POLICY BRIEF AUGUST 2020 CARE WORK AND INTRA-HOUSEHOLD TENSIONS DURING COVID-19 EVIDENCE FROM AN ONLINE SURVEY OF GIG WORKERS IN INDIA SUMMARY This note examines gender disparities in care work and intra-household tensions among online gig workers in India. The data was collected as part of an online experiment in April 2020, shortly after lockdown measures were implemented to mitigate the spread of the coronavirus disease (COVID-19). The findings show that childcare and eldercare responsibilities have increased for everyone during the lockdown, but women have disproportionately felt the burden of increased care work. Further, there was an increase in domestic violence, pointing to added stress and intra-household tensions. Policy makers need to incorporate a gender lens in emergency responses in order to promote women’s safety and wellbeing during COVID-19 and beyond. BACKGROUND WOMEN ARE OVERREPRESENTED IN: Across South Asia, impacts of the pandemic caused by the coronavirus disease (COVID-19) are anticipated to go CHILDCARE beyond the immediate public health risks. From a gender perspective, women are typically overrepresented in insecure forms of employment, as well as perform the bulk of care work at home, including childcare, eldercare and ELDERCARE other household activities. With governments implementing full or partial lockdowns, children are out of school and women are experiencing an increase in unpaid work at home, especially in contexts where inter-generational living HOUSEHOLD ACTIVITIES and care arrangements are common. As intra-household relations come under pressure, there is also an increase in cases of gender-based violence (Peterman, et al., 2020; McLaren, Wong , Nguyen, & Mahamadachchi, 2020). WITH INDIA HAVING THE: This note examines gender disparities in care work and intra-household tensions among online gig workers in India. 2ND LARGEST MARKET The gig economy in India has grown massively since 2010, OF FREELANCERS IN making the country the second largest market of freelance THE WORLD WITH workers in the world, with an estimated 15 million workers 15 MILLION WORKERS, (Kasliwal, 2020). Digital jobs provide women an opportunity to enter the paid labor market while undertaking the bulk of unpaid household work. The flexible modality of work THE FLEXIBLE MODALITY OF WORK has attracted many women to this sector. Yet, gig work is HAS ATTRACTED MANY WOMEN TO usually demand-driven and can be uncertain at times, THIS SECTOR. especially during a pandemic when businesses are not fully AUGUST 2020 | 1 functional. Even though the nature of gig tasks performed by women tend to be low-skilled, women may work for longer hours and get paid less than men for the same jobs (Mehta, BOX 1: SURVEY QUESTIONS ON CHILDCARE, Mehta, & Kumar, 2020). The flexible work arrangements of ELDERCARE, DOMESTIC VIOLENCE, WORK the gig economy, such as having a job which does not require ARRANGEMENTS AND JOB SECURITY a physical space or specific timings, also do not address the unequal household roles that are mediated by cultural norms (1) Do you have any children aged 12 or under (Kasliwal, 2020).1 living at home? If yes, normally, who provides most of the childcare for your children? (yourself; you In the context of a global pandemic where remote work and your partner together; other children in the arrangements are expected to become more common, it is family; relative, friend, nanny or caregiver in your important to understand the realities faced by gig workers. home; other caregiver outside your home including This is important for the wellbeing of women gig workers, an individual care giver, childcare center, nursery, but also for the broader workforce who might need to work kindergarten, school, etc.) remotely in the aftermath of COVID-19. (2) As people grow older, it sometimes becomes difficult for them to perform some activities without help. Normally, do you provide care or assistance for DATA an elderly adult who needs help? If yes, since the beginning of March 2020, when coronavirus spread The data analyzed in this note was collected as part of an across India and you had to stay home, have you online experiment2 on Amazon Mechanical Turk in April 2020, provided any additional care or assistance for an a crowdsourcing website where businesses can hire remotely elderly adult who needed help? (yes/no) located workers to perform discrete tasks. The survey was implemented shortly after the government enacted lockdown (3) When people feel stressed and economically measures. Participants included 1,168 online workers3, who insecure during crises, they often knowingly or had an average age of 32.7, were highly educated (over 95 unknowingly hurt those around them. Since the percent had either college or university degree), and largely beginning of March 2020, when coronavirus lived in southern India (56 percent in Tamil Nadu and 17 spread across India, have you experienced any percent in Kerala).4 Most of the participants belonged to of the following? (a) your partner saying or doing disadvantaged castes (Scheduled Castes, Scheduled Tribes something to humiliate you in front of others; (b) and Other Backward Castes). threaten to hurt or harm you or someone you care about; (c) insult you or make you feel bad about The survey was administered to both male and female yourself; (d) hit, slap, kick or do anything else to respondents. The respondents were asked three sets of hurt you physically; (e) physically force you to have questions about the pre-lockdown and post-lockdown period: intercourse or force you to perform any other sexual (i) intra-household distribution of childcare and eldercare; (ii) acts against your will. Did you experience any of the instances of verbal or physical violence in their households; above in the month before the lockdown started?* and (iii) perceived job insecurity and ability to work from home during the pandemic (Box 1). (4) Are you at risk of losing your job or being temporarily laid off due to the coronavirus? (5) Is it easy for you (both in terms of your living situation and the flexibility of your employer) to work from home? 1 For example, a study on women gig workers in South Africa and Kenya showed that * Following the recommendation of the Institutional women workers’ spouses contributed little to care and domestic work (Hunt, et al., 2019). Review Board at Middlebury College, those who Another study focusing of flexible work arrangements in Malaysia shows that flexible work arrangements did not impact lifestyle (Subramaniam, Tan, Maniam, & Ali, 2013). responded positively to the domestic violence 2 Abel, M, T. Byker, J. Carpenter (2020) “Coronavirus: Risk Perceptions, Availability and Pro- question were provided the contact information of social Behavior.” Mimeo. 3 Since not all respondents filled out the relevant modules of the survey, the subsamples varied counseling organizations and shelters in their state. for each strand of analysis. 4 The remaining respondents came from 25 different states, each of which represented less than 4 percent of the sample. 2 | POLICY BRIEF FINDINGS reported that the additional care work was shared between them, men tended to overvalue their own contribution to care work, both before and after the lockdown. The share of CHILDCARE AND ELDER CARE men reporting themselves as the primary caregiver increased from 17 percent to 21 percent during the lockdown, whereas Before the pandemic, women were more likely than only 1 percent of women reported their male spouse as the men to provide childcare and eldercare. Following the primary caregiver in either period. lockdown, these responsibilities have increased for everyone, but women have disproportionately felt the Gender differences in childcare and eldercare are not solely burden of increased care work. the result of socioeconomic factors. The disparities hold after controlling for individuals’ characteristics, including age, caste, education and location (Table 1). Based on a In times of crisis, women’s responsibilities increase in the multivariate analysis of the likelihood of reporting childcare productive domain (labor or paid work) and the reproductive responsibilities, women were 6 percentage points more likely domain (domestic or care work) (McLaren, Wong, Nguyen, & to report an increase in childcare needs during the lockdown. Mahamadachchi, 2020). In India, even before the COVID-19 Women were also 11 percentage points more likely than crisis, 66 percent of women’s work was estimated to be unpaid men to be the sole primary caregiver for children. Similar compared to only 12 percent of men’s (World Economic disparities were found in eldercare. Conditional on having Forum, 2017). Women’s ability to take up paid employment an elder that the respondent was responsible for, women partly depends on availability of childcare and care for elderly were 6 percentage points more likely to report an increase in relatives. However, the pandemic has led to the closure of eldercare during the lockdown, but not any more likely than schools, nurseries, and daycares. While it is common in many men to be the sole primary care giver.5 Asian countries for childcare to be provided by grandparents or other close family and friends (Mazza, Marano, Lai, Janiri, During times of duress, care responsibilities may also increase & Sani, 2020), the pandemic has increased the burden of disproportionately for disadvantaged groups. Amidst the childcare on women, as families distance themselves from COVID crisis, it has been reported that discrimination against older people (as older people are at a higher risk of getting certain castes has intensified, constraining their access to infected by COVID-19). testing and treatment services (Muralidharan, 2020). In the sample of online gig workers, 61 percent of the participants Before the lockdown, almost twice as many women in the belong to socially disadvantaged castes (OBCs, SCs, and STs), sample, as compared to men, reported providing childcare of which almost 41 percent are women. Women from SCs and eldercare. About 25 percent of women reported that were particularly affected by the lockdown: they were 19 they were the sole primary caregiver for children (compared and 34 percentage points more likely to report an increase to only 11 percent of men), while almost half of women in childcare and eldercare, respectively, after controlling for reported that they shared care responsibilities with their individual characteristics. These findings are consistent with spouse. The lockdown has resulted in an increase in care the fact that SCs tend to have poorer socioeconomic status responsibilities for both men and women. However, women and limited access to basic services (Kumar, 2014; Bhagat, have reported taking on additional responsibilities at a higher 2013). rate than men during the lockdown (23 percent versus 17 percent in childcare and 24 percent versus 19 percent in The job flexibility afforded by the gig economy may have eldercare) (Figure 1). Further, even though most couples affected the distribution of care work within households. In Fig 1: Incidence of Childcare and Eldercare Responsibilities During Lockdown (self-reported) Fig 1: Incidence of Childcare and Eldercare Responsibilities During Lockdown (self-reported) Additional Childcare 17.0% Responsibilities 23.0% Additional Eldercare 19.0% Responsibilities 24.0% 0 10 20 30 40 50 60 70 80 90 100 Men Women 5 The age variables were also significant, suggesting older individuals were less likely to be responsible for increased eldercare. AUGUST 2020 | 3 Table 1: Multivariate Analysis for Elder and Childcare Responsibilities 1 2 3 4 5 6 7 8 9 10 Sole Primary Caregiver Increase in Childcare Sole Person Taking on Increase in Eldercare Sole Person Taking on (Baseline) Responsibilities Additional Childcare Responsibilities Additional Eldercare Female 0.169*** 0.169*** 0.055** 0.055** 0.095** 0.107** 0.074** 0.062** 0.050 0.050 (0.000) (0.000) (0.024) (0.035) 0.016 (0.010) (0.012) (0.041) (0.106) (0.116) Age 30-39 -0.018 -0.015 -0.041 -0.002 -0.068* (0.712) (0.594) (0.358) (0.949) (0.055) Age 40-49 -0.072 -0.043 -0.043 -0.102** -0.084** (0.209) (0.326) (0.469) (0.025) (0.047) Caste (Other) 0.080 -0.033 0.013 0.051 0.167** (0.374) (0.586) (0.883) (0.434) (0.043) Other Backward Caste (OBC) -0.011 -0.012 -0.098* 0.016 -0.027 (0.834) (0.726) (0.087) (0.700) (0.517) Scheduled Caste (SC) 0.240** 0.030 0.204** 0.147*** 0.081 (0.016) (0.545) (0.050) (0.002) (0.308) Scheduled Tribe (ST) 0.394* 0.067 0.190 -0.023 -0.058 (0.075) (0.123) (0.434) (0.896) (0.645) University Education -0.052 0.015 0.054 0.046 0.091*** (0.225) (0.571) (0.199) (0.141) (0.004) Tamil 0.025 -0.007 0.024 0.091** -0.029 (0.618) (0.816) (0.660) (0.025) 0.469 Kerala 0.040 -0.072 -0.031 0.024 -0.031 (0.515) (0.166) (0.601) (0.649) 0.538 R-square 0.039 0.089 0.011 0.034 0.014 0.071 0.009 0.050 0.004 0.044 Observations 436 429 423 416 423 416 649 640 649 640 Male Mean 0.239 0.239 0.929 0.929 0.196 0.196 0.823 0.823 0.185 0.185 Std Dev 0.427 0.427 0.257 0.257 0.398 0.398 0.382 0.382 0.389 0.389 P-value Caste 0.084 0.817 0.451 0.049 0.597 *p <0.10, **p<0.05, ***p<0.01 Source: Author’s calculations the sample of online gig workers, a greater share of women Although women working in the gig economy in India were (77 percent) than men (72 percent) have reported being able working from home before the lockdown, heightened to work from home. After controlling for various covariates, economic insecurity may have exacerbated poor coping with women were 6.2 percentage points more likely to be able to new stressors and intra-household tension. While it is too early work from home compared to men. to understand the full impact of COVID-19, impacted groups are likely to face increased unemployment, food insecurity, However, among the sub-sample of women who reported and reduced income. Job insecurity among males is known to an increase in childcare responsibilities during the lockdown, trigger higher domestic violence rates (UN Women, 2020). A 80 percent reported that they were able to work from home. recent study by Bhalotra, Kambhampati, Rawlings, & Siddique Women reporting increased childcare responsibilities as a (2020), using data from 31 developing countries, find that a result of the lockdown were 37.7 percentage points more 1 percent increase in male unemployment rate is associated likely to be able to work from home compared to men. With with an increase in the incidence of physical violence against mandatory quarantines and rising household responsibilities women by 0.50 percentage points or 2.75 percent. to continue for several months to come, these women may find it increasingly difficult to manage paid work Among the gig workers sampled for this study, only 30 percent and care work. of men and 38 percent of women reported being secure about job continuity after the pandemic. The multivariate analysis shows women to be 8.2 percentage points more likely INTRA-HOUSEHOLD TENSIONS AND VIOLENCE to feel secure about future job continuity in comparison to men, which might indicate that gig employment may provide a greater sense of security for women (Figure 2). This is There was an increase in domestic violence during consistent with previous literature which showed that women the lockdown, pointing to increased stress and intra- with flexible work arrangements were more empowered and household tensions. Women were more likely than men consequently more secure in their jobs (Subramaniam, Tan, to report feeling humiliated and experiencing sexual Maniam, & Ali 2013). violence, both before and after the lockdown. 4 | POLICY BRIEF Fig 2: Women’s Work Preferences and Security - 10 -8 -6 -4 -2 0 2 4 6 Both men and women reported an increase in domestic of domestic violence among women. Figure 4 shows that violence during the lockdown. However, both the pre- and compared to the 20-29 age group, being in the 30-39 and 40- post-COVID values were larger for women. Figure 3 shows 49 age groups reduced this probability by 10-20 percentage how incidents of domestic violence changed in the month points. Women with a university education were also less likely after the lockdown compared to the previous month. Across to experience domestic violence. In India, socio-demographic all forms of domestic violence, there was an increase to the factors, such as young age, low levels of education, poverty, order of 15-20 percent. These rates are 2-3 times higher than urban domicile, lower caste and unemployment were those reported by men in either period. These findings might previously identified as the key risk factors for domestic be partly related to the job insecurity felt by the sampled violence (Neena, et al., 2015, Kalokhe, et al., 2017). online workers. Figures 3 and 4 are consistent with past public health crises. For instance, during the Ebola outbreak in the Democratic Republic of Congo, there was an escalation in domestic Table 2: Multivariate Analysis for Job Security and Work from Home 1 2 violence, sexual violence and aggression against women in Work from home Job Insecurity communities (IRC, 2019). In Guinea, gender-based violence Female 0.0623** -0.0821*** increased by 4.5 percent during the Ebola outbreak, and (0.0299) (0.0296) led to long-term mental and emotional consequences for Age 30-39 -0.0650** -0.00319 women (Onyango & Regan, 2020). During times of crises, (0.0318) (0.0315) governments can redirect resources related to gender-based Age 40-69 -0.0225 -0.0737* violence services to disease management as was the case in (0.0445) (0.0438) Caste (Other) -0.0102 -0.0279 Sierra Leone during the Ebola epidemic. Furthermore, as a (0.0646) (0.0622) result of government lockdowns during the recent pandemic, Other Backward Caste (OBC) 0.0615* 0.0870*** access to safe spaces for survivors of domestic violence such (0.0350) (0.0336) as homes of parents and close families has also been limited Scheduled Caste (SC) 0.259*** 0.176*** (Shalu, 2020). (0.0533) (0.0557) Scheduled Tribe (ST) 0.274*** 0.254** For some women, social distancing and quarantine measures (0.0932) (0.0988) may have increased day-to-day exposure to potential University Education -0.0194 0.0459 (0.0294) (0.0294) perpetrators, limiting their ability to engage in paid work, as Tamil -0.0428 0.108*** well as their access to essential services. In the data analyzed (0.0380) (0.0373) for this note, women who reported experiencing domestic Kerala 0.0138 0.166*** violence were 33 percentage points less likely to report (0.0491) (0.0488) being able to work from home than men. Previous literature Constant 0.716*** 0.535*** suggests that intimate partner violence gets interrupted (0.0452) (0.0446) when men migrate away from home (Mobarak and Ramos Observations 907 1,057 2019), but this may not be feasible during the lockdown. A R-squared 0.045 0.048 study from Bangladesh found that social isolation and forced quarantine measures with perpetuators of domestic violence *** p<0.01, ** p<0.05, * p<0.10 negatively affects a woman’s freedom and privacy within her house. Furthermore, confinement within homes also restricts a woman’s access to social services that address Multivariate analysis of female respondents indicates that domestic violence (Jejeebhoy, et al., 2017). The online survey age and caste were statistically significant determinants analyzed in this note did not capture other potential impacts of violence on women. AUGUST 2020 | 5 Fig 3: Incidence of Domestic Violence Before and After Lockdown (self reported by women) Fig 3: Incidence of Domestic Violence Before and After Lockdown (self-reported by women) Humiliation 24% 29% Insulted 17% 19% Physical Violence 11% 13% Threatened 16% 19% Sexual Violence 11% 12% Before Lockdown 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 During Lockdown Source: Author’s calculations In comparison to Fig 4: Likelihood of Experiencing Domestic Violence Age 30-39 (Before Lockdown) (women ages 20-29) In comparison to Age 30-39 (During Lockdown) In compa Fig 4: Likelihood Experiencing of Domestic Fig Domestic 4: LikelihoodViolence of Experiencing Domestic Violence Age Fig 4: Likelihood of Experiencing Violence (women ages 20-29) Age30-39 In comparison 20-29(Before to Lockdown) Age 30 (women ages 20-29) Fig 4: Likelihood of Experiencing (women ages Violence Domestic 20-29) 30-39(Before Age30-39 Age (During Lockdown) Lockdown) Age 30 13.3PP Age30-3926.1PP 20-29 Age 20 (women ages 20-29) Violence Index Age (During Lockdown) 13.3PP Age 20-29 13.3PP26.1PP Violence Index Violence Index 14.5PP 13.3PP 26.1PP Humiliated Violence Index 14.5PP 14.5PP Humiliated Humiliated 14.5PP 17.9PP 23.6PP Insulted Humiliated 17.9PP 23.6PP 17.9PP 23.6PP Insulted Insulted 17.0PP 17.9PP 23.6PP Sexual Violence Insulted 17.0PP 17.0PP Sexual Violence Sexual Violence 17.0PP 18.4PP Sexual Violence Threatened 18.4PP 18.4PP Threatened Threatened 0 2 comparison In 4 6 to 8 10 12 14 18.4PP 16 18 20 22 24 26 28 30 Threatened ViolenceIn comparison to 0 2 Age 4 30-39 6(Before 8 Lockdown) 10 0 12 2 14 4 16 6 18 8 20 10 22 12 24 14 26 16 28 18 30 20 22 Age 30-39 (Before Lockdown) 0 2 Age 4 30-39 6(During 8 Lockdown) 10 12 14 16 18 20 22 24 26 28 30 Age 30-39 (During Lockdown) Age 20-29 24.0PP Violence Index Age 20-29 24.0PP 24.0 13.3PP Violence Index 26.1PP Index Violence 26.1PP 19.3PP 24.0PP 23.6PP Violence Index Insulted 19.3PP 23.6PP 19.3PP 23.6 14.5PP Insulted Insulted 16.1PP 19.3PP 23.6PP 22.6PP Insulted Sexual Violence 16.1PP 22.6PP 16.1PP 22.6PP Sexual Violence Sexual Violence 17.9PP 23.6PP 16.1PP 13.8PP 13.9PP 22.6PP 23.6PP Sexual Violence Threatened 13.8PP 13.9PP 13.8PP 13.9PP Threatened Threatened 17.0PP 13.8PP 13.9PP Threatened 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 0 2 4 6 to 8 10 0 12 2 14 4 16 6 18 8 2010 12 22 14 24 2616 28 30 20 18 22 Age 40-49 (Before Lockdown) In comparison Violence 18.4PP 0 2 4 Age 30-39 6 40-49 (Before 8 10 12 14 16 18 20 22 24 26 28 30 In Age comparison Age 40-49 (During to (Before Lockdown) Lockdown) Age (Before Lockdown) Lockdown) P Age Age 30-39 40-49 Lockdown) (Before Lockdown) (Before Age Age 40-49 30-39 (During (During Lockdown) Lockdown) Age 40-49 (During Lockdown) 8 10 Age Age Source: 30-39 40-49 Authors’ 12 (During calculations. 14(During 16 Lockdown) Lockdown) Note: Figure 18 shows OLS 20 Age estimates 22 20-29 for women’s 24 26 experiencing likelihood of 28 30 domestic violence, controlling for age, caste, education and location. Only the statistically 20 Age 20-29 significant coefficients are shown. 18 22 13.3PP 24 26 28 30 26.1PP 26.1PP DISCUSSION 14.5PP 24.0PP AND POLICY 24.0PP IMPLICATIONS 19.3PP 23.6PP 9.3PP 23.6PP 17.9PP 23.6PP While lockdowns and quarantine measures are key to 23.6PP wellbeing of the sampled online gig workers in India during 16.1PP 22.6PP controlling 22.6PP spread of the coronavirus, they have the 17.0PP the COVID-19 pandemic. These patterns are similar to those unintended consequences on women’s wellbeing and intra- documented in previous crises. 13.8PP 13.9PP household relations. In addition to increased care work, many 18.4PP Learning from past crises and outbreak responses can help P women are further confined to their homes and are isolated 8 10 12 14 16 18 20 22 24 26 from their support networks. The fragmented nature of gig 28 30policy planners design effective response strategies in a 18 8 20 22 work, 12 10 24 14 with coupled 26 16 28 18 30 the gendered20 distribution 22 24 26 work of care 28 30 timely manner. Particularly, policy makers need to prioritize 18 20 22 24 26 28 30 gender-based violence within their response to COVID-19, and increased domestic violence, have adversely affected the 24.0PP 24.0PP 6 | POLICY BRIEF 19.3PP 23.6PP 9.3PP 23.6PP and provide interventions to support the safety and resilience or subsidized care arrangements, investments in women of women. So far, several states in India have implemented and girls’ capabilities, and institutional responses to gender-sensitive responses to COVID-19. In Tamil Nadu, gender-based violence. protection officers now have the designated responsibility to provide support to women who have experienced gender- based violence. In Kerala, the state women’s commission has BIBLIOGRAPHY started a tele-counselling facility for women suffering from stress and anxiety (Manorama, 2020) and a WhatsApp service Abel, M., T. Byker, J.Carpenter (2020). Risk Perceptions and to report instances of violence (Economic Times, 2020). Prosocial Behavior. Mimeo. In Odisha, the police are actively following up with women who reported instances of domestic violence before the Bhagat, R. B. (2013, October 12). Conditions of SC/ST lockdown to ensure their safety during the lockdown (Orissa Households A Story of Unequal Improvement. Economic and Post, 2020). High courts in Jammu and Kashmir, Delhi, and Political Weekly, 62-66. Karnataka have directed their state governments to assess the Bhalotra, S., Kambhampati, U., Rawlings, S., & Siddique, Z. growing domestic violence reports and designate safe spaces (2020, January ). Intimate Partner Violence: The Influence of Job for women to report violence (Shalu, 2020). Opportunities for Men and Women. Policy Research Working Paper 9118. Timely data is crucial to understand the pathways that lead to gender-differentiated impacts from institutional Economic Times. (2020, April 11). Kerala government launches responses to the pandemic. For this to happen effectively, WhatsApp number to report domestic violence during lockdown. it is critical for interventions to collect sex disaggregated economic times. Retrieved from https://economictimes. data. However, lockdowns and mandatory quarantines as indiatimes.com/news/politics-and-nation/kerala-government- a result of the COVID-19 have hindered conventional data launches-whatsapp-number-to-report-domestic-violence- collection methods that require face to face interaction. during-lockdown/articleshow/75095648.cms Remote data collection through apps and tablets is a viable Hunt, A., Samman, E., Tapfuma, S., Mwaura, G., Omenya, R., option, however ensuring privacy and confidentiality will be a Kim, K., & Roumer, A. (2019). Women in the gig economy: paid challenge (UN Women , 2020). For instance, gender violence work, care and flexibility in Kenya and South Africa. data is underreported, particularly in South Asia where IRC. (2019). Everything on her shoulders: Rapid assessment on socio-cultural beliefs normalize violence (Menon, 2018). gender and violence against women and girls in the Ebola. IRC. Generalizing trends based on administrative data comes with certain limitations, including seasonality, significant events Jejeebhoy, S. J., Santhya, K. G., Acharya, R., Zavier, A. J., Pandey, such as COVID-19 policies, accuracy of administrative data, N., Singh, S. K., . . . Ojha, S. (2017). Empowering women and in addition to social stigma and fear (Peterman, O’Donnell, & addressing violence against them through self-help groups Palermo, 2020). From past experiences, it is possible that the (SHGs). Population Council. impacts of the pandemic will negatively weigh on women’s Kalokhe, A., Rio, C. d., Dunkle, K., Stephenson, R., Metheny, N., labor force participation as well as on gender-based violence Paranjape, A., & Sahay, S. (2017). Domestic violence against and abuse. women in India: A systematic review of a decade of quantitative studies. Glob Public Health, 498–51. Although various interventions are being tested by governments and development organizations in response Kasliwal, R. (2020). Gender and the Gig Economy: A Qualitative to COVID-19, the effectiveness of these interventions Study of Gig Platforms for Women Workers. Observer Research will become clearer in the coming months and years. In Foundatio, ORF Issue Brief No. 359. the meantime, potential short-term interventions can Manorama. (2020, March 29). Kerala Women’s Commission include technology-based tools to nudge behaviors and to Offer Tele-counselling Amidst Lockdown. Retrieved from remind household members to share domestic chores and https://www.onmanorama.com/news/kerala/2020/03/29/ redistribute childcare and eldercare responsibilities. Similarly, kerala-women-commission-tele-counselling-amidst-lockdown. various informational channels can be used to disseminate html information about the impact of anxiety and quarrelling and Mazza, M., Marano, G., Lai, C., Janiri, L., & Sani, G. (2020). Danger teach men, women and youth better coping mechanisms in danger: Interpersonal violence during COVID-19 quarantine. to maintain a healthy home environment. However, such Psychiatry Research. interventions should not be considered as a substitute for longer-term institutional solutions, such as state-sponsored AUGUST 2020 | 7 McLaren, H. J., Wong , K. R., Nguyen, K. N., & Mahamadachchi, Peterman, A., O’Donnell, M., & Palermo, T. (2020). COVID-19 K. N. (2020). Covid-19 and Women’s Triple Burden: VigneĴes and Violence against Women and Children What Have We from Sri Lanka, Malaysia, Vietnam and Australia. Social Learned So Far? Center for Global Development. Retrieved Sciences. Mehta, B., Mehta, S., & Kumar, A. (2020, January 17). The joys violence-against-women-and-children-what-have-we- learned-so-far.pdf dailypioneer.com/2020/columnists/the-joys-and-perils-of- the-gig-economy.html N., Oertelt-Prigione, S., & Gelder, N. V. (2020). Pandemics Menon, V. (2018). 52% Indian women think it’s okay for their and Violence Against Women and Children. Center for Global husbands to beat them. Only 42% men agree. The Print. Development . Violence Needs Rethinking. South Asia Journal. Subramaniam, G., Tan, P.-L., Maniam, B., & Ali, E. (2013). Workplace Flexibility, Empowerment and Quality of Life. Procedia- Social and Behavioral Sciences. Muralidharan, K. (2020, May 03). Across Tamil Nadu, Caste UN Women . (2020). Violence Against Women and Girls Data UN Women. (2020). COVID-19 and Ending Violence against Women and Girls. UN Women. UN Women. (2020). PREVENTION: Violence against women U., Parial , S., . . . Kataria, D. (2015). Violence against women. and girls & COVID-19. UN Women . Indian J Psychiatry. World Economic Forum. (2017). The Global Gender Gap Onyango, M. A., & Regan, A. (2020, May 10). Sexual and gender-based violence during COVID-19: lessons from Ebola. weforum.org/docs/WEF_GGGR_2017.pdf com/sexual-and-gender-based-violence-during-covid-19- lessons-from-ebola-137541 ini� ate-phone-up-programme-to-address-domes � c- violence-during-lockdown/ Contributors: Observations 15,864 8,139 7,725 R-squared 0.308 0.310 0.318 0.005 -0.005 STAY CONNECTED Mean of dep var 0.000 We gratefully acknowledge the support of the Umbrella Facility for Gender Equality (UFGE). The in control P-value: T x Hindi 0.482 0.530 0.452 = T x English 0.625 0.551 0.744 SP-value: T x Hindi World Bank Economic Review = T x math P-value: T x English 0.681 0.809 Denmark, 0.587 Germany, Iceland, the Netherlands, Norway, the Republic of Latvia, Spain, Sweden, WORLDBANK.ORG/SARGENDERLAB = T x math 4 | POLICY BRIEF