LOCUST INVASION IN ETHIOPIA: SCOPE & IMPACT Evidence from the World Bank-Supported High-Frequency Phone Surveys John Ilukor and Sydney Gourlay ©FAO/Sven Torfinn This document has been prepared by John Ilukor (Economist, Development Data Group, World Bank, Kampala, Uganda) and Sydney Gourlay (Economist, Development Data Group, World Bank, Rome, Italy). Valuable comments were received from Alemayehu Ambel (Development Data Group, World Bank), Melissa Williams (Agriculture Global Practice, World Bank), and Alberto Zezza (Development Data Group, World Bank). BACKGROUND This brief reports on the scope, intensity, and type of damage incurred by rural The desert locust, the most destructive households as a result of both the first and migratory pest in the world, is highly mobile second locust invasions in Ethiopia, based and feeds on large quantities of any kind of on high-frequency phone survey data. green vegetation, including crops, pasture, and fodder with great potential to cause At the end of April 2020, the World Bank deterioration in the food security situation Living Standard Measurement Study team across the East Africa region. Ethiopia piggy-backed on the ongoing COVID-19 experienced two invasions of locust in 2020. High Frequency Phone Survey (HFPS) by The first invasion, which spread from Yemen integrating a short module on locusts to to Ethiopia, Kenya, Somalia, Uganda, Sudan, monitor the spread and impact of the and Tanzania, took place from January to desert locust. To implement the HFPS, May. This first invasion of desert locust is the team leveraged the Living Standards reported to have invaded 180-240 Woredas Measurement Study - Integrated Surveys and laid to waste over 200,000 hectares on Agriculture (LSMS-ISA) program by of land in Ethiopia alone, primarily in crop drawing a sample of households interviewed land in eastern and southern Ethiopia. The in the 2018/2019 round of the Ethiopia second invasion of locust in Ethiopia, which Socioeconomic Survey (ESS).2 The extensive started in late September and peaked in information collected in the ESS, less than October-November is the worst locust one year prior to the pandemic, provides a invasion and feared to be 20 times more rich set of background information on the severe that than the first invasion and it COVID-19 High Frequency Phone Survey has destroyed crops, grazing lands and of households which can be leveraged trees. The impacts of locust invasions are to assess the differential impacts of the exacerbated by COVID-19 restrictions, COVID-19 and desert locust invasion. A war, and flooding which are expected to module on locust was integrated into substantially reduce Ethiopia’s economic the fourth, sixth, and seventh rounds of growth. Poverty levels are expected to rise the HFPS, conducted in April/May 2020, and over 2 million people estimated to fall September/October 2020, and October/ into poverty while the scale of Productive November 2020 respectively. Safety Net Programme is expected to rise from 9 million to as high as 15 million.1 2 The first round of the Ethiopia HFPS covered a sample of 3,249 households from both urban and rural areas. For a description 1 One UN Assessment: Socio-Economic Impacts of COVID 19 in of the Ethiopia HFPS methodology see the Monitoring COVID- Ethiopia (2020). Available here: https://ethiopia.un.org/en/87497- 19 Impacts on Households in Ethiopia: Survey Methodology one-un-assessment-socio-economic-impact-covid-19-ethiopia Document, by Ambel et al. (2020; https://bit.ly/3rigwvK ). 2 The HFPS round four survey asked The highest proportion of Ethiopia’s households about the presence of locusts in rural households interviewed are from their kebele and on their farm, separately, Oromia (37%), followed by Amhara (27%), and the extent of damage observed on and Southern Nations, Nationalities, and their farms as a result of the first invasion Peoples’ Region (SNNPR, 22%), with the of locusts. The same module was repeated smallest share in Harar, Dire Dawa and in round seven of the HFPS, with a slightly Gambela as shown in Figure 1a. Overall, expanded set of questions following the the share of households cultivating peak of the second invasion. A set of crops (57%) is higher than those engaged questions was also included in round six in livestock (53%). Most of the HFPS of the HFPS but given that it was early households engaged in crop and livestock on in the locust invasion, that data is not farming are from SNNPR, Oromia, and presented here. To provide context to the Amhara as shown in Figure 1b. In the locust findings, the breakdown of the rural Somali region, very few households in the population and agricultural participation of HFPS sample are engaged in agriculture the HFPS sample is provided by region in and only 7% kept livestock.45 Figure 1. Note that these regional agricultural participation rates are not necessarily representative of the country as the HFPS sample is not regionally representative.3 Figure 1. Share of rural households interviewed by region and engagement in crop and livestock a. Share of total rural HFPS households, by region4 b. Household engagement in crop and livestock (% of HFPS households)5 Addis Ababa Total Oromia SNNPR Oromia Amhara Benishangul-Gumuz SNNPR Amhara Tigray Tigray Somali Gambela Benishangul-Gumuz Harar Afar Dire Dawa Gambela Afar Dire Dawa Addis Ababa Harar Somali 0% 10% 20% 30% 40% 0% 10% 20% 30% 40% 50% 60% 70% 80% Livestock Ag Crop Ag Note: Ethiopia HFPS is not representative at the regional level; for illustrative purposes only. 3 For a complete picture of agricultural participation by region, 4 Share of total rural households in Ethiopia, as of the Ethiopia see the Ethiopia Socioeconomic Survey (ESS) 2018/19 Survey HFPS Round 4 survey. Report (available in the World Bank’s Microdata Catalog: https://microdata.worldbank.org/index.php/catalog/3823/ 5 Crop and livestock engagement for HFPS households is related-materials) determined from the Ethiopia Socioeconomic Survey (ESS), in which HFPS households were included. 3 INCIDENCE OF LOCUSTS Afar, Somali, and Harar regions having the highest incidence of reported locusts in the Over half of all rural households kebele and on the farm. Figure 2 illustrates experienced locusts in their kebele, the location of locusts reported in the HFPS and nearly 30% experienced locusts on sample, with the icons representing the rural their own farms, during the first locust households who reported locusts in the invasion. At the peak of the second kebele and those who did not, in both the invasion, 37% of rural households first and second invasions.6 observed locusts in the kebele and 20% on their farm. In both invasions, incidence Figure 3 illustrates the incidence of of locusts was widespread impacting all locusts for both the first and second regions with rural populations, with the invasion by region (note that this is for Figure 2. Location of locusts reported in the HFPS, by invasion First Invasion Second Invasion First invasion Second invasion #X X ## X #X# # # # X X# X # # X X X XX# X # X X# X # # X # # # X## X X X # X ## ## XX# # X X ### X# # XX X # # X # X## X ## X # ## # X X X # X# X X X X XX # X X # X# X X ## #X#X # #X# # XXX# X # # # # XX # X # # X# X X # X X XX X X # ## # # X X # X X X # X# # X XX X X # XX # X X # # X ### X # X # X # X X XXX X X X X X # XX X X X X### X# X # X X X # X # # # #X X XX X X X # X # X # X # # X ##X # # # # ## ## # X X # X XX #X X ### # X X # X # #X # X # X# ### X XX XX X XX # XX# # X# #### # XX X # XX # X # X X # # X XX XX # X X X # # XX X X # X# ### X # X # ## X X XX X X # # # X # XX # # # X X## # X # # # XX # # # X # # X X # XX X # X ## # XX# # X XX# ## # # ## # # XX X X # X # # X XX#X XX# # # X X ## # X X # X XX # X # X # X ## X # X X # X # X X#X # # # # # X X X #X # # # ## # # # # # X X # # # # # # # Locusts reported X X No locusts reported # Locusts reported X No locusts reported # Locusts reported No locusts reported # Locusts reported X No locusts reported Figure 3. Proportion of rural Households reporting locust invasion in the two rounds at kebele and farm level, by region a. Presence in the kebele b. Presence on the farm National National Somali Somali Harar Afar Afar Harar Oromia Oromia Tigray Amhara SNNPR SNNPR Amhara Tigray Gambela Gambela Dire Dawa Dire Dawa Benishangul-Gumuz Benishangul-Gumuz 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% Second invasion First invasion Second invasion First invasion Note: Ethiopia HFPS is not representative at the regional level; for illustrative purposes only. 6 Note that figure includes only the rural households interviewed in the HFPS and is not necessarily representative of the rural population. GPS locations are offset from the true location to protect respondent anonymity. 4 illustrative purposes only as the HFPS is than the first invasion, as evidenced also by not representative at the regional level). the smaller proportion of rural households The higher number of rural households that reported no damage in the second reporting locusts at the kebele and at farm invasion relative to the first invasion. level in the first invasion relative to the This is also supported by the difference second invasion may be attributed to the in the intensity of damage reported by fact that second invasion occurred in regions households. Complete damage was like Afar, Harar, and Somali which are less reported by 13% of households at the peak populated than Oromia, Amhara, and of the second invasion, while only 6% of SNNPR where the first invasion hit harder. households with damage in the first wave reported it to be complete damage (Figure LOCUST DAMAGE AND 4b). Given both the incidence and intensity DAMAGE INTENSITY of reported damage, it appears the second locust invasion was more destructive than Fifty-nine percent of households the first, as predicted in the locust watch.7 that reported locusts on their farm in the second invasion also reported Eighteen percent of households who damage to crops as a result, with 63% reported crop damage as a result of reporting damage to grazing lands and locusts in the second invasion reported 52% reporting damage to trees. Damage complete crop damage (compared to incurred as a result of the first locust 11% in the first invasion). Damage to invasion was reportedly lower, with 44% grazing lands as a result of the second of households with locusts on their farm invasion was significant, with 20% of reporting crop damage, 38% reporting households who reported damage to grazing damage to grazing lands, and 23% reporting lands indicating complete damage, and damage to trees (Figure 4a). The results 25% indicating severe damage (Figure 5). suggest that the second invasion may be Trees fared relatively well, with over 40% of more detrimental to the agricultural sector households reporting light damage to trees Figure 4. Proportion of rural households reporting locusts to have damaged crops, grazing lands, and trees as well as intensity of damage a. Damage b. Intensity No damage Complete Trees Severe Grazing lands Moderate Crops Light 0% 20% 40% 60% 80% 0% 10% 20% 30% 40% 50% Second invasion First invasion Second invasion First invasion 7 Desert locusts in East Africa: A plague of another order (2020). Available here: https://reliefweb.int/report/ethiopia/ desert-locusts-east-africa-plague-another-order 5 in both the first and second invasions. spraying (10%), aerial spraying (8%), With 57% of households engaged in crop and burying of the locusts (4%), in the agriculture and 53% engaged in livestock, second invasion (data not available for the damage to crops, trees, and grazing lands first invasion; Figure 6a). Nine percent of as a result of locusts is both inevitable the rural households with locusts in the and detrimental. The timing of the second kebele reported to have done nothing invasion, after the crops were planted, to control the locust. Fifty-six percent of nearing maturity or at the start of the the households using some control or harvest, a period when crops were vulnerable prevention method reported that method to damage, likely attributed to the increased was very effective (Figure 6b). Six percent of crop damage observed in the second the households using control or prevention invasion relative to the first, where the methods reported them to be ineffective, as invasion came on before the planting season. shown in Figure 6b. The HFPS data allows for an understanding of the effectiveness of LOCUST CONTROL MEASURES the various methods separately but given the low incidence of households using the Eighty-three percent of rural households burying or spraying methods, effectiveness with locusts in the kebele reported is presented in aggregate. However, note noise making or smoke as a control that only 42% of households employing the or prevention method, with a smaller noise or smoke making method reported it share of households using backpack to be very effective. Figure 5. Intensity of locust-induced damage Crops Grazing Lands Trees Complete Complete Complete Severe Severe Severe Moderate Moderate Moderate Light Light Light 0% 10% 20% 30% 40% 50% 60% 0% 10% 20% 30% 40% 50% 60% 0% 10% 20% 30% 40% 50% 60% Second invasion First invasion Second invasion First invasion Second invasion First invasion Figure 6. Locust control methods reported by rural households with locusts in the kebele, and their effectiveness a. Control and prevention methods b. Effectiveness 6% None Buried Made noise/smoke 56% 38% Backpack spraying Aerial spraying 0% 50% 100% Not at all effective Moderately effective Very effective 6 SUPPORT NEEDS FOR LOCUST- Looking at households that experienced AFFECTED HOUSEHOLDS locusts on their farm, we find that support in the form of food was most frequently Sixty-four percent of households that reported as necessary by households reported locust in their kebele indicated who incurred crop damage (42% of that some form of support was needed households), followed by households who in order to cope with the second locust incurred damage to grazing lands (37% of invasion, as shown in Figure 7 (data not households), as seen in Figure 8a, which available for the first invasion). The most is consistent with households engaged in requested form of support was food items agriculture for subsistence. Cash was sought (32%) followed by cash (21%). The least most frequently by households that suffered requested forms of support were livestock tree damage (38% of households with tree feed (7%), seed for next season cultivation damage). Where complete or severe damage (3%), and work to earn income (1%). The need was incurred as a result of the locusts, food for support in the form of food was highest support was the primary need, with 69% and in Somali and Harar regions. The share of 39% of households who reported complete households indicating a need for support or severe damage also indicating a need for via hat not needing any form of support, food, respectively (Figure 8b). None of the the majority of these households did not rural households that reported completed experience locusts on their own farm. damage requested livestock feed or seed. Figure 7. Reported support needs (% of households with locust reported in kebele) Nothing Food items Cash Livestock feed/fodder Seed for next season Work to earn income 0% 10% 20% 30% 40% Figure 8. Support needs for rural households with locust damage on farm, by type and intensity of damage a. Support needs by type of damage b. Support needs by intensity of damage Nothing Nothing Food items Food items Cash Cash Livestock feed/fodder Livestock feed/fodder Seed for next season Seed for next season Work to earn income Work to earn income 0% 20% 40% 60% 0% 20% 40% 60% 80% Trees Grazing lands Crops Complete Severe Moderate Light 7 LOCUST INVASION IN ETHIOPIA: SCOPE & IMPACT EVIDENCE FROM THE WORLD BANK-SUPPORTED HIGH-FREQUENCY PHONE SURVEYS MAY 2021 www.worldbank.org/lsms