. 2) or in the same size group (table 9, col. 3). If formality is not associated with differences in TFP, it will be of interest to identify—and possibly affect through policy—factors that differ systematically between formal and informal farms and that may, by affecting either the cost of the benefits from formalization, may be a reason for farms to stay in informality. A model suggests that identifying and eliminating such factors could eliminate deadweight losses and in doing so increase competition, aggregate production, and skilled formal sector wages (Ulyssea 2018). Restrictions on land given to small farms in the 1990s or early 2000s for ‘permanent use’ under favorable conditions but, until 2022, 30 without mechanisms to acquire ownership rights to such land or transfer it to others, are one candidate. As they are non-transferrable and assigned to individuals, they will not survive the death of the original assignee. 31 creating tenure insecurity that is likely to reduce incentives for land- attached investments. Moreover, as most of these rights were assigned 2-3 decades ago, such restrictions 28 Appendix table 4 tabulates mean and dispersion of TFP overall and separately for legal entities and ‘informal’ farms. While TFP increases and dispersion decreases with size, for the 50-500 ha categories, the hypothesis of levels and dispersion of TFP being equal between formal and informal farms cannot be rejected. 29 Note that credit constraints are estimated to decrease TFP for 50-120 ha farms by 5 points, an estimate that is significant at 10%. 30 Short-term (1 year) leases of land operated under permanent use rights were allowed in March 2022. Legislation adopted in November 2022 provides citizens and legal entities with the right to permanently use a land plot of state or communal property to buy the corresponding land without land auctions. The purchase of agricultural land plots is carried out at a price equal to the normative monetary value of the land plot (UAH 27,500/ha) and the buyer has the right to opt for paying for the land in 10 equal instalments that are indexed to inflation. 31 Formally, in case of death, land will revert to the community or state, the tenure while village councils can-and have in some cases-provided permanent use rights to the assignee’s heirs, they are under no obligation to do so. 17 may prevent many of the original recipients from formally exiting farming and transferring their land to younger farmers able to make better use of it and mechanisms to ensure such transfer will be important. As smaller farms rely disproportionately on land held under use rights, high levels of tenure insecurity and limited investment incentives for land held under permanent use could be a plausible explanation for the low levels of TFP by farms below 50 ha. Using the cadastral designation of land can provide a quick check and suggests that 58% of land by farms below 20 ha and 36% for those below 50 ha, as compared to 20% for farms in the 50-120 ha class, is for ‘personal farming’ rather than ‘commercial agriculture’, a difference that is significant at 1% in both cases. Making use rights more transferable, possibly by helping the smallest farm size group acquire full ownership rights to the land they farm thus seems to have potential to boost Ukraine’s future development by increasing the welfare of this group, competitiveness of the land market, land-related investment and use of land as collateral, as well as the amount of local land tax receipts. 5. Conclusion and policy implications This paper aimed to complement existing studies by assessing impacts of Russia’s invasion of Ukraine on agricultural performance and food supply based on a micro-level economic data and by including informal farms, a segment of Ukraine’s agriculture sector that is often overlooked in official statistics. Two findings are noteworthy. First, our data show that surveyed farmers kept their cultivated area in 2022 almost unchanged, a sign of farmers’ resilience. At the same time, worsening terms of trade for agriculture led to a precipitous regionally differentiated decline in farm profitability and productivity. Monitoring changes in planted area for winter as well as summer crops in 2023 based on satellite imagery will be important to identify the extent to which such changes in profitability triggered longer-term adjustments like decapitalization. Second, our data suggest that, especially with the country undergoing a significant crisis, an exclusive focus on (large) farms that are formally registered risks permeating old biases and in doing so tilting the playing field against potential sources of dynamism in rural areas that will be needed if Ukraine is to emerge from the war with a more diversified rural economy that focuses on high value crops in addition to grains and oilseeds. Three implications are particularly relevant: First, efforts to improve factor and output market competitiveness will be essential to prevent war-induced problems of market access and logistics being exacerbated by exercise of market power and efforts to curtail competition. Ukraine’s success in facilitating competitive electronic procurement and transfer of rights to state assets including land via e-auctions (Deininger et al. 2022) together with the advances made in setting up the SAR and using it to disburse emergency productive cash support to small farmers during the war, suggest MAPF is well placed to quickly work with input suppliers, output traders, banks, and crop insurance providers, to explore how SAR can be used as an electronic platform to increase competitiveness and reduce risk, including via forward contracts, even at a time of war. 18 Second, our data suggest productivity of farms below 50 ha is low in part due to investment disincentives and transferability restrictions associated with land under permanent use rights, they also show that informal farms are as productive as formally registered ones and that productivity differences between farmers of different size groups above 50 ha are much narrower than policy makers seem to believe. The rapid growth of SAR registrations suggests that farmers will come out of the shadows if doing so provides benefits. Measures to expand this into a digital marketplace and facilitate land transfers and boost mortgage lending using land as collateral, including greater transferability of land given under rights of permanent use rights could thus have a transformative effect by setting in motion a wave of investment and growth in higher value products by small and medium-sized farms to solidify Ukraine’s comparative advantage in agriculture and improve rural living conditions. Finally, evidence-based decision-making to help the country address the unprecedented current challenges and maximize economic potential and public accountability during reconstruction requires that quality farm survey data like the ones analyzed here will need to be provided regularly, ideally by the country’s State Statistics Service (SSSU) including better coverage of the entire farm size spectrum, and in ways that can be linked to data from other sources. To move in this direction, there is a need to amend the legal basis for agricultural statistics to be in line with global best practice in terms of harnessing satellite imagery, computerized data entry, and dissemination of micro-data for analysis without sacrificing confidentiality. An expeditious move towards such a framework would allow alignment with EU membership requirements and help make information collected by SSSU more representative of Ukraine’s farm sector, timelier, and more meaningful for policy decisions to navigate the profound war-induced challenges to the country’s economic and social fabric in a way that is based on evidence rather than ideology and encourages participation and input by all stakeholders. 19 Table 1: Sample frame compared to national statistics Total Farm size in ha < 50 50-120 120-500 > 500 Panel A: No. of farms & median/average farm size Total Form 29 36,184 16,138 5,401 7,144 7,501 ACUs % 78.9 82.0 79.7 76.2 74.0 AACs % 21.1 18.0 20.3 23.8 26.0 Total SAR 75,571 61,705 6,127 4,444 3,295 ….Legal Entities % 20.8 9.0 58.4 76.3 97.1 FOPs % 11.2 9.3 28.2 21.1 2.0 Individuals % 68.0 81.7 13.4 2.7 0.8 Ratio SAR/F29 2.09 3.82 1.13 0.62 0.44 Median farm size, SAR 6.65 4.92 74 .99 206.60 1307.9 Med. farm size, Form 29 68.57 24.00 77.09 241.00 1298.6 Avg. farm size SAR 154.0 9.75 88.70 283.11 2802.5 Avg. farm size Form 29 777.7 25.07 79.43 264.67 3393.0 Panel B: Area cult. (mn. ha) Total Form 29 27.85 0.40 0.43 1.89 25.14 ACUs % 75.4 79.8 79.5 75.9 75.3 AACs % 24.6 20.2 20.5 24.1 24.7 Total SAR 11.64 0.60 0.54 1.26 9.23 ….Legal Entities % 90.5 33.0 63.9 82.8 96.9 FOPs % 4.4 13.6 25.5 15.3 1.1 Individuals % 5.1 53.4 10.7 1.9 2.0 Ratio SAR/F29 0.42 1.49 1.27 0.67 0.37 Panel C: Informality (ACUs) Area cult. 2020 (mn. ha) 33.87 F29 overlapping (mn. ha) 6.42 0.09 0.20 0.67 5.46 F29 non-overlap (mn. ha) 14.59 0.23 0.14 0.76 13.45 SAR only (mn. ha) 3.14 0.43 0.24 0.28 2.20 Informal (mn. ha) 9.71 Panel D: Informality (AACs) Area cult. 2020 (mn. ha) 11.86 F29 overlapping (mn. ha) 1.09 0.01 0.02 0.08 1.00 F29 non-overlap (mn. ha) 5.75 0.08 0.07 0.38 5.23 SAR only (mn. ha) 0.22 0.02 0.03 0.05 0.12 Informal (mn. ha) 4.80 Source: Own computation from SAR, 2020 Form 29 data from SSSU and crop maps based on satellite imagery for 2020 as discussed in the text. Note: In panels C and D, ACU is areas controlled by Ukraine and AAC are areas affected by conflict. Region is not reported in the SAR for 1,629 farms cultivating a total of 39,735 ha. a Area cultivated 2020 is based on satellite imagery. 20 Table 2: Basic farm characteristics by region and farm size group Total Center North South East West <50 50-120 120-500 >500 Household characteristics Lives in villagea 0.70 0.70 0.62 0.71 0.68 0.76 0.77 0.71 0.64 0.64 Hh qnr applieda 0.82 0.85 0.74 0.84 0.81 0.78 0.94 0.91 0.75 0.52 Male farm owner 0.75 0.76 0.85 0.64 0.82 0.77 0.61 0.82 0.80 0.83 Age of owner 48.97 49.21 47.65 50.31 49.24 46.92 49.14 49.02 48.09 50.11 Education < secondary 0.01 0.01 0.01 0.02 0.00 0.02 0.02 0.01 0.00 0.00 Secondary education 0.10 0.09 0.10 0.13 0.16 0.07 0.14 0.09 0.09 0.01 Education > secondary 0.89 0.90 0.89 0.85 0.84 0.91 0.83 0.90 0.91 0.99 Owner has bank account 0.96 0.95 0.96 0.95 0.98 0.97 0.92 0.97 0.98 0.99 Other member w. bank acct 0.72 0.71 0.71 0.77 0.68 0.69 0.71 0.73 0.78 0.60 Owner has FB account 0.67 0.66 0.65 0.69 0.64 0.71 0.64 0.68 0.70 0.66 Other member w. FB acct 0.64 0.64 0.63 0.65 0.65 0.63 0.63 0.67 0.69 0.46 War impact & coping strategies War damage to landa 0.08 0.03 0.21 0.10 0.26 0.03 0.03 0.06 0.13 0.18 War damage to structuresa 0.04 0.01 0.11 0.05 0.15 0.01 0.02 0.02 0.06 0.09 Member migrated out in oblast 0.08 0.08 0.10 0.06 0.09 0.10 0.06 0.09 0.07 0.07 Member migrated beyond oblast 0.09 0.08 0.07 0.11 0.18 0.08 0.10 0.09 0.08 0.08 Could get informal support 0.46 0.45 0.44 0.48 0.48 0.48 0.46 0.50 0.44 0.32 if yes, amount (US$) 7,311 7,505 9,922 5,461 9,276 6,260 4,145 7,735 10,490 12,141 $ 200 lottery: Consumption 0.17 0.14 0.20 0.18 0.20 0.21 0.18 0.15 0.17 0.19 $ 200 lottery: Working capital 0.40 0.39 0.42 0.44 0.39 0.39 0.46 0.42 0.34 0.26 $ 200 lottery: Investment/savings 0.15 0.17 0.14 0.13 0.15 0.15 0.15 0.13 0.17 0.23 $ 5,000 lottery: Consumption 0.10 0.09 0.09 0.11 0.14 0.08 0.15 0.08 0.07 0.06 $ 5,000 lottery: Working capital 0.65 0.64 0.66 0.69 0.57 0.61 0.67 0.69 0.63 0.43 $ 5,000 lottery: Investment/saving 0.19 0.20 0.17 0.17 0.23 0.22 0.19 0.16 0.20 0.35 No. of obs. (farms) 2,251 1,059 316 464 124 288 619 790 501 341 Source: Own computation from MAPF/WB/EU small producer survey. Note: Answers marked by a were answered by all farms while the remainder is only for the 82% owner-operated farms that answered the household section of the questionnaire. 21 Table 3: Household composition, income sources & coping mechanisms by region and farm size group Total Center North South East West <50 50-120 120-500 >500 Panel A: 2021 Income sources/attitudes No. adult household members 2.60 2.58 2.53 2.61 2.43 2.82 2.61 2.63 2.69 2.28 …. # working full-time in agric. 1.45 1.49 1.36 1.41 1.43 1.51 1.43 1.51 1.61 1.09 …. # working part- time in agric. 0.62 0.61 0.67 0.59 0.48 0.67 0.83 0.67 0.44 0.30 Had permanent employees (y/n) 0.54 0.53 0.64 0.44 0.66 0.55 0.18 0.45 0.82 0.98 … if yes, how many? 9.02 7.68 14.92 3.39 13.28 14.94 1.05 1.15 4.56 49.34 Member w. non-agr. job (y/n) 0.94 0.96 0.94 0.90 0.95 0.94 0.87 0.96 0.98 0.98 if yes, avg. mthly income 253 255 269 240 284 237 217 256 269 314 Member self-emp. in non-ag (y/n) 0.62 0.64 0.69 0.57 0.72 0.57 0.34 0.71 0.83 0.8 if yes, avg. mthly income 288 286 297 282 316 281 246 284 293 349 Personal situation (1-10) 7.34 7.39 7.36 7.29 7.31 7.21 6.84 7.37 7.55 7.81 Country situation (1-10) 7.15 7.22 7.06 7.16 7.02 7.04 6.88 7.24 7.21 7.34 Coping mechanisms Member received social assistance 0.32 0.30 0.30 0.33 0.33 0.37 0.36 0.32 0.28 0.22 if yes, avg. monthly (US$) 69.0 72.7 64.6 67.4 61.2 66.1 64.5 70.3 76.9 65.5 Member received remittances 0.04 0.04 0.02 0.03 0.07 0.07 0.04 0.05 0.04 0.02 if yes, avg. monthly receipt 75.2 81.0 92.6 61.7 88.0 66.1 48.6 88.0 84.2 92.6 Panel B: 2022 Income sources/attitudes No. adult household members 2.59 2.57 2.52 2.61 2.41 2.80 2.59 2.62 2.69 2.27 …. # working full-time in agric. 1.47 1.50 1.44 1.41 1.37 1.55 1.47 1.55 1.58 1.07 …. # working part- time in agric. 0.63 0.64 0.56 0.61 0.52 0.76 0.89 0.71 0.45 0.18 Had permanent employees (y/n) 0.55 0.54 0.68 0.44 0.61 0.57 0.20 0.46 0.81 0.98 … if yes, how many? 8.78 6.99 14.63 3.27 11.16 16.90 0.96 1.16 4.53 48.01 Member w. non-agr. job (y/n) 0.93 0.94 0.93 0.89 0.93 0.94 0.85 0.95 0.98 0.98 if yes, avg. mthly income 232 234 254 207 246 233 198 222 256 311 Member self-emp. in non-ag (y/n) 0.62 0.64 0.69 0.56 0.7 0.58 0.34 0.71 0.8 0.82 if yes, avg. mthly income 269 270 281 238 296 282 229 254 279 349 Personal situation (1-10) 3.66 3.67 3.69 3.61 3.27 3.88 3.52 3.57 3.89 3.81 Country situation (1-10) 3.24 3.25 3.19 3.29 2.83 3.36 3.20 3.22 3.25 3.34 Coping mechanisms Member received social assistance 0.39 0.34 0.39 0.47 0.39 0.42 0.42 0.39 0.37 0.29 if yes, avg. monthly (US$) 66.4 69.6 60.5 62.1 54.7 75.2 63.0 69.4 70.8 53.4 Member received remittances 0.05 0.05 0.02 0.06 0.04 0.05 0.05 0.04 0.06 0.03 if yes, avg. monthly receipt 74.1 76.5 84.9 61.0 92.6 84.2 54.3 89.2 77.2 86.8 No. of obs. (farms) 2,251 1,059 316 464 124 288 619 790 501 341 Source: Own computation from MAPF/WB/EU small producer survey. 22 Table 4: Factor market participation by region and farm size group By region By farm size Total Center North South East West <50 50-120 120-500 > 500 Land market Leases land 0.84 0.85 0.84 0.82 0.84 0.84 0.72 0.87 0.87 0.85 if yes, lease paid ($/ha) 105.34 110.24 107.16 92.54 96.65 107.96 85.62 105.16 108.38 117.02 Wants to lease more (y/n) 0.85 0.87 0.86 0.80 0.84 0.84 0.71 0.89 0.89 0.94 Wants to buy land (y/n) 0.78 0.80 0.82 0.70 0.82 0.76 0.56 0.86 0.85 0.89 if yes, price ($/ha) 1,432 1,483 1,355 1,340 1,353 1,512 1,353 1,440 1,413 1,548 if yes, med. price ($/ha) 1,111 1,388 1,111 1,111 1,111 1,111 1,111 1,250 1,111 1,389 Wants to sell land (y/n) 0.03 0.03 0.04 0.03 0.05 0.05 0.06 0.02 0.03 0.00 if yes, selling price ($/ha) 3,142 4,627 1,369 1,728 1,852 3,254 4,787 1,848 1,206 if yes, med. Price ($/ha) 1,111 1,389 1,111 1,111 1,111 1,111 2,500 1,806 1,111 Assets Area cultivated (ha) 418.06 339.87 822.50 289.66 467.01 445.07 20.36 86.68 256.49 2,118.21 Own tractor/equip 0.84 0.84 0.84 0.86 0.78 0.86 0.82 0.88 0.88 0.72 if yes, value ($) 187,991 183,610 268,727 149,566 198,787 167,105 21,033 64,220 179,167 785,646 .. if yes, value/ha ($) 774.79 787.09 835.41 551.16 705.56 1023.28 1212.75 672.42 680.82 609.29 Credit markets Ever received credit 0.55 0.56 0.57 0.51 0.58 0.54 0.34 0.49 0.69 0.84 if yes, years since last 2.29 2.28 1.96 2.82 2.01 2.07 4.78 2.85 1.57 0.78 size (US$) 75,709 62,469 149,683 45,655 74,600 89,671 6,946 17,547 53,253 253,041 interest rate 7.59 7.71 7.36 7.28 9.69 6.96 11.47 8.31 7.04 5.44 used for working capital 0.94 0.93 0.95 0.94 0.92 0.91 0.79 0.95 0.97 0.98 Would want to borrow 0.78 0.79 0.83 0.72 0.78 0.81 0.62 0.84 0.84 0.89 if yes, amount 140,530 116,669 188,375 130,829 148,939 184,438 34,508 74,299 128,475 446,956 interest rate 6.08 6.18 6.76 5.32 6.55 5.84 5.20 5.89 6.64 6.75 Credit constrained? 0.54 0.54 0.55 0.53 0.62 0.50 0.56 0.67 0.46 0.33 Support received/needed Got advisory service (y/n) 0.18 0.19 0.20 0.14 0.13 0.20 0.14 0.19 0.19 0.22 if yes, from Gov’t 0.49 0.50 0.49 0.44 0.50 0.49 0.41 0.48 0.53 0.53 Would want more TA 0.91 0.90 0.92 0.92 0.94 0.93 0.88 0.93 0.90 0.93 Priorities for gov’t to … regulate input prices 0.29 0.30 0.24 0.31 0.24 0.30 0.34 0.28 0.26 0.25 … procurement/markets 0.25 0.27 0.26 0.24 0.24 0.23 0.24 0.26 0.25 0.26 … provide credit 0.15 0.14 0.17 0.15 0.14 0.15 0.12 0.16 0.16 0.16 … tax reduction 0.12 0.11 0.12 0.13 0.14 0.10 0.11 0.10 0.12 0.15 No. of obs. (farms) 2,251 1,059 336 464 124 268 619 790 501 341 Source: Own computation from MAPF/WB/EU small producer survey. Note: As discussed in the text, a farm is considered credit constrained in 2022 if it indicated that it would want to borrow but did not have access to credit in this year. Area cultivated at the farm level reported for 2022. 23 Table 5: Area allocation and changes in aggregate crop mix Total Center North South East West <50 50-120 120-500 >500 Panel A: 2021 Average area cult. (ha) 346.46 269.29 676.69 248.54 438.45 410.76 21.52 77.93 225.19 1713.87 % of farmers cultivating … wheat (ha) 0.70 0.67 0.58 0.76 0.81 0.77 0.53 0.71 0.78 0.81 … barley (ha) 0.30 0.28 0.15 0.46 0.33 0.24 0.22 0.29 0.33 0.41 … sunflower (ha) 0.59 0.64 0.55 0.70 0.71 0.18 0.38 0.60 0.68 0.76 … maize (ha) 0.46 0.56 0.66 0.20 0.42 0.33 0.25 0.46 0.52 0.72 … soybean (ha) 0.24 0.27 0.28 0.01 0.12 0.59 0.16 0.28 0.25 0.30 … rapeseed (ha) 0.10 0.08 0.07 0.11 0.01 0.23 0.00 0.05 0.12 0.33 … other (ha) 0.19 0.14 0.28 0.11 0.18 0.39 0.13 0.16 0.22 0.29 Total area grown with … wheat (ha) 176,082 61,484 29,619 42,179 19,128 23,672 3,029 15,569 31,615 125,869 … barley (ha) 34,175 12,326 2,533 14,496 2,025 2,795 986 3,595 7,405 22,189 … sunflower (ha) 171,289 74,257 37,624 29,469 17,483 12,456 2,945 14,364 27,197 126,783 … maize (ha) 172,414 68,517 71,189 4,752 6,714 21,242 1,651 11,401 20,846 138,516 … soybean (ha) 48,211 13,042 18,631 105 2,098 14,335 1,088 5,487 6,984 34,652 … rapeseed (ha) 33,570 7,913 4,490 7,830 490 12,847 0 766 3,997 28,807 … other (ha) 29,117 8,858 11,853 2,574 1,168 4,664 544 2,824 5,542 20,207 Panel B: 2022 Average area cult. (ha) 370.70 301.35 716.61 261.69 400.48 436.18 22.60 83.67 242.02 1832.09 % of farmers cultivating … wheat (ha) 0.75 0.74 0.60 0.82 0.89 0.79 0.56 0.78 0.83 0.88 … barley (ha) 0.32 0.32 0.15 0.49 0.32 0.26 0.22 0.30 0.38 0.45 … sunflower (ha) 0.65 0.70 0.66 0.75 0.71 0.21 0.40 0.69 0.75 0.80 … maize (ha) 0.48 0.60 0.63 0.22 0.38 0.39 0.25 0.50 0.55 0.73 … soybean (ha) 0.30 0.34 0.37 0.01 0.12 0.64 0.19 0.33 0.30 0.38 … rapeseed (ha) 0.12 0.10 0.10 0.13 0.02 0.27 0.01 0.07 0.15 0.38 … other (ha) 0.21 0.16 0.32 0.13 0.21 0.42 0.15 0.18 0.25 0.31 Area/farm grown with … wheat (ha) 183,742 65,695 29,231 40,966 23,887 23,963 3,072 16,116 32,644 131,910 … barley (ha) 33,119 12,581 2,416 13,855 1,376 2,891 863 3,691 6,922 21,643 … sunflower (ha) 201,675 88,306 59,642 30,983 10,955 11,789 3,096 16,282 31,800 150,497 … maize (ha) 154,938 70,159 53,187 5,693 5,660 20,239 1,607 10,481 18,841 124,009 … soybean (ha) 68,504 19,754 24,823 477 947 22,503 1,502 6,750 9,238 51,014 … rapeseed (ha) 41,068 11,366 5,977 11,295 983 11,447 70 1,354 4,677 34,967 … other (ha) 28,331 7,870 11,042 3,501 1,046 4,872 549 3,310 7,207 17,265 No. of obs. (farms) 1,919 915 260 408 112 224 476 693 460 290 Source: Own computation from MAPF/WB/EU small producer survey Note: Data are for the balanced panel of farms that report cultivating any crop in both the 2021 and 2022 agricultural seasons. 24 Table 6: Output and input prices and cash flow Total Center North South East West <50 50-120 120-500 >500 Panel A: 2021 Output value (US$/ha) 872.0 903.7 815.1 1,051.6 661.3 962.4 704.1 866.3 907.2 1,084.6 Input cost (US$/ha) 154.03 153.04 153.80 200.92 124.44 166.98 110.96 140.36 189.67 261.76 Net revenue (US$/ha) 670.4 706.3 656.5 790.0 506.2 735.1 563.7 710.9 692.5 778.2 Sold anything in market 0.94 0.94 0.97 0.95 0.96 0.90 0.84 0.96 0.98 1.00 … if yes, % of value 0.83 0.84 0.85 0.81 0.83 0.80 0.83 0.85 0.85 0.77 Cash flow (US$/ha) 432.2 487.2 376.1 532.9 331.7 276.1 401.5 431.3 442.6 481.6 % w. negative cash flow 0.10 0.09 0.09 0.15 0.09 0.17 0.12 0.08 0.11 0.13 Panel B: 2022 Output value (US$/ha) 571.3 595.7 498.1 680.2 384.5 738.6 471.4 565.8 591.3 704.4 … at 2021 price ($/ha) 704.8 735.1 688.9 864.8 468.2 855.9 576.2 698.9 732.3 870.0 Input cost (US$/ha) 183.61 187.02 171.19 226.47 153.76 187.36 149.05 169.51 222.21 260.91 … real input value ($/ha) 94.88 94.04 81.83 124.50 82.42 97.36 77.08 87.12 115.63 134.75 Net revenue (US$/ha) 345.9 363.5 311.5 417.8 216.6 496.0 294.2 379.6 345.3 368.0 Sold anything in market 0.77 0.78 0.76 0.75 0.80 0.71 0.58 0.80 0.85 0.89 … if yes, % of value 0.57 0.57 0.57 0.58 0.59 0.50 0.69 0.55 0.57 0.51 Cash flow (US$/ha) -25.6 -30.3 -18.9 -36.2 -6.6 -38.3 89.1 8.1 -99.5 -188.6 % w. negative cash flow 0.46 0.43 0.46 0.51 0.47 0.51 0.40 0.43 0.50 0.56 No. of obs. (farms) 1,714 852 213 367 89 192 413 632 419 250 Source: Own computation from MAPF/WB/EU small producer survey Note: Cash flow is defined as the value of output sold minus cost of purchased inputs 25 Table 7: Regression (area/yield etc.) for main crops Area Output value Market Share sold Input Net revenue Cash flow Negative cash Share of area (US$/ha) participation flow with oilseeds Central # 2022 0.032 -0.237*** -0.096*** -0.318*** -154.089*** -425.184*** -680.161*** 0.460*** 0.044 (0.051) (0.034) (0.032) (0.037) (24.701) (45.034) (139.359) (0.076) (0.028) East # 2022 -0.073 -0.170** -0.158** -0.371*** -159.596*** -422.076*** -630.566*** 0.531*** 0.025 (0.069) (0.069) (0.074) (0.069) (39.537) (66.861) (149.905) (0.123) (0.047) North # 2022 -0.036 -0.192*** -0.150*** -0.311*** -183.705*** -451.522*** -819.191*** 0.478*** 0.113*** (0.068) (0.047) (0.047) (0.053) (32.468) (61.006) (152.796) (0.103) (0.043) South # 2022 -0.026 -0.375*** -0.085** -0.322*** -155.485*** -380.965*** -556.685*** 0.513*** -0.005 (0.054) (0.044) (0.041) (0.047) (25.644) (46.814) (130.939) (0.087) (0.036) West # 2022 0.079 -0.144*** -0.124** -0.316*** -178.273*** -356.133*** -616.454*** 0.462*** 0.036 (0.067) (0.040) (0.050) (0.052) (32.230) (54.478) (184.906) (0.105) (0.038) < 50 ha # 2022 0.054 0.065 -0.153*** 0.035 111.805*** 117.286** 271.982* -0.092 -0.002 (0.057) (0.046) (0.047) (0.048) (26.207) (49.686) (138.623) (0.086) (0.042) 50-120 ha # 2022 0.066 0.016 -0.056 -0.049 78.871*** 69.108 169.584 -0.070 0.013 (0.054) (0.037) (0.037) (0.041) (26.275) (46.584) (136.112) (0.081) (0.030) 120-500 ha # 2022 0.051 0.026 -0.021 -0.014 70.529** 46.797 60.736 -0.007 0.043 (0.057) (0.039) (0.039) (0.044) (27.617) (49.472) (145.668) (0.087) (0.032) Structure or land -0.124** -0.111** 0.007 0.015 -11.703 -45.517 -34.700 0.082 -0.026 affected by war (0.062) (0.054) (0.052) (0.054) (23.629) (45.629) (96.038) (0.089) (0.039) Constant 4.376*** 6.682*** 0.942*** 0.767*** 249.480*** 569.715*** 402.366*** 0.140*** 0.390*** (0.008) (0.007) (0.007) (0.007) (3.094) (6.481) (12.908) (0.012) (0.006) No. of farms 3,428 3,358 3,288 3,006 2,710 2,460 2,168 2,172 3,422 R-squared 0.977 0.826 0.657 0.710 0.813 0.803 0.712 0.677 0.678 Note: Main crops include wheat, barley, rapeseed, maize, sunflower and soybean. The dependent variables in cols. 1 and 2 are in natural logs. The dependent variable in col. 2 is real farm output of these crops at the 2021 prices per ha of cultivated land. The dependent variable in col. 5 is the real value of hired labor and machinery services as well as purchased material inputs per ha deflated at 2021 regional level fertilizer prices. The dependent variable in col. 6 is nominal value of output minus value of hired labor and machinery services and purchased material inputs per ha. The dependent variable in col. 9 is the share of the farm’s cultivated area devoted to oilseeds. All regressions include farm fixed effects. The sample is restricted to the balanced panel based on valid observations the dependent variable in the respective column. Robust standard errors clustered at farm level in parentheses: * significant at 10%, ** significant at 5% and *** significant at 1%. 26 Table 8: Production function regressions 2031 2022 2021 and 2022 combined OLS Farm fixed effects Land (ha) 0.081*** 0.114*** 0.080*** 0.114*** 0.077*** 0.095*** -0.031 0.014 (0.009) (0.012) (0.011) (0.014) (0.008) (0.010) (0.032) (0.039) Affected by war -0.111** -0.122** -0.098* -0.114** -0.129** -0.122** (0.055) (0.053) (0.055) (0.055) (0.050) (0.051) Labor (US$/ha) 0.063*** 0.068*** 0.034** 0.019 (0.016) (0.019) (0.014) (0.033) Fertilizer (US$/ha) 0.081*** 0.084*** 0.069*** 0.050** (0.017) (0.018) (0.015) (0.021) Machinery (US$/ha) 0.065*** 0.055*** 0.056*** 0.029 (0.015) (0.018) (0.017) (0.022) Central # Year 2022 -0.184*** -0.137*** -0.232*** -0.197*** (0.020) (0.021) (0.018) (0.023) East # Year 2022 -0.196*** -0.118*** -0.152*** -0.114** (0.043) (0.044) (0.055) (0.058) North # Year 2022 0.010 0.031 -0.120*** -0.088** (0.042) (0.042) (0.039) (0.043) Southern # Year 2022 -0.602*** -0.486*** -0.332*** -0.285*** (0.031) (0.034) (0.031) (0.036) West # Year 2022 0.007 0.042 -0.123*** -0.072* (0.049) (0.050) (0.037) (0.039) Constant 6.322*** 5.165*** 6.092*** 5.012*** 6.341*** 5.497*** 6.803*** 6.134*** (0.043) (0.136) (0.049) (0.146) (0.040) (0.118) (0.138) (0.332) No. of farms 954 954 959 959 1,913 1,913 1,913 1,913 R-squared 0.073 0.170 0.055 0.133 0.221 0.270 0.278 0.289 Note: Regressions are for monetary yield from main crops (wheat, barley, rapeseed, maize, sunflower and soybean) per hectare using the balanced panel of farms cultivating any crop in the 2021 and 2022 cropping season. Robust standard errors clustered at the farm level in parentheses: * significant at 10%, ** significant at 5% and *** significant at 1%. All regressions include farm fixed effects. 27 Table 9: Regression of farm TFP on farm size, formality, and credit constrained status Farm size <50 ha 0.908*** 0.909*** 0.884*** (0.016) (0.019) (0.026) Farm size 50-120 ha 1.043*** 1.046*** 1.086*** (0.012) (0.020) (0.027) Farm size 120-500 ha 1.085*** 1.087*** 1.076*** (0.015) (0.021) (0.037) Farm size >=500 ha 1.155*** 1.162*** 1.126*** (0.023) (0.030) (0.109) Legal entity 0.003 (0.017) Credit constrained -0.005 (0.016) Farm size <50 ha * Legal entity 0.056 (0.038) Farm size 50-120 ha * Legal entity -0.010 (0.025) Farm size 120-500 ha * Legal entity -0.011 (0.036) Farm size >=500 ha * Legal entity 0.052 (0.108) Farm size <50 ha * Credit constrained 0.017 (0.032) Farm size 50-120 ha * Credit constrained -0.053* (0.027) Farm size 120-500 ha * Credit constrained 0.040 (0.033) Farm size >=500 ha * Credit constrained -0.042 (0.052) No. of farms 941 918 918 R-squared 0.952 0.952 0.953 Note: Dependent variable is farm TFP (i.e., exponential of farm fixed effects extracted from col. If of table 8 after purging village fixed effects). A constant term is not included in all the specifications. The equality of farm size coefficients in cols. 1 and 2 between any category and the greater than 500 ha category is rejected at 1% level. Standard errors clustered at the farm level in parentheses: * significant at 10%, ** significant at 5% and *** significant at 1%. 28 Figure 1: Geographical distribution of sample 29 Figure 2: Plot of farm-level productivity against farm size Note: Farm fixed effect (net of rayon fixed effects) is extracted from the regression of output/ha in col. 8 of table 8. 30 APPENDIX TABLES: Appendix table 1: Response rate by farm size Total Farm size category < 50 50-120 120-500 > 500 Sampled farms 11,756 2,383 4,609 2,743 2,021 Farms contacted 10,982 2,302 4,174 2,626 1,880 No response (%) 45.76 50.26 47.22 43.18 40.59 Refusal (%) 33.76 27.89 30.43 37.43 43.19 Interviewed (%) 20.49 21.85 22.35 19.38 16.22 … 1st attempt (%) 66.13 68.39 67.52 68.17 54.75 … 2nd attempt (%) 22.22 20.48 22.51 22.20 24.26 > 2 attempts (%) 11.64 11.13 9.97 9.63 20.98 Note: Categories coded as non-response include ‘cellphone out of range/not reachable’ (4,536 cases), ‘no answer’, ‘enterprise closed or did not operate in 2022’, ‘respondent moved abroad’, ‘non-response due to security’, and ‘not relevant contact’. 31 Appendix table 2: Prices for key farm inputs, 2021 and 2022 National Regional Center North South East West Panel A: 2021 Fertilizer price (US$/50kg) 15.57 15.45 15.28 15.91 15.28 15.94 Diesel (US$/liter) 0.78 0.78 0.78 0.78 0.79 0.79 Machinery services (US$/ha) 37.52 35.42 31.85 30.56 28.65 61.15 Daily wage (US$/day) 11.55 13.89 12.50 7.98 12.36 11.03 Panel B: 2022 Fertilizer price (US$/50kg) 40.17 38.90 40.28 40.42 38.19 43.06 Diesel (US$/liter) 1.49 1.48 1.49 1.49 1.50 1.48 Machinery services (US$/ha) 49.23 45.30 39.18 42.03 39.76 79.88 Daily wage (US$/day) 14.38 15.30 12.50 13.89 16.32 13.89 Source: Survey of community leaders and local dealers’ adverts, complemented by completed auctions for machinery services on the Prozorro sales platform. Prices are for ammonium nitrate and urea for fertilizer and the cost of harvesting and ploughing/soil preparation for wheat and maize, respectively, for machinery services. Wages are for agricultural labor. 32 Appendix table 3: Production outcomes by crop Total Center North South East West <50 50-120 120-500 >500 Panel A: 2021 Produced wheat 0.63 0.61 0.77 0.52 0.70 0.68 0.48 0.64 0.74 0.72 Area (ha) 103.30 84.82 128.44 144.98 108.76 112.94 10.88 30.62 84.10 385.56 Yield of wheat (t/ha) 4.28 4.42 4.16 4.78 3.52 4.75 3.72 4.13 4.38 4.92 Price of wheat ($/t) 170.59 167.11 162.97 180.77 171.93 177.52 146.77 160.46 178.31 192.92 Sold wheat (y/n) 0.90 0.90 0.95 0.90 0.94 0.79 0.72 0.92 0.95 0.95 if yes, share of harvest sold 0.87 0.88 0.93 0.86 0.86 0.87 0.87 0.89 0.89 0.82 Input cost 158.47 155.75 139.60 191.06 149.02 170.96 117.65 148.07 190.17 203.09 Net revenue 535.34 552.43 551.06 652.23 423.51 584.53 485.74 543.01 529.06 609.72 Produced barley 0.27 0.26 0.30 0.13 0.41 0.23 0.19 0.26 0.31 0.36 Area (ha) 49.56 41.43 41.36 56.25 63.02 47.15 8.12 16.61 49.40 146.72 Yield of barley (t/ha) 3.76 3.87 3.15 4.09 3.50 4.16 3.05 3.58 3.81 4.59 Price of barley ($/t) 145.03 140.50 131.94 162.22 145.87 163.10 120.58 133.04 156.57 162.53 Sold barley (y/n) 0.84 0.85 0.73 0.64 0.94 0.67 0.56 0.90 0.91 0.87 if yes, share of harvest sold 0.87 0.88 0.80 0.74 0.87 0.92 0.82 0.85 0.89 0.90 Input cost 136.11 134.59 107.14 121.97 136.89 171.50 107.43 118.10 149.56 225.21 Net revenue 388.35 408.88 330.44 424.04 342.71 489.83 340.92 398.86 393.89 425.80 Produced sunflower 0.52 0.57 0.65 0.47 0.63 0.14 0.30 0.54 0.63 0.68 Area (ha) 108.99 90.84 128.79 156.97 96.40 313.26 17.38 33.99 83.85 376.57 Yield of sunflower (t/ha) 2.38 2.41 2.32 2.79 2.08 2.78 2.06 2.34 2.39 2.69 Price of sunflower ($/t) 441.06 442.19 409.51 452.18 433.96 512.88 396.04 415.63 455.21 494.12 Sold sunflower (y/n) 0.95 0.92 0.96 0.98 0.98 1.00 0.93 0.95 0.97 0.93 if yes, share of harvest sold 0.92 0.93 0.95 0.92 0.90 1.00 0.94 0.92 0.95 0.86 Input cost 154.06 154.17 141.53 203.61 128.81 230.93 120.98 122.62 180.71 242.52 Net revenue 845.63 868.25 823.78 1,036.91 697.61 1,094.86 751.97 850.55 825.38 971.54 Produced maize 0.41 0.50 0.38 0.55 0.19 0.27 0.21 0.41 0.49 0.64 Area (ha) 141.37 108.72 101.69 240.88 66.41 289.24 14.39 31.84 80.93 428.69 Yield of maize (t/ha) 7.28 7.25 5.27 8.17 6.10 8.27 5.93 7.14 7.24 8.03 Price of maize ($/t) 162.65 162.40 138.54 172.33 151.45 169.52 144.28 151.80 162.64 182.57 Sold maize (y/n) 0.92 0.92 0.75 0.93 0.92 0.93 0.82 0.92 0.92 0.95 if yes, share of harvest sold 0.88 0.90 0.86 0.86 0.90 0.82 0.90 0.92 0.90 0.81 Input cost 219.39 196.77 219.53 304.47 153.45 308.64 147.83 169.47 270.32 314.82 Net revenue 915.36 920.93 613.87 1,115.42 691.44 992.01 763.55 984.60 862.97 955.90 Produced soybeans 0.22 0.24 0.10 0.25 0.01 0.49 0.14 0.25 0.24 0.27 Area (ha) 74.46 49.56 97.67 114.40 58.33 99.42 14.39 29.57 60.19 259.08 Yield of soybeans (t/ha) 2.44 2.40 2.43 2.36 1.58 2.58 2.36 2.40 2.38 2.70 Price of soybeans ($/t) 389.49 378.46 403.22 378.46 277.78 415.61 340.51 374.48 400.24 434.65 Sold soybeans (y/n) 0.89 0.89 1.00 0.91 0.50 0.90 0.94 0.89 0.92 0.83 if yes, share of harvest sold 0.90 0.91 0.97 0.84 1.00 0.93 0.93 0.90 0.93 0.87 Input cost 135.44 113.75 270.92 179.40 172.22 145.72 96.27 121.38 123.31 272.00 Net revenue 749.15 736.24 607.50 636.79 179.63 868.51 742.68 750.03 751.75 752.68 Produced rapeseed 0.09 0.07 0.01 0.06 0.10 0.18 0.00 0.05 0.12 0.29 Area (ha) 175.47 110.32 490.00 236.32 192.24 218.93 21.78 81.49 282.52 Yield of rapeseed (t/ha) 2.74 2.65 2.60 3.27 2.45 2.79 2.55 2.57 2.89 Price of rapeseed ($/t) 465.56 445.44 476.36 488.30 471.56 476.22 446.54 448.92 484.81 Sold rapeseed (y/n) 0.99 0.98 1.00 1.00 0.95 1.00 1.00 0.97 0.99 if yes, share of harvest sold 0.98 0.99 0.98 0.97 0.97 1.00 0.99 0.96 Input cost 285.24 227.42 589.57 342.48 254.00 333.55 301.72 285.99 272.35 Net revenue 946.34 890.36 650.72 1,054.44 804.10 1,065.59 924.06 890.02 1,005.38 Source: Own computation from MAPF/WB/EU small producer survey 33 Appendix table 3 (cont.): Production outcomes by crop Total Center North South East West <50 50-120 120-500 >500 Panel B: 2022 Produced wheat 0.73 0.71 0.88 0.58 0.82 0.77 0.56 0.76 0.82 0.86 Area (ha) 100.93 85.31 121.86 150.30 97.36 111.30 10.59 30.88 81.84 375.12 Yield of wheat (t/ha) 3.52 3.56 3.64 4.09 2.62 4.48 3.05 3.41 3.52 4.17 Price of wheat ($/t) 128.94 122.76 102.81 132.93 134.69 152.76 106.63 115.93 136.31 156.48 Sold wheat (y/n) 0.57 0.57 0.56 0.53 0.65 0.49 0.40 0.57 0.61 0.69 if yes, share of harvest sold 0.66 0.66 0.67 0.67 0.69 0.59 0.76 0.65 0.69 0.61 Input cost 204.22 208.56 160.14 230.12 174.90 250.99 170.69 192.53 230.06 251.52 Net revenue 214.69 201.81 213.38 251.10 157.83 357.85 215.48 219.71 191.94 242.39 Produced barley 0.31 0.31 0.31 0.14 0.46 0.26 0.22 0.29 0.37 0.43 Area (ha) 45.81 38.02 35.48 62.50 55.95 45.98 8.18 16.13 43.08 137.44 Yield of barley (t/ha) 3.20 3.38 3.09 3.73 2.57 3.99 2.76 3.15 3.05 3.86 Price of barley ($/t) 113.39 106.66 103.13 131.49 111.95 153.36 90.66 102.25 116.67 141.95 Sold barley (y/n) 0.54 0.52 0.45 0.54 0.64 0.38 0.37 0.56 0.61 0.53 if yes, share of harvest sold 0.79 0.77 0.61 0.95 0.81 0.73 0.80 0.76 0.82 0.77 Input cost 155.53 154.74 145.56 161.83 151.60 180.41 130.68 140.57 172.54 216.68 Net revenue 178.07 190.85 92.66 257.58 129.02 342.38 177.30 186.22 169.18 171.46 Produced sunflower 0.63 0.68 0.70 0.64 0.73 0.20 0.36 0.68 0.75 0.79 Area (ha) 117.70 98.24 136.13 181.67 94.11 376.43 16.70 33.77 92.51 411.40 Yield of sunflower (t/ha) 1.87 1.98 2.02 2.04 1.42 2.34 1.53 1.86 1.87 2.08 Price of sunflower ($/t) 364.79 365.24 330.75 366.53 372.06 382.52 345.50 354.37 368.11 391.48 Sold sunflower (y/n) 0.68 0.71 0.71 0.67 0.61 0.70 0.68 0.63 0.72 0.72 if yes, share of harvest sold 0.78 0.76 0.76 0.77 0.84 0.91 0.92 0.81 0.73 0.71 Input cost 174.00 176.82 140.91 253.05 139.38 200.66 127.59 151.49 205.44 239.46 Net revenue 460.79 507.08 513.39 447.60 346.61 591.82 396.21 488.39 429.72 501.97 Produced maize 0.47 0.59 0.38 0.61 0.22 0.36 0.26 0.48 0.55 0.73 Area (ha) 134.22 110.15 83.00 191.20 75.90 283.28 14.24 31.28 72.63 409.75 Yield of maize (t/ha) 5.78 5.64 5.46 7.09 3.27 7.38 4.66 5.52 6.01 6.34 Price of maize ($/t) 136.17 127.36 119.10 145.53 150.85 156.93 115.33 123.28 137.99 157.66 Sold maize (y/n) 0.66 0.65 0.58 0.65 0.75 0.66 0.59 0.63 0.69 0.69 if yes, share of harvest sold 0.79 0.81 0.81 0.74 0.79 0.79 0.87 0.82 0.77 0.75 Input cost 280.61 261.09 360.24 382.89 154.78 340.43 205.42 218.81 354.99 372.05 Net revenue 449.63 433.18 179.41 606.31 320.97 627.86 382.77 516.51 364.13 472.58 Produced soybeans 0.29 0.32 0.11 0.35 0.02 0.60 0.18 0.32 0.31 0.38 Area (ha) 83.76 56.99 113.50 94.95 80.00 126.71 14.54 30.91 69.77 295.85 Yield of soybeans (t/ha) 2.05 1.96 2.32 2.16 1.47 2.17 1.85 1.90 2.20 2.39 Price of soybeans ($/t) 333.14 331.43 347.06 346.63 333.33 326.94 300.93 319.97 341.12 368.82 Sold soybeans (y/n) 0.60 0.60 0.83 0.61 0.50 0.57 0.49 0.59 0.65 0.63 if yes, share of harvest sold 0.77 0.81 0.85 0.76 0.01 0.70 0.91 0.78 0.74 0.69 Input cost 176.69 158.96 250.32 222.17 177.08 182.00 156.94 169.59 169.69 247.11 Net revenue 471.76 451.33 463.97 430.47 281.25 545.76 438.46 457.71 510.40 519.12 Produced rapeseed 0.12 0.10 0.02 0.09 0.14 0.23 0.01 0.07 0.15 0.38 Area (ha) 181.63 126.59 935.00 283.11 184.05 189.95 22.41 78.54 295.59 Yield of rapeseed (t/ha) 2.76 2.68 4.81 3.30 1.86 3.05 2.47 2.41 3.04 Price of rapeseed ($/t) 391.44 369.30 473.17 419.13 382.02 407.03 350.87 356.64 432.03 Sold rapeseed (y/n) 0.88 0.82 1.00 0.95 0.86 0.95 0.89 0.97 0.84 if yes, share of harvest sold 0.93 0.95 0.88 0.92 0.92 0.96 0.94 0.90 Input cost 275.45 232.79 143.52 253.93 310.66 313.38 262.90 331.12 245.09 Net revenue 723.33 641.72 1,942.96 775.82 390.07 934.77 759.87 507.48 854.81 Source: Own computation from MAPF/WB/EU small producer survey 34 Appendix table 4: Distribution of farm TFP by farm size and ownership type Size group Total Legally registered farms Informal farms TFP No. TFP No. TFP No. Mean CV obs. Mean CV obs. Mean CV obs. <50 ha 0.908 0.266 226 0.951 0.255 49 0.896 0.268 177 50-120 ha 1.043 0.218 375 1.045 0.216 183 1.042 0.221 192 120-500 ha 1.085 0.225 240 1.078 0.221 166 1.100 0.234 74 >=500 ha 1.155 0.184 100 1.158 0.187 95 Total 1.033 0.238 941 1.068 0.221 493 0.995 0.253 448 Note: As discussed in the text, informal farms are either operations run by individuals or by FOPs. For legally registered farms, mean farm TFP in the less than 50 ha, 50-120 ha and 120-500 ha category is statistically different from the mean TFP in the greater than 500 ha category at 1%. For informal farms, mean farm TFP in the less than 50 ha category is statistically different from that of 120-500 ha category at 1% while the equality of mean TFP between 50-120 ha and the 120-500 ha category can only be rejected at 10%. 35 References: Akresh, R., Bhalotra, S., Leone, M., Osili, U., 2023. First- and Second-Generation Impacts of the Biafran War. Journal of Human Resources 58, 1-45 Akresh, R., Caruso, G.D., Thirumurthy, H., 2022. Detailed geographic information, conflict exposure, and health impacts. World Development 155, 105890 Akresh, R., Verwimp, P., Bundervoet, T., 2011. Civil War, Crop Failure, and Child Stunting in Rwanda. Economic Development and Cultural Change 59, 777-810 Amodio, F., Di Maio, M., 2018. Making Do with What You Have: Conflict, Input Misallocation and Firm Performance. Economic Journal 128, 2559-2612 Arcand, J.-L., Rodella-Boitreaud, A.-S., Rieger, M., 2015. The Impact of Land Mines on Child Health: Evidence from Angola. Economic Development and Cultural Change 63, 249-279 Arias, M.A., Ibáñez, A.M., Zambrano, A., 2019. Agricultural production amid conflict: Separating the effects of conflict into shocks and uncertainty. World Development 119, 165-184 Balma, L., Heidland, T., Jävervall, S., Mahlkow, H., Mukasa, A.N., Woldemichael, A., 2022. Long-Run Impacts of the Conflict in Ukraine on Food Security in Africa. Kiel Institute for the World Economy, Kiel Bauer, M., Blattman, C., Chytilova, J., Henrich, J., Miguel, E., Mitts, T., 2016. Can War Foster Cooperation? Journal of Economic Perspectives 30, 249-274 Ben Hassen, T., El Bilali, H., 2022. Impacts of the Russia-Ukraine War on Global Food Security: Towards More Sustainable and Resilient Food Systems? Foods 11 Benhassine, N., McKenzie, D., Pouliquen, V., Santini, M., 2018. Does inducing informal firms to formalize make sense? Experimental evidence from Benin. Journal of Public Economics 157, 1-14 Bruhn, M., McKenzie, D., 2014. Entry Regulation and the Formalization of Microenterprises in Developing Countries. World Bank Research Observer 29, 186-201 Büttner, N., Grimm, M., Soubeiga, S., 2022. Political instability and households’ investment behavior: Evidence from Burkina Faso. Journal of Comparative Economics 50, 350-368 Capasso, S., Ohnsorge, F., Yu, S., 2022. Informality and Financial Development: A Literature Review. Manchester School 90, 587-608 Couttenier, M., Monnet, N., Piemontese, L., 2022. The Economic Costs of Conflict: A Production Network Approach. Centre for Economic Policy Research, Discussion Paper DP16984, London de Roux, N., Martinez, L.R., 2021. Forgone Investment: Civil Conflict and Agricultural Credit in Colombia. Johns Hopkins University, Washington DC Deininger, K., Ali, D., 2023. Land and mortgage markets in Ukraine: Pre-war performance, war effects, and implications for recovery. World Bank Policy Research Working Paper, Washington DC Deininger, K., Ali, D.A., Kussul, N., Shelestov, A., Lemoine, G., Yailimova, H., 2023. Quantifying war-induced crop losses in Ukraine in near real time to strengthen local and global food security. Food Policy 115, 102418 Deininger, K., Ali, D.A., Neyter, R., 2022. Impacts of Transparent Online Auctions on Public Land Lease Revenue: Evidence from Legal and Administrative Changes in Ukraine. World Bank Policy Research Working Paper 10201, Washington DC Deininger, K., Nizalov, D., Singh, S.K., 2018. Determinants of Productivity and Structural Change in a Large Commercial Farm Environment: Evidence from Ukraine. The World Bank Economic Review 32, 287-306 Duncan, E., Skakun, S., Becker-Reshef, I., 2023. Identification and mapping of missiles craters and unexploded ordnance in agricultural fields in Ukraine using very high spatial resolution satellite imagery. In: Paper Presented at the 9th International Conference on Agricultural Statistics. World Bank, Washington DC FAO, 2022a. The importance of Ukraine and the Russian Federation for global agricultural markets and the risks associated with the war in Ukraine. In: Information Note, Dec. 5, 2022 update. Food and Agriculture Organization of the United Nations, Rome FAO, 2022b. Note on the impact of the war on food security in Ukraine. Food and Agriculture Organization of the United Nations, Rome FAO, 2022c. Ukraine: Impact of the war on agriculture and rural livelihoods in Ukraine: Findings of a nation-wide rural household survey. Food and Agriculture Organization of the United Nations, Rome Federle, J., Meier, A., Muller, G., Sehn, V., 2022. Proximity to War: The stock market response to the Russian invasion of Ukraine. CEPR Discussion Paper 17185, London Fergusson, L., Ibanez, A.M., Riano, J.F., 2020. Conflict, Educational Attainment, and Structural Transformation: La Violencia in Colombia. Economic Development and Cultural Change 69, 335-371 George, J., Adelaja, A., Awokuse, T.O., 2021. The Agricultural Impacts of Armed Conflicts: The Case of Fulani Militia. European Review of Agricultural Economics 48, 538-572 George, J., Adelaja, A., Weatherspoon, D., 2020. Armed Conflicts and Food Insecurity: Evidence from Boko Haram's Attacks. American Journal of Agricultural Economics 102, 114-131 Glauben, T., Svanidze, M., Gotz, L., Prehn, S., Jamali Jaghdani, T., Duric, I., 2022. The War in Ukraine, Agricultural Trade and Risks to Global Food Security. Review of European Economic Policy 57, 157-163 36 Ibanez, A.M., Moya, A., Velasquez, A., 2022. Promoting Recovery and Resilience for Internally Displaced Persons: Lessons from Colombia. Oxford Review of Economic Policy 38, 595-624 Ihle, R., Bar-Nahum, Z., Nivievskyi, O., Rubin, O.D., 2022. Russia's Invasion of Ukraine Increased the Synchronisation of Global Commodity Prices. Australian Journal of Agricultural and Resource Economics 66, 775-796 Khoshnood, K., Raymond, N.A., Howard, C., 2022. Ukraine's crop storage infrastructure: Post-invasion damage assessment. Yale School of Public Health Humanitarian Research Lab and Oak Ridge National Laboratory, New Haven Korovkin, V., Makarin, A., 2023. Conflict and Intergroup Trade: Evidence from the 2014 Russia-Ukraine Crisis. American Economic Review 113, 34-70 Kussul, N., Shelestov, A., Lemoine, G., 2023. Using freely available imagery to identify damage from military action: A machine learning approach applied to Ukraine In: Paper Presented at the 9th International Conference on Agricultural Statistics. World Bank, Washington DC La Porta, R., Shleifer, A., 2014. Informality and Development. Journal of Economic Perspectives 28, 109-126 Lin, F., Li, X., Jia, N., Feng, F., Huang, H., Huang, J., Fan, S., Ciais, P., Song, X.-P., 2023. The impact of Russia-Ukraine conflict on global food security. Global Food Security 36, 100661 Miguel, E., Roland, G., 2011. The Long-Run Impact of Bombing Vietnam. Journal of Development Economics 96, 1-15 Muriuki, J., Hudson, D., Fuad, S., March, R.J., Lacombe, D.J., 2023. Spillover Effect of Violent Conflicts on Food Insecurity in Sub-Saharan Africa. Food Policy 115 Nguimkeu, P., 2022. A Structural Model of Informality with Constrained Entrepreneurship. Economic Development and Cultural Change 70, 941-980 Nivievskyi, O., Donchenko, O., Iavorskyi, P., 2021. Assessing the role of small farmers and households in agriculture and the rural economy and measures to support their sustainable development. Kyiv School of Economics, Kyiv Ohnsorge, F., Yu, S., 2021. The long shadow of informality: Challenges and Policies. World Bank, Washington, DC. Rockmore, M., 2020. Conflict-Risk and Agricultural Portfolios: Evidence from Northern Uganda. Journal of Development Studies 56, 1856-1876 Sinha, R., Aghabarari, L., Rostom, A., 2022. Conflict and the Nature of Precautionary Wealth. Oxford Economic Papers 74, 567- 593 Tapsoba, A., 2023. The cost of fear: Impact of violence risk on child health during conflict. Journal of Development Economics 160, 102975 Tellez, J.F., 2022. Land, Opportunism, and Displacement in Civil Wars: Evidence from Colombia. American Political Science Review 116, 403-418 Ulyssea, G., 2018. Firms, Informality, and Development: Theory and Evidence from Brazil. American Economic Review 108, 2015-2047 Ulyssea, G., 2020. Informality: Causes and Consequences for Development. Annual Review of Economics 12, 525-546 Voors, M.J., Nillesen, E.E.M., Verwimp, P., Bulte, E.H., Lensink, R., Van Soest, D.P., 2012. Violent Conflict and Behavior: A Field Experiment in Burundi. American Economic Review 102, 941-964 Zhu, W., Paudel, K.P., Luo, B., 2021. The Influence of Land Titling on the Disparity between Willingness to Accept and Willingness to Pay Values. Journal of Environmental Planning and Management 64, 930-953 37