REPORT NO. 2 NOVEMBER 2020 Results from the High-Frequency Phone Survey of Households Round 2 (17 August–7 September 2020)1 Authors: Wendy Karamba, Kimsun Tong INTRODUCTION The coronavirus disease 2019 (COVID-19) pandemic and its effects on households create an urgent need for timely data and evidence to help monitor and mitigate the social and economic impacts of the crisis on the Cambodian people, especially the poor and most vulnerable. To monitor the socioeconomic impacts of the COVID-19 pandemic and inform policy responses and interventions, the World Bank designed and conducted a nationally representative High-Frequency Phone Survey (HFPS) of households in Cambodia. The survey covers important and relevant topics, including knowledge of COVID-19 and adoption of preventative behavior, economic activity and income sources, access to basic goods and services, exposure to shocks and coping mechanisms, and access to social assistance. This brief summarizes the findings of round 2 (R2) of the HFPS of households. The information presented is based on a sample of 1,667 households, of which 612 were drawn from the nationally representative Living Standard Measurement Study Plus (LSMS+) and 1,055 from the list of beneficiaries of the conditional cash transfer program for pregnant women and children under 2 with an IDPoor equity card (IDPoor sample). IDPoor is Cambodia’s national poverty identification program and official targeting mechanism for programs that support the poor. Unless otherwise noted, the results presented are primarily drawn from the LSMS sample. The results of round 2 (R2) implemented August–September 2020 are compared against results of round 1 (R1) implemented May–June 2020 to trace the evolution of key indicators. The same households will be tracked over 10 months, with selected respondents—typically the household head—completing interviews every 8 weeks. Monitoring the well-being of households over time will improve understanding of the effects of, and household responses to the COVID-19 pandemic in near-real time. HIGHLIGHTS -- HFPS ROUND 2 Employment remained steady between May and August 2020 as Cambodians faced fewer disruptions to work activities. About 70 ❖ percent of the respondents were working, many of whom remained in the same jobs. Among those who stopped working, reasons were largely related to seasonality or illness rather than to business closures due to COVID-19. Despite relatively steady employment, losses to household income remain widespread. ❖ COVID-19 continues to negatively affect non-farm business activities although less severely compared to the onset of the pandemic. The ❖ share of non-farm household businesses reporting having generated less or no revenues declined from 4 in 5 households in May to about 3 in 5 households in August. Weak demand continues to be the main constraint household businesses face. Migrants experienced losses in wages and jobs since the COVID-19 outbreak in March resulting in diminished remittances to migrant-sending ❖ households. However, the declines in household income have slowed following the government’s launch of a new COVID-19 cash transfer program to ❖ assist poor and vulnerable households during COVID-19. Fewer households reported food shortages since the launch of the cash transfer program on 24 June 2020. 1 Wendy Karamba and Kimsun Tong led the Cambodia High Frequency Phone Survey (HFPS) team that comprised of Maheshwor Shrestha and Sokbunthoeun So. Nuppun Research Consulting implemented the survey with technical and financial support from the World Bank. Additional contributions for the HFPS were received from the Public Financial Management and Service Delivery Trust Fund contributed by Australia and the European Union. The team is grateful to the National Institute of Statistics of the Ministry of Planning and the Ministry of Social Affairs, Veterans and Youth Rehabilitation for their collaboration. The team benefitted from advice and comments from World Bank Group colleagues including Rinku Murgai. REPORT NO 2 November 2020 Access to food staples and health services remained robust as markets continued to function well. Virtually all households that sought to buy ❖ staple foods and medicines, and access medical treatment were able to do so. Children’s engagement in learning activities increased between May and August as households adapted to the COVID-19 situation, and ❖ schools offered households alternative methods of instructions. About 75 percent of households with children aged 6–17 years were engaged in learning activities in August (R2) compared with 63 percent in May (R1). More households relied on face-to-face meetings with teachers, while fewer relied on mobile learning applications or television programs. EMPLOYMENT Employment remained steady between May and August following an initial shock due to the COVID-19 pandemic. PATTERNS AND While the share of respondents who were working declined sharply from 82 percent before the COVID-19 outbreak HOUSEHOLD to 71 percent in May (R1), it remained relatively unchanged in August (R2) at 70 percent. Most respondents INCOME remained within the same sector of employment. Only 5 percent of the respondents switched jobs, largely moving from the industry sector into the agricultural sector. The share of respondents who stopped working remained around 10 percent between the two rounds. Among those who stopped working, around 32 percent reported that it was due to seasonality in employment and 21 percent reported it was due to illness or quarantining.2 The COVID- 19 related reasons mostly cited in May such as business closures or temporary layoffs had largely disappeared. EMPLOYMENT STATUS Figure 1: Respondent working in the Figure 2: Non-farm business households last 7 days reporting changes in sales revenues 17% 20% 11% 10% 6% 5% 48% 61% 73% 67% 66% 65% 42% 31% 18% 15% Round 1 Round 2 R1 R2 R1 R2 Did not work LSMS ID Poor Currently not working Currently working (change job) Higher The same Less No revenue Currently working (same job) COVID-19 continues to negatively affect non-farm business activities albeit less severely compared to the onset of the pandemic. The share of non-farm household businesses reporting less or no revenues declined. 4 in 5 households reported earning less or no revenues in April compared to March, while about 3 in 5 households did so in July compared to June. Household businesses reporting that their revenues remained the same doubled, reaching 31 percent in July. Similar trends are observed among IDPoor households although at different magnitudes. The share of households operating a non-farm family business remained relatively unchanged. About one-third (32–34 percent) of households in the LSMS sample operated non-farm family businesses during May– August, and around 16–19 percent of IDPoor households did so during June–August. Most of the businesses were in retail sales. Nearly 80 percent of the businesses operating in August were in retail, compared to 67 percent in May. Weak consumer demand continues to drive the loss or lack of revenues for non-farm family businesses. Among household businesses reporting less or no revenues, about 88 percent reported having fewer or no customers. Lower demand is equally affecting all household businesses across urban and rural areas and among households in the bottom 40 and top 60 percent. 2Note Cambodia has thus far contained the spread of the coronavirus. As of November 17, 2020 there were 303 confirmed cases according to the World Health Organization (https://www.who.int/countries/khm/). Page 2 REPORT NO 2 November 2020 Most migrants have lost jobs or wages because of the COVID-19 pandemic. COVID-19 has forced the return of some migrants. Pre-COVID-19, 3 in 10 Cambodian households had a migrant. Since the COVID-19 outbreak, 1 in 10 households have a migrant that has returned home, while 2 in 10 households have a migrant that has remained at their destination. Among the households whose members are currently away, 80 percent had migrant workers working as wage employees. More than 60 percent of migrant households reported declines in the migrant member’s income and estimated that this income fell by 47 percent since the COVID-19 outbreak, which in turn reduced the amount of remittances received. IDPoor households reported that the income of their migrant members had declined by 42 percent compared to the pre-COVID-19 outbreak. Among the 1 in 10 return migrant households, most had a migrant return home because of factory closures (36 percent) and lack of work (24 percent). One-third (34 percent) of the return migrant households had a returnee who was currently out of work, but the estimate was only 16 percent for IDPoor households. The other third (29 percent) of the returnees were engaged in family farm activity, while 16 percent worked as wage employees. Conversely, only 16 percent of the return migrant households from IDPoor households engaged in family farm activity, while 58 percent worked as wage employee (probably because poor households are likely to own only small parcels of agricultural land). Wage and job losses among migrants have had an adverse impact on remittance flows, which are an important source of income for rural households. Over 70 percent of migrant households had received remittances since March. However, migrant households received a median of US$250 since March, which was lower than the pre- COVID-19 outbreak. IDPoor households had only received US$150 since the onset of the pandemic. The survey did not inquire the magnitude of the change. Figure 3: Migrant households reporting changes in income since mid-March (% of households) LSMS IDPoor Don't know Increased Increased Don't know 4% 3% 6% 8% Stayed the Stayed the same same 26% 28% Reduced Reduced 65% 60% Figure 4: Migrant households reporting changes in remittances since mid-March (% of households) LSMS ID Poor Increased Don't know Increased 5% 1% 2% Stayed the Stayed same Reduced the same Reduced 36% 54% 41% 61% Page 3 REPORT NO 2 November 2020 INCOME Income losses remain widespread although the decline in household income has slowed since May due to fewer disruptions to work activities and increased government assistance. About 3 in 5 households reported a decline in household income in R2 as compared to 4 in 5 households in R1. The decline in income has slowed for income from all sources, but most prominently for income from farming. Only 32 percent of households that derive their livelihoods from farming, livestock or fishing reported that their income declined in R2, which is a 30 percentage point reduction since R1. The share of households reporting a decline in income from wage employment and non- farm family businesses also fell by 13–14 percentage points. Remittances and assistance from family have not further deteriorated as the share of households reporting a decline fell 17 and 44 percentage points respectively from about 75 percent. The share of households reporting an increase in assistance from government and non- governmental organizations increased 10-fold between May and August from 4 percent to 42 percent. Figure 5: Changes in household income between May and August 16 15 16 15 16 29 33 29 27 31 83 85 83 84 83 63 61 63 64 61 R1 R2 R1 R2 R1 R2 R1 R2 R1 R2 Cambodia Urban Rural Bottom 40 Top 60 Reduced Stayed the same Increased SOCIAL On June 24, 2020 Cambodia launched a nationwide relief cash transfer program to support poor and vulnerable ASSISTANCE households during the COVID-19 pandemic. The program provides cash transfers to households identified by the government as part of the Identification of Poor Households Program, known as “IDPoor,” The program is a temporary measure intended to alleviate the economic hardships the poor and vulnerable face from the impacts of COVID-19. The government would provide cash payments to households covered under the IDPoor program for an expected 7 months. In June, Prime Minister Hun Sen announced that the program would spend $25 million a month to support IDPoor households and would benefit around 560,000 families.3 Since June 2020, there has been a marked increase in the share of households receiving social assistance (SA) from the government. The share of IDPoor households that received social support regardless of source increased 41 percentage points from 50 percent to 91 percent between June and August 2020. The government provided almost all of the assistance (97 percent) and most of the assistance was in the form of cash transfers (90 percent). The COVID-19 relief cash transfers for poor and vulnerable households accounted for more than 60 percent of the social assistance programs that beneficiaries reported in the survey, while the conditional cash transfer for pregnant women and children under 2 accounted for nearly 30 percent. 3 Associated Press. June 24, 2020. “Cambodia to provide cash to poor hit by COVID-19 outbreak.” Page 4 REPORT NO 2 November 2020 Figure 6: Share of SA Figure 7: Sources of SA Figure 8: Types of SA received beneficiaries 1% 1% 1% 89% 91% 2% 8% 50% 30% 97% 90% LSMS ID Poor Government Food NGO Direct cash transfers Round 1 Round 2 Religious body Other Other in-kind transfers (excluding food) The relief cash transfer program has reached many IDPoor households, but a small fraction are still uncovered. About 90 percent of IDPoor households reported having received the relief cash transfers during the months of August and September. While most IDPoor households have received these transfers, about 10 percent of households with an equity card have yet to receive the cash transfers. Only 1 in 4 of these households had verified the validity of their equity card with the commune council, which is a prerequisite for receiving the COVID-19 relief cash transfers. The remaining 3 in 4 households had not registered at the commune council because they were not aware of the program or did not know that the cash transfers was for them. To ensure the remaining IDPoor households are aware of the relief available during COVID-19, further outreach and awareness campaigns are needed. Figure 11: COVID-19 cash Figure 12: Registered at Figure 13: Reasons for not transfer receipt commune council registering 14 10 20% 26% 86 90 53% 27% 74% LSMS ID Poor Unaware of the program Transfer No transfer Yes No Didn’t know the transfer is for them Other Most IDPoor households have received multiple installments of the COVID-19 relief cash transfers averaging US$133. Those who verified the validity of their equity card with the local authorities immediately following the launch of the program would have received about 3 installments by the time of the survey. Among the cash transfer beneficiary households with an equity card, 53 percent had received transfers three times, 43 percent two times, and 4 percent one time since June 2020. Poor households with an equity card received an average of US$133 from the COVID-19 relief cash transfer program since June 2020. According to the HFPS, “very poor” households that are classified as IDPoor1 in the program have received an average of $67 per month while “poor” households that are classified as IDPoor2 have received $52 per month. Page 5 REPORT NO 2 November 2020 Figure 8: Number of times cash Figure 9: How much cash was received transfers were received so far in total $136 4% $113 $91 53% 43% $47 1 time 2 times 3 times 1 time 2 times 3 times Total According to the program documents, IDPoor households in rural areas receive a base transfer of $20 a month, and IDPoor households in Phnom Penh and other urban areas receive $30 due to higher cost of living. These transfers are supplemented by smaller payments to each family member, and additional payments to each vulnerable family member consisting of elderly persons aged 60 and above, children under 5 years old, and/or persons living with HIV or a disability. These supplemental payments vary depending on area of residence and classification of the household as IDPoor1 or IDPoor2. For instance, rural households will receive $20, and each family member will receive an additional $6 or $4 depending on whether the household is IDPoor1 or IDPoor2 respectively. For households in Phnom Penh, a family will receive $30, and each family member will receive $13 for IDPoor1 or $9 for IDPoor2. A family in other urban areas will receive $30, but each family member will receive $10 and $7 for IDPoor1 and IDPoor2, respectively. Beneficiaries of the COVID-19 cash transfers largely spent their cash on food. Consequently, the food security situation improved since the launch of the COVID-19 relief cash transfer program for poor and vulnerable households on 24 June 2020. The proportion of IDPoor households which reported that they ran out of food in the last 30 days because of lack of money or resources decreased by 37 percentage points between June and August. There was also an improvement in other indicators such as fewer households eating less frequently. However, some households continued to eat a less diverse diet. Households also used their cash transfers to purchase other essential items (58 percent), and for loan repayments (15 percent). Figure 10: How the COVID-19 cash tranfers were used 100% Food 98% 58% Other essential items 47% 15% Paid back loan 12% 2% Other 13% ID Poor LSMS Page 6 REPORT NO 2 November 2020 ACCESS Markets remained functional and able to supply basic necessities despite the COVID-19 pandemic. Virtually all TO BASIC households regardless of poverty status and area of residence were able to buy rice, fish/meat, and vegetable/fruit NECESSITIES in both May and August when they sought these staples. Similarly, ninety-nine percent of households that tried to purchase medicine were able to do so in both periods. Among households that needed medical treatment, virtually all of them were able to access medical services. Female respondents were more likely to report they needed medical treatment for themselves or other household members than male respondents. FOOD STAPLES Figure 14: Household able to buy medicine in the last 7 days 30% 34% 57% 55% 69% 66% 42% 45% R1 R2 R1 R2 LSMS ID Poor Yes No Not try Figure 15: Households needing and able to access medical treatment 45 44 60 73 98 100 100 98 55 56 39 26 R1 R2 R1 R2 R1 R2 R1 R2 LSMS ID Poor LSMS ID Poor Need medical treatment Access medical treatment Yes No SCHOOLING Children’s engagement in learning activities is rising following initial disruptions brought about by the COVID-19 outbreak. On March 16, 2020, the Ministry of Education, Youth and Sport (MoEYS) directed the closure of all public and private schools and required students to study remotely. Households largely relied on television or radio programs to complement learning at home. Since September 7, 2020, schools have partially reopened and must follow Standard Operating Procedures recommended by MoEYS to prevent the spread of COVID-19. These procedures included limits to class sizes, hygiene practices, social distancing, temperature screening, and the use of masks. At the time of the survey, a partial reopening had not yet been in effect. Instead, greater adaptations of households and alternatives from schools offered more opportunities for learning at home. The share of households with children aged 6–17 who had been in school engaged in learning activities rose from 63 percent to 75 percent. Despite this improvement, 1 in 4 households with children continue to have few opportunities to learn during school closures. Page 7 REPORT NO 2 November 2020 Figure 16: In the last 7 days, have the children engaged in education or learning activities? 80% 75% 76% 69% 70% 68% 66% 63% 63% 64% 62% 59% All Urban Rural Bottom 40 Top 60 LSMS ID Poor Round 1 Round 2 School closures disproportionately affected learning activities of the most vulnerable students. While school closures are a logical solution to enforcing social distancing within communities, households with few opportunities for remote learning or households facing economic burdens because of the crisis may not be able to mitigate the loss of learning. Households in the bottom 40 percent of the income distribution were less able to access learning opportunities than the top 60 percent. While participation in learning activities for children in the bottom 40 percent increased from 59 percent in May 2020 (Round 1) to 70 percent in August (Round 2), their participation is still 10 percentage points lower than that of the top 60 percent. Prolonged school closures without access to effective alternative options may cause further loss in human capital and diminish long-run economic opportunities. There is strong evidence that schools are adopting alternatives to mitigate the loss of learning. In August 2020 (Round 2), most households with children had had a meeting with teachers (66 percent), completed assignments (32 percent), used mobile learning applications (32 percent) and watched educational television programs (14 percent). The proportion of households with children who met with teachers rose by 46 percentage points from the previous round, while those who used mobile learning applications and watched television programs declined by 26 and 33 percentage points, respectively. However, children in the poorest households are less likely to have met with a teacher than richer households. Page 8 REPORT NO 2 November 2020 BOX: SURVEY METHODOLOGY The HFPS consists of two separate samples: (a) Living Standard Measurement Study Plus (LSMS+) and (b) IDPoor households. LSMS+ is a nationally representative household survey implemented October–December 2019 by the National Institute of Statistics (NIS) with technical and financial support from the World Bank. LSMS+ consists of 1,512 households, of which 1,364 have a phone number. The phone survey successfully reached and completed interviews for 700 of 1,364 households in May, generating a response rate of 51 percent. In August, 612 households were successfully re-interviewed. Sampling weights were adjusted to ensure representativeness at the national and urban/rural level and according to the steps outlined in Himelein, K (2014) to obtain unbiased estimates from the sample. To monitor and evaluate the Cash Transfer Program for Poor and Vulnerable Households during COVID-19, which was launched in June, an additional 1,000 IDPoor households were sampled. Before selection, the beneficiary list of the conditional cash transfer for pregnant and child under 2 was assigned into 5 strata i.e. Phnom Penh and other urban areas, Plain, Tonle Sap, Coastal, Plateau and Mountain. The sample was randomly selected with proportional to the size of the IDPoor households in each strata. The phone survey successfully interviewed 984 households in June. In August, 784 households responded to the calls, where 271 were replacement households. Sampling weights were computed to ensure the representativeness of IDPoor households at national and regional level. The HFPS was implemented using Computer Assisted Telephone Interview (CATI) techniques and the questionnaire was programmed using the Survey Solutions CAPI software package. Enumerators used mobile phone devices. Enumerators were given data bundles, allowing for internet connectivity and for daily data transfer and synchronization with the server. Field supervisors reviewed the survey responses with enumerators via one-on-one calls daily and addressed concerns that arose immediately following enumerators interview. At the same time, a Research Analyst was in charge of checking the uploaded data daily to identify the errors so as to inform the field supervisors and enumerators. Data collection parameters, LSMS round 2 ❖ Data collection period: 17 August–7 September 2020 ❖ Completed interviews: 612 households (199 urban, 413 rural) ❖ Average duration of interview: 24 minutes Data collection parameters, IDPoor round 2 ❖ Data collection period: 17 August–7 September 2020 ❖ Completed interviews: 1,055 households (56 urban, 999 rural) ❖ Average duration of interview: 24 minutes Page 9